Explaining executive pay: The roles of managerial power and complexity

DISSERTATION of the University of St. Gallen, Graduate School of Business Administration, Economics, Law and Social Sciences (HSG) to obtain the title of Doctor of Business Administration

submitted by

Lukas Hengartner from Waldkirch (St. Gallen)

Approved on the application of

Prof. Winfried Ruigrok, PhD and Prof. Simon Peck, PhD

Dissertation no. 3217

Publisher: Deutscher Universitts-Verlag, Wiesbaden 2006

The University of St. Gallen, Graduate School of Business Administration, Economics, Law and Social Sciences (HSG) hereby consents to the printing of the present dissertation, without hereby expressing any opinion on the views herein expressed.

St. Gallen, June 12, 2006

The President:

Prof. Ernst Mohr, PhD

PREFACE

Preface
The idea of studying executive pay was born in a lecture on option valuation in spring 2002 at the Norwegian School of Economics and Business Administration NHH. At that time, I started wondering how companies that had granted large amounts of stock options to their executives would react to the sharp decreases in share prices that could be observed since 2000. Based on this interest, I wrote my NHH master thesis about Employee stock options and falling share prices investigating a sample of Norwegian IT companies. These companies were especially likely to both have granted substantial amounts of options and suffered from severe declines in share prices. The topic of executive pay continued to attract my interest, mainly because I felt that it offered opportunities to make a real contribution to academic researchers and policy makers alike. In summer 2002, I wrote my master thesis at the University of St. Gallen HSG. I investigated the determinants of executive stock option plan adoptions of Swiss stocklisted companies. Major results were that both large outside shareholders and CEO duality influenced whether companies would adopt stock option plans. At that time, very little information about executive pay was disclosed in Switzerland. However, the Swiss Stock Exchange SWX issued new listing guidelines that required companies to disclose detailed information on Corporate Governance starting from the annual report 2002 to be published in 2003. Part of this information was the total compensation of the highestpaid member of the board of directors and the total compensation for the top management team. With these guidelines, the way was open to empirical studies on executive pay in Switzerland, a topic that had received much attention in the US for a long time. The goal of this dissertation was twofold. On one hand, I wanted to make a contribution to the international research community interested in top executive pay. By developing and testing an innovative and broad measure of the link between managerial power and pay as well as complexity and pay, I am confident to have made such a contribution. On the other hand, this work should provide a detailed and comprehensive overview of the executive pay in Switzerland. By presenting an extensive and updated review of the literature and detailed statistical data, I hope to provide valuable input for compensation consultants, HR managers and policy makers on this highly contentious issue.

VI

PREFACE

I am indebted to my interview partners and all participants of the survey used in this work. Interview partners were Gary Steel, Head of HR and secretary to the compensation committee of ABB, David Tankel, Managing Director of New Bridge Street Consultants, Hanspeter Fssler, member of the compensation committee of Daetwyler, and Prof. Dr. Rolf Dubs, chairman of the compensation committee of Schindler. Further, I would like to thank all persons who supported me in writing this study. My special thanks go to my academic advisors, Prof. Dr. Winfried Ruigrok and Prof. Dr. Simon Peck, for their valuable guidance and suggestions as well as for the academic freedom I was granted during the preparation of this paper. I also appreciated the instructive comments of Elin Therese Dahl and Peder Greve, who reviewed drafts of the manuscript.

Zurich, June 15, 2006

Lukas Hengartner

OVERVIEW

VII

Overview
Preface ..............................................................................................................................V Overview.........................................................................................................................VII Table of Contents............................................................................................................ IX List of tables ................................................................................................................ XVII List of figures ................................................................................................................XIX List of abbreviations .....................................................................................................XXI Summary ....................................................................................................................XXIII 1 2 3 4 5 6 7 8 9 10 Introduction...............................................................................................................1 The compensation setting process ............................................................................9 Research review on executive pay...........................................................................13 Research review on director pay .............................................................................48 Are executives paid for the complexity of the job they have?................................53 Can powerful managers extract rents? ..................................................................66 Research methods..................................................................................................102 Research results.....................................................................................................119 Discussion..............................................................................................................159 Conclusions ...........................................................................................................169

Appendix - questionnaire ..............................................................................................179 References......................................................................................................................183

VIII

TABLE OF CONTENTS

IX

Table of Contents
Preface ..............................................................................................................................V Overview.........................................................................................................................VII Table of Contents............................................................................................................ IX List of tables ................................................................................................................ XVII List of figures ................................................................................................................XIX List of abbreviations .....................................................................................................XXI Summary ....................................................................................................................XXIII 1 Introduction...............................................................................................................1 1.1 1.2 1.3 1.4 2 Why study executive pay?...............................................................................1 The evolution of executive compensation research.......................................2 Literature gap and research questions ..........................................................4 Overview of the paper .....................................................................................6

The compensation setting process ............................................................................9 2.1 2.2 2.3 2.4 2.5 Purpose of compensation policies ...................................................................9 Roles and responsibilities ................................................................................9 Organization of the compensation committee.............................................10 Elements and allocation of executive pay ....................................................10 The influence of executives on their compensation.....................................11

Research review on executive pay...........................................................................13 3.1 3.1.1 3.1.2 3.1.3 3.1.4 3.1.5 The relationship between pay and performance.........................................14 The debate over size versus profits..............................................................14 Agency theory and optimal contracting.......................................................15 The pay-for-performance sensitivity ...........................................................16 Relative performance evaluation .................................................................21 Non-financial firm performance and pay ....................................................22

TABLE OF CONTENTS

3.1.6 3.1.7 3.2 3.2.1

Pay structure and subsequent performance..................................................23 The certainty-equivalent approach ..............................................................26 Other determinants of executive pay ...........................................................27 Economic explanations................................................................................27

Complexity...........................................................................................................27 The influence of risk in optimal contracting........................................................27 Managerial labor market ......................................................................................29 Investment opportunities......................................................................................30 Managerial discretion...........................................................................................31 Human capital ......................................................................................................33 Market forces .......................................................................................................34 Executives personal characteristics ....................................................................34 Stewardship theory...............................................................................................35 3.2.2 Political explanations...................................................................................35 Managerial power ................................................................................................35 Impression management ......................................................................................35 3.2.3 Social explanations ......................................................................................36 Social comparison................................................................................................36 Isomorphism ........................................................................................................36 3.3 3.3.1 3.3.2 3.3.3 3.3.4 3.3.5 3.3.6 3.3.7 3.3.8 3.3.9 3.4 3.4.1 3.4.2 3.4.3 Other effects of executive pay .......................................................................37 Stock options, dividend payments and share repurchases ...........................37 Unvested stock options, restricted stock and manager retention.................37 Equity incentives and risk............................................................................37 Compensation, M&A and divestitures ........................................................39 Long-term compensation and capital investments ......................................40 Contingent pay and expectations management............................................40 Incentive compensation and earnings management.....................................41 Stock options and litigation .........................................................................41 Crowding out ...............................................................................................41 Determinants and consequences of pay differentials..................................42 Explaining pay differentials tournament theory .......................................42 Pay differentials and performance ...............................................................43 Other determinants and consequences of pay differentials .........................44

TABLE OF CONTENTS

XI

3.5 3.6 3.7 4

Stock option repricing ...................................................................................44 Stock options versus stock grants.................................................................46 International comparison of pay levels and mix .........................................47

Research review on director pay .............................................................................48 4.1 4.2 4.3 4.4 Directors as the shareholders agents ..........................................................48 Overview of outside director compensation ................................................49 Determinants and consequences of director compensation .......................49 Summary of director pay studies .................................................................51

Are executives paid for the complexity of the job they have?................................53 5.1 5.1.1 5.1.2 5.1.3 5.1.4 5.2 5.3 5.4 5.5 5.6 5.7 5.8 Towards a framework of complexity and pay.............................................53 Defining complexity ....................................................................................53 Complexity as a determinant of pay level and structure..............................53 Prior operationalizations of complexity.......................................................54 Conceptual framework of complexity and executive pay ...........................55 Firm internationalization and executive pay...............................................56 Firm diversification and executive pay........................................................58 Firm size and executive pay ..........................................................................59 Market uncertainty and executive pay ........................................................61 Politicized environment and executive pay..................................................62 Summary of the literature on complexity-pay relation ..............................63 Literature gap and complexity hypotheses..................................................65

Can powerful managers extract rents? ..................................................................66 6.1 6.1.1 6.1.2 6.1.3 6.1.4 6.1.5 Towards a framework of power and pay.....................................................67 Defining executive power............................................................................67 Executive pay theories and executive power...............................................69 Prior studies on the link between executive power and pay ........................70 Related studies on CEO power ....................................................................72 Conceptual framework of executive power and compensation ...................74

XII

TABLE OF CONTENTS

6.2 6.2.1 6.2.2 6.2.3 6.3 6.3.1 6.3.2 6.3.3 6.3.4 6.3.5 6.4 6.4.1 6.4.2 6.5 6.5.1 6.5.2 6.5.3 6.6 6.6.1 6.6.2 6.6.3 6.7 6.8 7

Ownership power and executive compensation ..........................................76 Executive ownership....................................................................................76 Non-executive directors ownership............................................................78 Shareholder concentration ...........................................................................79 Structural power and compensation ............................................................81 Non-executive and independent directors ...................................................81 CEO duality .................................................................................................83 Presence of a compensation committee.......................................................84 Composition of the compensation committee .............................................86 Board size ....................................................................................................87 Tenure power and executive compensation.................................................88 CEO tenure ..................................................................................................88 Interdependent directors ..............................................................................90 Network power and executive compensation ..............................................91 Executive board memberships.....................................................................92 Reciprocal and social interlocks ..................................................................93 Non-executive board memberships .............................................................94 Credibility power and executive compensation ..........................................94 CEO celebrity status ....................................................................................94 Prior firm performance ................................................................................95 Expertise ......................................................................................................97 Summary of the literature on power pay relation ...................................97 Literature gap and power hypotheses........................................................100

Research methods..................................................................................................102 7.1 7.2 7.2.1 7.2.2 7.3 7.3.1 7.3.2 Development of hypotheses .........................................................................102 Sample and data...........................................................................................102 Archival data..............................................................................................102 Survey........................................................................................................104 Model specification ......................................................................................105 Regression analysis....................................................................................105 Unit of analysis ..........................................................................................106

TABLE OF CONTENTS

XIII

7.4 7.4.1 7.4.2 7.4.3 7.5 7.5.1 7.5.2 7.6 7.6.1 7.6.2 7.6.3 7.6.4 7.6.5 7.6.6 7.7 7.7.1 7.7.2 7.7.3 7.7.4 7.7.5 7.7.6 7.7.7 8

Measuring executive compensation............................................................107 Cash compensation ....................................................................................108 Share awards..............................................................................................108 Stock options .............................................................................................108 Measuring complexity .................................................................................109 Archival data..............................................................................................109 Survey data ................................................................................................110 Measuring executive power.........................................................................111 Ownership power.......................................................................................111 Structural power.........................................................................................112 Tenure power.............................................................................................113 Network power ..........................................................................................113 Credibility power .......................................................................................114 Survey data ................................................................................................114 Other variables.............................................................................................115 Firm financial performance .......................................................................115 Firm risk ....................................................................................................116 CEO age.....................................................................................................116 Anglo-American board members and top managers .................................116 Disclosure effects.......................................................................................117 Industry effects ..........................................................................................117 Consultants ................................................................................................118

Research results.....................................................................................................119 8.1 8.1.1 Descriptive statistics ....................................................................................119 Overview of the Swiss pay scene ..............................................................119 CEO compensation ............................................................................................119 Top management compensation.........................................................................120 Non-executive compensation.............................................................................122 8.1.2 8.1.3 8.1.4 8.2 Evidence on complexity variables .............................................................123 Evidence on managerial power variables ..................................................124 Other variables...........................................................................................126 Basic determinants of executive pay ..........................................................127

XIV

TABLE OF CONTENTS

8.2.1 8.2.2 8.3 8.3.1 8.3.2 8.3.3 8.3.4 8.4 8.4.1 8.4.2 8.4.3 8.4.4 8.5 8.6 8.6.1 8.6.2 8.6.3 8.6.4 9

The determinants of the level of executive pay .........................................127 The determinants of equity-compensation mix .........................................130 The influence of complexity on executive pay ...........................................131 The influence of individual complexity variables on pay levels ...............131 The influence of individual complexity variables on pay mix ..................134 Survey measures of complexity and pay levels and mix ...........................135 The complexity construct and executive pay.............................................137 The influence of power on executive pay ...................................................140 The influence of individual power variables on pay levels .......................140 The influence of individual power variables on pay mix ..........................143 Survey measures of power.........................................................................145 The power construct and executive pay.....................................................147 The overall influence of power and complexity on executive pay ...........150 Sensitivity analysis .......................................................................................155 Does the composition of the compensation committee matter? ................155 Alternative measures for complexity variables .........................................156 Alternative measures for power variables .................................................157 Alternative measure of firm performance..................................................157

Discussion..............................................................................................................159 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 9.9 9.10 Complexity and executive pay ....................................................................159 The incremental contribution of firm size .................................................160 Technological complexity and executive pay.............................................160 The components of complexity ...................................................................161 The components of managerial power .......................................................162 Interdependent directors and executive pay .............................................162 Compensation committee existence and executive pay ............................163 Managerial power and pay structure.........................................................164 CEO duality as a self-serving opportunity? ..............................................166 Board size and executive pay ......................................................................167

TABLE OF CONTENTS

XV

9.11 10

The influence of Anglo-Americans .............................................................167

Conclusions ...........................................................................................................169 10.1 10.2 Contributions to the literature....................................................................169 Summary of results......................................................................................170 Overview of executive and director compensation ...............................170 Basic determinants of executive pay .....................................................171 Complexity and executive pay ..............................................................172 Power and executive pay .......................................................................173

10.2.1 10.2.2 10.2.3 10.2.4 10.3 10.4 10.5

Implications for policy makers and practitioners.....................................175 Limitations....................................................................................................176 Future research ............................................................................................177

Appendix - questionnaire ..............................................................................................179 References......................................................................................................................183

XVI

LIST OF TABLES

XVII

List of tables
Table 1: Empirical studies on the pay-performance relation. ................................................ 20 Table 2: Overview of studies on directors pay....................................................................... 52 Table 3: Empirical studies on complexity-CEO pay level relation........................................ 64 Table 4: Empirical studies on complexity executive pay mix relation. .............................. 65 Table 5: Major empirical studies on the CEO/board power variables and CEO pay level. ... 97 Table 6: Empirical studies on CEO power pay mix relation............................................... 99 Table 7: Descriptive statistics on CEO compensation. ........................................................ 120 Table 8: Descriptive statistics on top management compensation....................................... 121 Table 9: Descriptive statistics on non-executive director compensation. ............................ 122 Table 10: Descriptive statistics on complexity variables. .................................................... 124 Table 11: Descriptive statistics of executive power variables. ............................................ 125 Table 12: Summary statistics of other variables. ................................................................. 127 Table 13: Correlation matrix of basic determinants............................................................. 128 Table 14: OLS estimation of common determinants on total executive compensation. ...... 129 Table 15: Tobit model estimates of common determinants on equity-pay mix................... 130 Table 16: Correlation matrix based on individual complexity variables. ............................ 131 Table 17: OLS estimation of individual complexity variables on compensation levels. ..... 133 Table 18: Tobit estimates of individual complexity variables on compensation mix.......... 135 Table 19: Descriptive statistics on survey complexity measures......................................... 136 Table 20: Univariate estimates of surveyed complexity on CEO compensation. ................ 136 Table 21: Univariate Tobit estimates of CEO equity-based compensation on survey complexity variables. ........................................................................................................... 137 Table 22: Correlation matrix of archival complexity variables. .......................................... 138 Table 23: Principal component analysis and scoring coefficients of archival complexity construct. ........................................................................................................... 138 Table 24: Principal component analysis and scoring coefficients of survey complexity construct. ........................................................................................................... 139 Table 25: Relationship between complexity constructs and CEO compensation. ............... 140 Table 26: OLS estimation of individual power variables on compensation level................ 142

XVIII

LIST OF TABLES

Table 27: Tobit estimates of individual power variables on executive compensation mix.. 144 Table 28: Descriptive statistics on survey power measures................................................. 146 Table 29: Univariate estimates of survey power variables on CEO compensation. ............ 147 Table 30: Principal component analysis of archival power construct.................................. 148 Table 31: Scoring coefficients of archival power construct................................................. 149 Table 32: Principal component analysis on survey power measures. .................................. 150 Table 33: Scoring coefficients of survey power construct................................................... 150 Table 34: Regressions of the archival complexity and power constructs on executive pay. 152 Table 35: Tobit regressions of the complexity and power constructs on executive pay structure................................................................................................................................ 154 Table 36: Regressions of the survey complexity and power constructs on CEO compensation. ...................................................................................................................... 155 Table 37: OLS estimation of the composition of the compensation committee on TMT compensation. ...................................................................................................................... 156

LIST OF FIGURES

XIX

List of figures
Figure 1: Dissertation overview ............................................................................................... 7 Figure 2: Determinants and consequences of executive compensation. ................................ 13 Figure 3: Conceptual framework of complexity and executive compensation. ..................... 55 Figure 4: Conceptual framework of executive power and compensation. ............................. 75

XX

LIST OF ABBREVIATIONS

XXI

List of abbreviations
CEO DOI EJPD FASB FETE FSTS GBP Ln LTIP M&A MCHF MUSD R&D SIC UK US USD S & P 500 SWX TMT Chief Executive Officer Degree of internationalization Eidgenssisches Justiz- und Polizeidepartement Financial Accounting Standards Board Foreign employees total employees Foreign sales total sales British Pound Natural logarithm Long-term incentive plan Mergers & acquisitions Million Swiss Francs Million US Dollar Research & Development Standard Industry Classification United Kingdom United States US Dollar Standard & Poors 500 Swiss Stock Exchange Top management team

XXII

SUMMARY

XXIII

Summary
This paper examines the roles of complexity and power as explanations for executive pay level and mix. It is hypothesized that complexity and power are positively associated with the level of compensation. Further, I hypothesize that complexity is positively and executive power negatively related to the proportion of executive compensation that is based on stock options and restricted stock. Despite numerous research articles investigating variables associated with complexity and executive power and their effect on executive pay, a broad conceptualization of both complexity and power with respect to the setting of executive pay is missing. This dissertation attempts to close this gap. For a sample of 199 Swiss stock-listed companies in 2002-2003, I collect a large number of archival variables to empirically test the effects of complexity and power constructs on executive compensation. I augment these results with a sample of 47 survey-based responses to obtain a richer conceptualization of executive power and firm complexity. Multiple regressions are run on an individual variables basis as well as on factoranalyzed categories. Results are broadly consistent with hypotheses. Complexity is positively associated with CEO and top management team compensation levels and equity-pay mix. Managerial power is generally positively associated with pay levels. Contrary to my hypothesis, I also find evidence that some dimensions of executive power are positively related to the proportion of equity in total compensation.

XXIV

INTRODUCTION

Introduction

1.1 Why study executive pay?

Executive compensation has attracted widespread attention in recent years and has become one of the focus topics in corporate governance (Felton, 2004). At least three reasons have contributed to this increased attention on executive compensation for academics and practitioners alike: A fascination with and sometimes lack of understanding for the high levels of CEO compensation, the importance of the CEO as the main strategic decision maker and the increased transparency of executive compensation data in many countries. First, the pay of corporate leaders has undisputedly escalated in the last few decades, while workers pay has stagnated. The average real pay for chief executive officers of S&P 500 firms increased significantly during the 1990s, growing from 3.5 MUSD in 1992 to 14.7 MUSD in 2000. Most of this increase reflects the escalation in stock options valued at the time of grant, which now constitute the single largest component of executive pay (Hall and Murphy, 2003). Average real CEO pay in the S&P 500 decreased again to 9.4 MUSD in 2002.1 Even for large stock-listed companies, these amounts are substantial. Average CEO compensation now constitutes around 8% of corporate profits for US firms (Balsam, 2002: 262; Bebchuk and Fried, 2003). Over the same period, the pay of most employees has increased only marginally, increasing the ratio of CEO pay to the average factory workers pay from 1:42 in 1980 to 1:531 in 2000 (Felton, 2004) and 1:431 in 2004 (Economist, 2005a). In Switzerland, some CEOs seem to have multiplied their total compensation in recent years, in spite of modest stock market developments (Schtz, 2005). At UBS, a large bank, top executives now earn 230 times the average pay of low-paid employees (Wittwer, 2005a: 27). The rapid increase in CEO compensation is difficult to comprehend for many people, given simultaneous corporate downsizing, employee layoffs, plant closings (Abowd and Kaplan, 1999; Dial and Murphy, 1995; Murphy, 1999; Murphy, 1997) and poor financial performance (e.g.

At the same time as average pay decreased, median pay rose, indicating that the declining average was the result of a significant fall in the earnings of a few extremely highly compensated executives (Useem, 2003).

CHAPTER 1

Krauer, 2004). In short, there is a widely held belief that top executives are overpaid (Gomez-Mejia, 1994). Anecdotal evidence indicates that CEOs of large Swiss corporations are the best paid in Europe (Schtz, 2005: 73). A second reason for the high attention paid to CEO compensation by practitioners and academics is the CEOs position at the apex of the firm. As the main strategist for the firm, his or her compensation may have a direct bearing on business decisions affecting the future of the entire organization. How a firms top managers are compensated and how much equity of their firm they hold has a significant impact on corporate strategic decisions (Sanders, 2001a; Datta et al., 2001; Bliss and Rosen, 2001), the value development of the firm (Morck et al., 1988; McConnell and Servaes, 1990) or the compensation policies for middle and lower-level managers (Gomez-Mejia, 1994). Furthermore, the structure of executive compensation mainly the large option grants has been linked to recent corporate collapses in the US. With large stock options payoffs in view, executives manipulated their books and inflated stated earnings in the search for higher share prices (Hall and Murphy, 2003; Bebchuk et al., 2001; Bolton et al., 2002; Felton, 2004; Frey, 2004; Hambrick et al., 2005). This has led experts to conclude that much may be wrong with executive compensation (Elson, 2003) and stock options in particular (Hall and Murphy, 2003) today. Third, executive pay is increasingly transparent, as companies are required to publish increasing details of remuneration data in various countries (e.g. Schildknecht, 2004: 140). The Swiss Stock Exchange SWX required its listed companies to disclose the compensation of the top management team and the board of directors in 2002. Subsequently, interest in top management pay rapidly increased and has been a regular topic in news publications. An amendment of the law to make more detailed disclosure mandatory is under way (Gerny, 2004; EJPD, 2003).

1.2 The evolution of executive compensation research

Along with the increase in executive compensation levels, the volume of related research papers has soared. Over the last several decades, hundreds of studies have been conducted on the determinants and to a lesser degree the consequences of executive compensation. Already more than a decade ago, Gomez-Mejia (1994) stated that probably no other single variable had received as much empirical attention across different business fields and related social sciences as executive compensation.

INTRODUCTION

Spurred by the separation of ownership and control (Berle and Means, 1932; Fama and Jensen, 1983) and agency theory (Jensen and Meckling, 1976), research has mainly focused on the strength of the relationship between pay and firm performance (e.g. Lewellen and Huntsman, 1970; Murphy, 1985). Conversely, managerialists (Tosi et al., 2000) were interested in whether pay was related to firm size and sales growth rather than financial performance (Baumol, 1959; Ciscel, 1974). This started the sales versus profit debate originally tested by McGuire et al. (1962), which returned inconsistent results. In a seminal study, Jensen and Murphy (1990a) showed that a change in total shareholder wealth of 1,000 dollars was associated with a change of 3.25 dollars of total CEO wealth, and concluded that the size of the relationship is too weak to provide useful incentives. Since then, numerous studies have replicated such investigations of the payperformance relationship (e.g. Lippert and More, 1994; Hall and Liebman, 1998; Murphy, 1999; Brunello et al., 2001). However, little consistency has been found for the relationship between pay and performance, partly due to differences in sample periods, methods applied, different performance measures and control variables included (Joskow and Rose, 1994). In a meta-analysis of 137 CEO pay studies, Tosi et al. (2000) documented that firm performance accounts for less than 5% of the variance, while firm size accounts for more than 40% of the variance in total CEO pay. The same authors noted that the findings of studies on executive pay as a control mechanism are remarkably inconsistent not only with theory but with each other (Tosi et al., 2000: 305). Due to these discouraging results, recent work has begun to move beyond the payperformance framework to suggest other potential factors influencing executive compensation (Barkema and Gomez-Mejia, 1998). These factors include the role of risk (Gray and Cannella, 1997; Carpenter, 2000; Aggrawal and Samwick, 1999; Miller et al., 2002), human capital (Harris and Helfat, 1997; Combs and Skill, 2003; Sanders et al., 2001), skill (Kornhauser et al., 2005), social similarity (Belliveau et al., 1996), power (Sridharan, 1996; Barkema and Pennings, 1998), CEO reputation (Milbourn, 2003), symbolism and impression management (Zajac and Westphal, 1995; Siegel and Brockner, 2005), the role of managerial discretion (Finkelstein and Boyd, 1998) and stakeholder management (Coombs and Gilley, 2005). Overall, these advances have

CHAPTER 1

shown that executive pay levels and structure are the consequence of the interplay between various firm- and CEO-specific factors. Economists and financial scholars have turned their focus towards design specifications of stock option plans (Brenner et al., 2000; Hall and Murphy, 2000; Johnson and Tian, 2000a; Archarya et al., 2000; Bettis et al., 2005), stock option repricing (Chen, 2004; Carter and Lynch, 2001; Chance et al., 2000; Challaghan et al., 2004; Chidambaran and Prabhala, 2003) and the valuation of executive stock options (Carpenter, 1998; Hall and Murphy, 2002; Johnson and Tian, 2000b; Cuny and Jorion, 1995). Furthermore, researchers have increasingly become interested in the consequences of compensation. They have studied how compensation policy affects managerial decisions such as investment and debt policy (Coles et al., 2002), risk taking (Carpenter, 2000) or corporate acquisition decisions (Datta et al., 2001) and how compensation is related to future firm performance (Sanders, 2001b; Hanlon et al., 2003).

1.3 Literature gap and research questions

This work moves beyond the basic agency-theoretic hypothesis of linking pay to firm performance. It extends the existing literature by investigating the concepts of complexity and power in the design of executive compensation contracts. Some CEOs have more complex and demanding jobs than others, and these CEOs may command a pay premium in the managerial labor market. Although complexity was early on suggested as an important determinant of CEO pay (Finkelstein and Hambrick, 1989), it has received very little theoretical and empirical attention in prior research on executive compensation. Few studies have explicitly invoked a complexity explanation for CEO pay. Moreover, operationalization of complexity has by and large been limited to firm diversification (Finkelstein and Hambrick, 1989) or a firms degree of internationalization (Sanders and Carpenter, 1998). Much remains unclear on how the notion of complexity affects executive compensation. Are executives really paid for the complexity they manage? What kind of complexity are executives paid for? Does complexity affect the structure of executive pay contracts? This dissertation attempts to shed more light on how complexity influences executive pay:

INTRODUCTION

What is the role of complexity in the determination of executive compensation level and structure? To answer this research question, I developed an integrative concept of complexity. This concept models complexity along different dimensions based on a large set of variables. The variables were collected through archival and survey data collection methods to capture both objective and subjective notions of complexity. Applying this thorough method, I have attempted to give a more comprehensive account of how complexity affects compensation level and structure. Power stands at the heart of agency theory (Fama, 1980; Jensen and Meckling, 1976). According to the agency perspective, the evolution of the public corporation has dispersed corporate ownership. This has created a separation between ownership and control and an opportunity for CEOs to maximize their personal wealth at the expense of shareholders (Berle and Means, 1932; Jensen and Meckling, 1976). To empirically test this basic agency assumption, researchers focused almost exclusively on the magnitude of interest alignment in terms of the pay-performance sensitivity. A significant and high pay-performance sensitivity would indicate that agency problems are effectively mitigated. The level of compensation, arguably an important indication of a CEOs pursuit of self-interest, has traditionally played a minor role, as agency theorists advocate that it is not how much you pay but how you pay (Jensen and Murphy, 1990b). The relatively scarce empirical literature on power has yielded ambiguous results. This is mainly due to insufficient operationalization of power variables (Grabke-Rundell and Gomez-Mejia, 2002) rather than poor theoretical ground. Better constructs and operationalization of variables are necessary to capture the full picture of how more powerful CEOs, as opposed to less powerful ones, manage to extract rents. Bebchuk and Fried (2003; 2004) argue that managerial power is the reason for the extraction of rent in recent years and decades. However, a valid construct measuring cross-sectional differences in managerial power and its influence on compensation is still missing. A call by Hambrick and Finkelstein (1996: 281) that a broader conceptualization of power is needed in the context of executive compensation is still unanswered. This dissertation attempts to close this gap by answering the following question: What is the role of executive power in the determination of executive compensation?

CHAPTER 1

I developed a conceptualization of board-CEO power grounded in theory and earlier empirical results and created a measurement method that adequately captures the multiple dimensions of executive power and its relation to compensation level and mix. I also augmented archival data with survey data, as archival data cannot adequately capture the nuances underlying executive compensation decisions (Gomez-Mejia, 1994).

1.4 Overview of the paper

The paper is structured as follows. In chapter 2, I describe the process of executive pay setting in Switzerland. Today, many Swiss companies have established a compensation committee. This committee is the body of the board mainly responsible for setting executive compensation. The influence of executive management on pay determination may vary substantially across companies. The vast literature on top management compensation is reviewed in chapter 3. Major streams of research are discussed. The most influential concept in studying executive compensation is agency theory. Due to the non-alignment of interests between shareholders and managers, performance-based compensation helps reduce agency costs. Researchers have focused on measuring how executive pay and wealth responds to concurrent changes in firm performance in determining the extent of reductions in agency costs. While significant relations between pay and performance have been detected, results were generally disappointingly low. As a result, other concepts have been studied to increase the understanding of executive pay. Human capital, risk, discretion, or managerial power have all been invoked and partially confirmed as determinants of executive pay. Academic research also investigated the effects of compensation structure and level on a firms strategic behavior. Chapter 4 contains a review of the literature on director pay. Closely linked to executive pay, non-executive director pay has only recently received more attention by academics. One reason for this lack of interest may be the implicit assumption of most studies on executive pay that the board of directors acts in the interests of shareholders. Empirical results show that the determinants of executive and director pay are remarkably similar. In Chapter 5, the notion of complexity and its application to executive compensation are discussed. Research hypotheses are formulated. While complexity has been examined as a determinant of CEO pay in earlier studies, operationalization has generally remained

INTRODUCTION

weak. Prior studies used proxies such as firm size and firm diversification to gauge the extent of complexity. This chapter develops a broad concept of complexity related to a CEOs job. In this respect, complexity is multidimensional. Firm size, diversification, internationalization, a politicized environment and market uncertainty are identified as different dimensions of complexity. It is hypothesized that complexity is associated with higher levels of pay and a higher proportion of equity-based pay out of total pay.
Figure 1: Dissertation overview

1. Introduction 2. Compensation setting process 3. Literature review on executive pay 4. Literature review on director pay

5. Are executives paid for the 6. Can more powerful managers complexity of the job they have? extract rents? 7. Research methods 8. Research results 9. Discussion 10. Conclusions
Chapter 6 contains the literature review and the development of the research gap and hypotheses with respect to managerial power. Notions of power have been frequently studied in executive compensation research. Agency theory, the dominant theoretical approach, argues that the separation of ownership and control in a large organization creates a power base for executive management. However, most prior studies have applied single measures of power, and results are often inconsistent and ambiguous. A more complete concept of power is developed. Research methods used to test the hypotheses developed in chapters 5 and 6 are discussed in chapter 7. The hypotheses are tested applying both archival data and survey data to capture a broad concept of power and complexity. I collected compensation data

CHAPTER 1

for 199 Swiss stock-listed companies in 2002 and 2003, resulting in a dataset of 398 firm-year observations. Both CEO compensation and top management team compensation are investigated. A large number of complexity and power variables were also collected for these years. In addition, I conducted a survey among compensation committee chairmen to capture more subtle and subjective elements of power and complexity. This second data set consists of 47 observations in 2004. To reduce the number of variables in regression analysis, I apply principal components factor analysis. Research results are provided in chapter 8. In general, results confirm that several dimensions of complexity are significantly associated with executive compensation levels and mix. The individual complexity dimensions size, internationalization, and politicized environment are all related to executive pay, while diversification and market uncertainty show no significant association with pay. Further, the subjective measure perceived complexity is significantly related to CEO pay. Results based on variables transformed by factor analysis show that these dimensions all load on one principal component and confirm a strong association between complexity and compensation. The fifteen power variables sort into six principal components. Several of these components, such as outside ownership power and structural power, are associated with executive compensation levels. I discuss the results in chapter 9. For instance, several dimensions of managerial power are significantly associated with compensation mix. However, the sign of the results is not consistent with the idea that more powerful managers prefer cash over equity-based compensation. One explanation may be that internal and external political constraints restrict the amount of cash compensation, and that stock options and share awards provide better opportunities for managers to extract rents. Conclusions are provided in chapter 10.

THE COMPENSATION SETTING PROCESS

The compensation setting process

The general roles of the board of directors are to set the companys strategic direction, to advise and monitor top management and to otherwise protect the interests of shareholders (e.g. Styles and Taylor, 2003; Hilb, 2005). Determining executive pay is an important task in fulfilling the boards duties (Finkelstein and Hambrick, 1996). This chapter describes institutional details of the executive pay setting process.

2.1 Purpose of compensation policies

Many companies have a charter for the board of directors and their committees. These charters outline in detail the purpose, tasks and responsibilities of the board and its committees. The purpose of compensation policies as frequently stated is that executive compensation is set such as to retain and attract executives who are needed to ensure the competitiveness and long-term success of the business. Most companies publishing a charter for the compensation committee further state the goal of establishing a link between pay and performance. For instance, the charter for the compensation committee of the UBS Board of Directors states that the compensation committee will support policies and practices to attract, motivate and retain executives, ensure competitiveness, long-term business success, shareholder interests and a strong pay-performance link. While other companies use a different wording, the contents largely remain the same. Companies that do not have or do not disclose a compensation committee charter often state the same purpose in their annual reports. This would imply that compensation committees only pay as much as needed to attract and retain the necessary individuals at the top and that they aim at establishing a strong link between pay and performance. Some companies further mention that the value of the position or individual qualification required for that position play a role in determining compensation levels. This indicates that complexity influences executive pay.

2.2 Roles and responsibilities

The compensation committees role is to either determine executive compensation or to make recommendations to the full board of directors. The board expects to adopt the compensation committees recommendations with possible modest modifications

10

CHAPTER 2

(Newman and Mozes, 1999). The compensation committee typically surveys market compensation levels, establishes performance benchmarks and salary policies, and evaluates managements performance against financial and non-financial goals. These activities, however, often depend on managements cooperation in providing background information and advice (Crystal, 1991). Even though the CEO usually is not a member of the compensation committee, he may sometimes attend its meetings. In companies that do not have a compensation committee, the full board assumes the role of determining executive pay. This is especially the case in small boards where all tasks are fulfilled by the combined board. Some Swiss companies further combine the specific responsibilities of nomination and compensation in a nomination & compensation committee. Many compensation committees use external compensation consultants to establish market conditions. These consultants typically compare the pay levels and structure of firms in the same industry and of similar size and make recommendations to the board of directors on how to compensate the CEO and his top management team. In some companies, the internal audit function regularly reviews the compensation setting process and submits a report to the board of directors.

2.3 Organization of the compensation committee

The compensation committee typically consists of three to four members of the board of directors, who are appointed for one year. A majority and sometimes all of its members are non-executive and independent. This should ensure that the objectivity required to perform their duties is not impaired. However, it is not uncommon that the CEO and other executives attend the meeting of the compensation committee without formal voting power. The head of HR occasionally functions as the secretary to the compensation committee. Furthermore, internal or external specialists may be invited to hold presentations if deemed necessary. Meeting minutes are usually made available to all members of the board of directors. Compensation committees meet less frequently than the full board, typically twice or three times a year, if circumstances do not require otherwise.

2.4 Elements and allocation of executive pay

Compensation is largely made up of five components: Salary, bonus, option grants, share grants and benefits. Salary and bonus are usually paid in cash. Bonuses are typically awarded for achievement of accounting-based performance targets of the prior year and

THE COMPENSATION SETTING PROCESS

11

are awarded once a year. Benefits include additional cash contributions to companies pension funds, company cars or other perquisites. Not all companies make use of equitybased option or shares grants. Similar to bonus payments, option and restricted stock grants are usually made once a year. Only a small minority of companies has an awardfrequency that is higher than that. The adjustment of the strike price of options, called repricing, is possible. However, little is known about repricing practice in Switzerland. Interestingly, a few companies explicitly mention that the board considers the awards given to the CEO and other top executives in past years when it determines current-year compensation. While it remains unclear what elements of compensation that would include, it indicates a certain slack in compensation levels. This suggests that once a certain level of CEO compensation has been established, for instance after a period of strong growth or profitability, a later significant downward adjustment seems unlikely. In Switzerland, employee options of stock-listed companies are taxed at grant. The taxable income is determined by the fair value of options at the time of grant adjusted for the vesting period. The fair market value is usually determined by the Black-Scholes option-pricing formula (Black and Scholes, 1973). For each year that the options are blocked for exercise, the tax value is reduced by roughly 6%. Taxation at grant means that employees are taxed for a potential benefit they may never actually be able to cash in.

2.5 The influence of executives on their compensation

Sections on executive compensation in annual reports revealed numerous instances where CEOs appeared to have direct or indirect influence upon the contracting process for their own pay. Some companies openly acknowledge in annual reports that CEOs and other corporate executives such as the head of HR helped structure their own compensation with the role of board committees potentially limited to ratifying management proposals. For example, Tamedias 2004 annual report states that the remuneration of the top management team is determined based on the proposal of the CEO to the Board of Directors. Similarly, Kaba described the role of its board of directors by writing in its 2004 annual report, it is the task of the compensation committee todetermine the compensation policies for the top management team upon proposal of the delegate to the board of directors and to approve the emoluments of the top management team... Even when CEOs cannot directly influence their own pay, they

12

CHAPTER 2

often propose the pay of their immediate subordinates, which in turn may influence their own pay, because paying regular members of the top management team more necessarily reduces the gap between the CEO and the average TMT member. This, in turn, provides a legitimate reason to increase CEO compensation to ensure internal fairness. For instance, SIGs 2004 annual report states that the level of compensation for the other members of the Group Executive Committee is setby the Nomination & Compensation Committee, based on the proposal of the CEO. In addition to these acknowledgments of direct management participation in setting the terms of compensation, many annual reports suggested the presence of conflicts of interests in the contracting process. As discussed above, a few Swiss stock-listed companies reported having CEOs or other executives who served as members of their own compensation committee. Other companies reported that non-executive directors benefited from personal consulting contracts or from diversion of company business to their principal employers. In addition to these channels for CEOs to grant favors to the directors who set their compensation, the process for recruiting and reappointing members of the board itself had long been understood to be influenced by the CEO in most companies. A central tenet of this paper is that CEOs exert influence over their board of directors and compensation committees in these and other ways, and that they exploit this power to increase the value and lower the riskiness of their compensation.

RESEARCH REVIEW ON EXECUTIVE PAY

13

Research review on executive pay

The number of research papers on executive pay has increased dramatically over the last few decades. Academics of such diverse fields as accounting (e.g. Antle and Smith, 1986), economics (e.g. Jensen and Murphy, 1990a), finance (e.g. Baker et al., 1988; Carpenter, 2000), human resources (e.g. Kostiuk, 1990), management (e.g. Barkema and Gomez-Mejia, 1998), industrial relations (e.g. Agarwal, 1981) and sociology (e.g. Allen, 1981a) have published studies on the determinants and effects of executive compensation.
Figure 2: Determinants and consequences of executive compensation.

Economic explanations
Performance

Complexity Risk Managerial labor market Investment opportunities Managerial discretion Human capital Market forces Executives personal characteristics Stewardship theory

Managerial power Impression management

Political explanations

Equity-based compensation Total compensation

Cash compensation

Executive compensation

Performance Dividend payments Stock repurchases Managerial retention Risk Mergers, acquisitions and divestitures Capital investments Expectations management Earnings management Litigation

Social explanations
Social comparison

Isomorphism

This vast body of literature has generated not only useful insights, but also many contradictory findings. Any review on executive pay will therefore have to remain incomplete. Good general reviews are provided by Gomez-Mejia (1994), Finkelstein and Hambrick (1996), Gomez-Mejia and Wiseman (1997) or Murphy (1999). Core and Guay (2003) review the literature on equity-based executive pay. Balsam (2002) provides a good introduction into and overview of executive pay. Figure 2 presents an overview of

14

CHAPTER 3

the determinants and consequences of executive compensation level and structure, as discussed in this and subsequent chapters.

3.1 The relationship between pay and performance

3.1.1 The debate over size versus profits
Prior to 1980, only a handful of studies of executive compensation were published. Most of these studies focused on whether pay was more closely tied to company size or company profits. Economists were interested in examining hypotheses derived from the traditional theory of the firm that top managers operate to maximize profits. It was contended that executive compensation (i.e. salary and bonus2) would be closely linked with profitability (e.g. Lewellen and Huntsman, 1970). According to this view, competition in the managerial labor market (Fama, 1980) and the structure of the managers compensation contracts unite managerial interests with shareholder interests. Consequently, managers act to maximize profits and shareholder wealth. To recognize elements of oligopolistic competition, an alternative managerialist hypothesis was introduced. Due to the separation of ownership in large firms, managers seek their own personal goals, such as maximization of perquisites, power and control, and they achieve these goals by maximizing sales and not profits or shareholder wealth. The managerialist hypothesis therefore stated that compensation would be more closely related to sales revenues subject to a minimum profit constraint (Baumol, 1959). This alternative view was also termed sales-maximization hypothesis. This started the sales versus profit debate. Researchers examined whether changes in executive pay were more closely related to changes in sales revenues or changes in profits. For instance, McGuire et al. (1962), Ciscel (1974) and Schmidt and Fowler (1990) among others concluded that top executive compensation appears to be driven more by organization size than performance. At the other extreme, Deckop (1988), Lewellen and Huntsman (1970) and Masson (1971) accorded firm performance a stronger role than size. These early studies suffered from severe methodological problems. For example, collinearity can make a difference in how one interprets regression coefficients, if both firm size and profits are correlated (Ciscel and Carroll,

At that time, the award of equity-based pay was still rare.

RESEARCH REVIEW ON EXECUTIVE PAY

15

1980). Also the type of firm performance measure used affects the results (Masson, 1971).

3.1.2 Agency theory and optimal contracting

Since the late 1970s, agency theory has emerged as the major theoretical approach to studying executive pay. The original foundation of agency theory, however, goes back to the early 1930s. At that time, Berle and Means (1932) argued that the owners of a company usually do not have the power to place important pressure upon management, since individual ownership interests are very small. They built their arguments on three propositions: First, economic power is concentrated with a few large corporations in each industry. Second, stock ownership of these corporations is dispersed. And third, executives manage companies without owning them. The relationship between shareholders of a publicly owned corporation and the corporations executives is a classic example of a principal-agent problem. Jensen and Meckling (1976: 308) define an agency relationship as a contract under which one or more persons (the principal(s)) engage another person (the agent) to perform some service on their behalf, which involves delegating some decision-making authority to the agent. Agency theory assumes that the agent is a self-interested rational individual and seeks to pursue personal value maximization, which may not necessarily be aligned with shareholder value maximization. There is goal incongruence between managers and shareholders. For instance, building a diversified and large company may reduce the managers risk of unemployment and increase his compensation (Murphy, 1999). However, shareholder value may not necessarily be maximized with such a strategy. Another example of agency cost is the purchase of a corporate jet or a luxury office building. Both of these actions are unlikely to result in a net increase in shareholder wealth. A further building block of agency theory is information asymmetry. If shareholders had complete information regarding the CEOs activities and the firms investment opportunities, they could design a contract specifying and enforcing the managerial action to be taken in each state of the world. In most cases, however, shareholders do not know what actions the CEO can take and they cannot perfectly observe managerial actions and investment opportunities. Therefore, information asymmetry excludes the design of a detailed action plan to be implemented by the CEO.

16

CHAPTER 3

Agency theorists argue that a compensation contract can be written that alleviates problems associated with goal incongruence and information asymmetry between managers and shareholders. The traditional agency-framework defines an optimal or efficient contract as one that maximizes the net expected economic value to shareholders after transaction costs (such as contracting costs) and payments to employees (Core et al., 2003). In this view, the board of directors sets executive pay to minimize agency costs. Compensation policy will be designed to give the manager incentives to select and implement actions that increase shareholder wealth. As actions cannot be directly specified in the compensation contract, the principal will negotiate outcome-based elements of compensation in the form of bonus payments, stock options and restricted stock awards (Eisenhardt, 1989). Equity-based compensation is a particularly effective means to solve the agency problem (Haugen and Senbet, 1981), especially as it is hard in practice to use salary and bonus to reward and penalize CEOs (Hall and Liebman, 1998). Agency theory further predicts that variable pay and equity holdings are used as complementary sources of pecuniary incentives, where their relative importance varies across firms and industries, depending on the relative costs and gains (Lambert and Larcker, 1987).

3.1.3 The pay-for-performance sensitivity

Optimal contracting suggests the alignment of shareholders and managers interests through the provision of salary revisions, bonus compensation, and equity-based compensation.3 As a measure of agency costs reduction, the sensitivity between pay and performance has developed as the most studied phenomenon in executive compensation research. Pay-performance sensitivity is defined as the dollar change in the CEOs pay or wealth associated with a dollar change in the wealth of shareholders. The underlying idea is that the existence of competition in capital markets makes the survival of the firm depend on the provision of incentive compensation, which encourages the CEO to act in the shareholders interest. Firms which fail to compensate managers in this way will face higher costs and thus will not compete successfully with firms whose managers act in the shareholders interest. Pay performance sensitivity is typically measured by tracking

Supporting incentive effects of equity compensation, the announcements of proposed changes in equity-based managerial compensation packages have been met with positive stock market reactions (Brickley et al., 1985; Larcker, 1983).

RESEARCH REVIEW ON EXECUTIVE PAY

17

current changes of firm financial performance (typically total shareholder return) to changes in CEO compensation or wealth. Murphy (1985) and Coughlan and Schmidt (1985) document that changes in executive cash compensation are positively related to current year stock price changes. While the relation between changes in executive cash pay and changes in shareholder wealth is statistically significant, the effect is small and explains little of the variance in executive compensation (Jensen and Zimmerman, 1985). Bentson (1985) establishes the importance shares and option holdings have in calculating the appropriate payperformance sensitivity. He investigates CEO wealth effects of 29 large conglomerates between 1970 and 1975 and finds that managers tend to gain and lose along with shareholders, if the changes in value in CEO share- and option holdings are included in the calculations. In a seminal article, Jensen and Murphy (1990a) provide estimates for the total wealthperformance sensitivity for a large sample of US companies between 1974 and 1986. Jensen and Murphy estimate the value-enhancing magnitude of different compensation elements such as salary revisions, cash bonus, stock options, stock ownership, and dismissal decisions. Both cash compensation and total compensation (including stock option grants and gains from exercising stock options) are positively related to firm performance. However, the authors argue that the economic significance of these compensation and wealth changes is low. The CEO receives an additional 1.35 cents of cash and 3.3 cents of total compensation for each 1,000 USD increase in shareholder wealth respectively. Their estimates of the total CEO pay-performance relation including pay, options, stockholdings and dismissal indicates that CEO wealth changes 3.25 USD for every 1,000 USD change in shareholder wealth. The results further demonstrate that the incentives generated by stock ownership are large relative to direct pay incentives. Showing a decline in real executive compensation since the 1930s and a decreased sensitivity in the pay-performance relation since then, Jensen and Murphy (1990a) use their findings to challenge the principal-agent paradigm. They conclude that these values are too low, and that boards of directors should seek to increase equity-based compensation and CEO shareholdings. In a related article, Jensen and Murphy (1990b) argue that, from a shareholder perspective, the level of compensation is of secondary importance to the appropriate mix of compensation. Compensation would be too low to attract the highly-talented individuals necessary to run big companies: The very best

18

CHAPTER 3

lawyers or investment bankers can earn substantially more than the very best corporate executives. Highly talented people who would succeed in any field are likely to shun the corporate sector, where pay and performance are weakly related, in favor of organizations where pay is more strongly related to performance and the prospect of big financial rewards more favorable (Jensen and Murphy, 1990b: 149). They further write that the resulting general absence of management incentives in public corporations presents a challenge for social scientists and compensation practitioners (Jensen and Murphy, 1990a: 262). By proposing incentive compensation, these contributions greatly influence companies adopting equity-based compensation plans, which consequently increased total executive pay.4 In the academic field, the contribution of Jensen and Murphy has spurred a vast amount of literature investigating the size of the pay-performance sensitivity. Lippert and More (1994) confirm Jensen and Murphys (1990a) results for a sample of 310 US firms between 1974 and 1988. Their total pay-performance sensitivity is 0.0035 or 3.5 USD CEO pay-change per 1,000 USD change in shareholder wealth. Haubrich (1994) studies whether principal-agent and optimal contracting theory is indeed inconsistent with the low value of pay-performance sensitivity reported by Jensen and Murphy (1990a) by calculating numerical solutions to agency models developed by Grossman and Hart (1983) and Holmstrom and Milgrom (1987). He concludes using reasonable assumptions about executive risk aversion, CEO effort and other variables, principalagent theory can yield quantitative solutions in line with the empirical results of Jensen and Murphy (Haubrich, 1994: 259). Hall and Liebman (1998) study 478 US large companies between 1980 and 1994. Investigating changes in total CEO wealth including holdings of stock and stock options they argue that there is a strong link between the fortunes of CEOs and the fortunes of the companies they manage (Hall and Liebman, 1998: 654). As changes in firm market value are often extremely large in absolute terms, even relatively low wealth-performance sensitivities may lead to very large dollar rewards and punishments for CEOs. Applying the Jensen/Murphy (1990a) statistic of CEO wealth changes, they estimate a change of 6.00 USD for every 1,000 USD change in firm value for 1994. Adjusting for increasing firm size over the

A further reason for the explosion of option grants is the 1994 reform that limited tax deductibility of executive cash pay to 1 MUSD. Options turned into an alternative form of tax-favored compensation.

RESEARCH REVIEW ON EXECUTIVE PAY

19

investigation period and the corresponding natural tendency for the Jensen/Murphy statistic to fall over time, they find that sensitivities increased fourfold between 1980 and 1994 to 12 USD. The inclusion of stock and stock option revaluations in addition to changes in direct pay increases compensation sensitivities by a factor of about 30 compared with direct pay-performance sensitivity. Murphy (1999) tests both cash payperformance sensitivity and elasticity for a sample of 500 US firms between 1990 and 1996 and reports an average sensitivity of 14 USD and elasticity of 0.263. In the UK, pay-performance sensitivities are lower than in the US. Conyon and Murphy (2000) report an average pay-performance sensitivity of 2.33 GBP per 1,000 GBP in the largest 510 UK companies. Buck et al. (2003) report a sensitivity of 1.55 GBP of CEO rewards per 1,000 GBP increase in shareholder value. Very few studies on the pay-performance relationship exist outside the US and the UK. For a sample of 48 German firms between 1968 and 1994, Conyon and Schwalbach (2000) find a cash pay-performance elasticity of 0.071. Their pay measure is per capita income of the top management team. Brunello et al. (2001) use survey data from 106 Italian companies and find a significant pay-performance sensitivity for upper- and middle-level managers. However, their estimate is considerably lower than the ones based on US data. In addition to cross-national differences, the lower sensitivity may be a result of including middle-level managers. The authors further report that the payperformance sensitivity is higher in foreign-owned firms, in listed firms, and in firms affiliated with a multinational group. Table 1 summarizes a selection of studies estimating cross-sectional differences in the pay-performance relation. These studies of the pay-performance link suffer from two major methodological weaknesses. First, there is no ex ante certainty on the time-lag of the pay-performance relationship. The arbitrary research design of regressing current-year changes in executive pay on current year shareholder wealth need not necessarily be the appropriate agency-theoretic proposition. Boschen and Smith (1995) show that CEO compensation responds to changes in firm performance over the next 4-5 years, with the cumulative response of pay to performanceroughly 10 times that of the contemporaneous response (p. 577). Second, relying on shareholder return or accounting performance measures provides an incomplete picture of how the board of directors evaluates CEO performance. Hayes and Schaefer (2000) investigate the proposition that employment contracts may be based on performance measures that are observable only to the parties

20

CHAPTER 3

of the contract (but not to the researcher). They show that unexplained variation in current compensation is positively related to future performance and that this relationship is stronger when there is more noise in the firms market and accounting returns. These results suggest that compensation is informative regarding future performance and that the board of directors takes into account subjective performance measures when setting executive pay. Together, both articles demonstrate that examining the contemporaneous relationship between pay and observable financial performance underestimates the strength of the pay-performance relationship. In short, the total payperformance relationship may be high enough to elicit suitable incentives for executives, especially when variations in equity holdings are considered.
Table 1: Empirical studies on the pay-performance relation.
STUDY Murphy (1985) Coughlan and Schmidt (1985) Jensen and Murphy (1990a) Lippert and More (1994) Hall and Liebman (1998) Murphy (1999) SAMPLE (FIRMS) 73 149 430 310 478 500 102 48 106 287 COUNTRY US US US US US US UK Germany Italy UK TIME PERIOD 1964-1981 1977-1980 1974-1986 1974-1988 1980-1994 1990-1996 1969-1995 1968-1994 1993-1996 1996-1997 COMPENSATION VARIABLES CEO total compensation CEO cash compensation CEO total compensation CEO total compensation CEO total compensation CEO cash compensation HPD* cash compensation TMT cash compensation Managerial cash compensation CEO total compensation P-P SENSITIVITY/
ELASTICITY
A

= 0.12 = 0.14 = 0.0033 = 0.0035 = 0.0060 = 0.0138 = 0.263 = 0.067 = 0.071 = 0.00024 = 0.0016

Conyon and Schwalbach (2000) Brunello et al (2001) Buck et al. (2003)

A

Measuring pay and performance in absolute values leads to interpreting the regression coefficients as sensitivities, while measuring pay in logarithms and performance in rates of return leads to interpreting the regression coefficients as elasticities. An advantage of the elasticity approach is that it is relatively invariant to firm size (Gibbons and Murphy, 1992). The primary advantage of the sensitivity approach is that sensitivities have a more natural economic interpretation. The pay-performance sensitivity represents the executives share of value creation. This sharing rate seems a natural measure of the severity of the agency problem. Elasticities have no corresponding agency-theoretic interpretation.*means highest-paid director

RESEARCH REVIEW ON EXECUTIVE PAY

21

3.1.4 Relative performance evaluation

A further limitation of the usual pay-performance research is that it does not take into account the agency theoretic suggestion that CEOs are compensated based on firmspecific returns, filtering out observable aggregate shocks over which the CEO has no control (Holmstrom, 1982). Holmstrm (1979) shows that any informative signal, regardless of how noisy it is, can be used to improve contracts, if obtained without cost. Specifically, a relative performance-based contract can remove the common shock term from an agents performance, which enables a better evaluation of the agents actions. Researchers investigating the relative performance evaluation hypothesis have typically studied the optimal contract as a function of the performance of the agents firm and the performance of a peer group. For relative performance evaluation to hold, executive pay should be positively related to firm performance, but negatively to industry or market returns (Gibbons and Murphy, 1990). Executives should be paid well only when their company outperforms related companies. Filtering out market-specific noise appears even more relevant in light of new understanding from finance theory. It suggests that market-wide movements in stock prices are to a large extent driven by psychological effects independent of fundamental values (Shiller, 2003). The few empirical tests for the presence of relative performance evaluation in executive compensation have produced mixed evidence. Supporting studies include Antle and Smith (1986), Gibbons and Murphy (1990) and Hall and Liebman (1998). In their study of 39 firms, Antle and Smith (1986) find weak evidence consistent with the theory of relative performance evaluation in 16 of a total of 39 firms. Gibbons and Murphy (1990) study a sample of more than 1,000 firms over a 13-year period and find evidence that both industry and market performance are negatively related to changes in CEO compensation. Hall and Liebman (1998) substantiate the negative link between cash pay and market returns consistent with relative performance evaluation. However, they acknowledge that changes in direct pay, which do have a relative pay component, are tiny when compared with changes in the value of stock and stock option holdings, which do not have a relative pay component. Little or no supportive evidence for the relative performance evaluation hypothesis is found in Janakiraman et al. (1992) and Main et al. (1996). Bertrand and Mullainathan (2001) argue that the lack of pay for relative performance in CEO compensation contracts can be explained by CEOs simply being rewarded for luck.

22

CHAPTER 3

While the extent of relative performance evaluation in direct compensation awards appears small, the requirement for the CEO to hold more than the preferred equity in her own firm provides effects consistent with the hypothesis of relative performance evaluation. Compulsory shareholdings expose the CEO to the firms idiosyncratic return in addition to market-specific risk and thereby filter out some of the market shocks. If the CEO were to choose, he or she would have a stock portfolio that is more reliant on the market portfolio (Core et al., 2003). Compulsory stock ownership can therefore be viewed as a substitution mechanism to direct pay in providing pay for relative performance.

3.1.5 Non-financial firm performance and pay

The pay-performance relationship has either been measured by accounting returns or shareholder returns. One possible reason for the difficulty of finding a strong link between organizational performance and CEO compensation may be the fields almost exclusive focus on a firms financial performance. If boards of directors reward CEOs for maximizing value for other stakeholders than shareholders, or based on individual performance evaluation, an exclusive focus on financial performance may underestimate the true pay-performance relationship (Gomez-Mejia and Wiseman, 1997). Coombs and Gilley (2005) investigate the relationship between stakeholder management and executive pay. Stakeholder management deals with the degree to which organizations move beyond their own needs and legal requirements to satisfy the needs of their various non-shareholding stakeholders such as employees, suppliers, customers, and individuals in the community whose primary benefit derived from the company is not from its shareholder returns (Coombs and Gilley, 2005: 827). These authors use measures of community, diversity, employee and environment performance as predictors of various measures of CEO compensation. Contrary to the effect hypothesized, stakeholder management exerts a negative influence on CEO salaries.5 Boards may also pay their executives for innovation performance. Balkin et al. (2000) use a composite index of innovation built up of R&D expenditure and number of patents and find that CEOs of more innovative firms get higher cash compensation.

Studies on the relationship between stakeholder performance and CEO pay may potentially be illspecified. Rather than receiving higher compensation in return for good stakeholder performance, accepting lower overall compensation may be an inherent element of good stakeholder management.

RESEARCH REVIEW ON EXECUTIVE PAY

23

Unfortunately, their composite measure of innovation does not disentangle the effects of R&D expenditure (input measure) and number of patents (output measure). This makes it difficult to make a statement of the causal direction between pay and performance. Further, firms may pay their executives for key performance indicators, which are different across industries. Davila and Venkatachalam (2004) report that passenger load factor is positively associated with CEO cash compensation in the airline industry. In addition to objective measures of financial, innovation and stakeholder performance, subjective performance evaluation may play a role in executive pay setting (Gibbons, 1998), as measured or objective performance differs from total contribution. However, empirical research on subjective performance evaluation is nearly inexistent. An exception is Hayes and Schaefer (1997) who report evidence consistent with the use of subjective performance assessments when boards of directors decide the salary and bonus of chief executives: Variation in the executives current cash compensation that is not explained by current financial performance measures predicts future variation in these performance measures.

3.1.6 Pay structure and subsequent performance

An interesting question in the pay-performance discussion is the one of the causal direction between compensation and firm performance. Are managers paid for higher firm performance? Or does equity-based compensation induce higher firm performance? Accounting or stock price performance usually is hypothesized to be a determinant of the level of executive compensation. And equity-based compensation predicts subsequent firm performance. That is, managers are rewarded for past performance by higher compensation levels, and incentivized by equity grants and share options to elicit higher subsequent firm performance. The main argument for giving the executive options and shares is that this will align their interests with those of the owners of the firm. As many option and equity grants now have vesting periods of several years, we may expect the stock price effect to materialize after several years. There is presently no theoretical or empirical consensus on how stock options and managerial equity ownership affect firm performance (Core et al., 2003). Studies of this issue generally take one of two underlying assumptions. The first perspective conjectures that firms and managers contract optimally and that managerial ownership levels are set, on average, at the value-maximizing level (Himmelberg et al., 1999; Core et al., 2003). From this perspective, no simple ex-ante relationship between ownership and firm

24

CHAPTER 3

performance is expected (Demsetz and Lehn, 1985; Core and Guay, 1999; Himmelberg et al., 1999). Due to the lack of continuous re-contracting, this view does not preclude the possibility that costly contracting allows incentives to periodically become misaligned or that some firms contract sub-optimally with their executives. However, this misalignment will always be of temporary nature (Core et al., 2003). Consistent with this optimal contracting view, Agrawal and Knoeber (1996) using simultaneous equations and Himmelberg et al. (1999) using firm fixed effects find no relation between firm value and managerial stockholdings. The second perspective argues that observed CEO equity ownership and incentives are too low (Jensen and Murphy, 1990a; Morck et al., 1988). A first stream of the literature has analyzed the relationship between equity ownership and performance, typically measured as Tobins Q.6 Habib and Ljungqvist (2003) study firms deviations from efficient Tobins Q. Applying a stochastic frontier analysis approach, their result indicates that deviations from efficient Tobins Q are larger, when CEO stock- and option holdings are lower. This suggests that, on average, stock- and option holdings are indeed too low. In this setting, CEO equity ownership and firm performance should exhibit a positive association because higher ownership CEOs are closer to optimal incentive levels. Consistent with the hypothesis that CEO incentives are, on average, too low, evidence exists that higher ownership is associated with a higher Tobins Q. Morck et al. (1988) find that higher incentives lead to higher firm value, except among CEOs with very large fractional equity ownership. McConnell and Servaes (1990) find evidence of a positive relationship between increases in ownership and firm performance as long as managerial ownership is less than 50 percent. A positive association between managerial ownership and various measures of firm performance is also found in China (Hengartner, 2003a) and Spain (DeMiguel et al., 2004). In a related article, Abowd (1990) has tried to link incentive compensation with subsequent shareholder performance. He analyzes the effects of bonus payments on subsequent shareholder return and finds that greater use of incentive pay is positively related to next years total shareholder return and gross economic return. A general problem with investigating incentives and subsequent shareholder performance is that it is unclear to the researcher,

Tobins Q is the ratio of the market value of equity plus the book value of debt to the book value of total assets.

RESEARCH REVIEW ON EXECUTIVE PAY

25

when the share price incorporates the incentive effect created. Incentive effects documented by this type of study therefore need to be interpreted with caution. A second stream of the literature measures the effect of compensation mix on accounting performance. This eliminates measurement problems associated with stock price reactions. Empirical research has been mainly supportive of a link between pay mix and subsequent accounting performance. Mehran (1995) finds that firm performance improves measurably, when firm managers have a higher percentage of equity-based compensation. Gerhardt and Milkovich (1990) find a positive association between the use of long-term incentives and subsequent ROA for a large sample of top- and middlelevel managers. Carpenter and Sanders (2002) documents a positive association between CEO pay structure and subsequent accounting performance. Frye (2004) finds evidence that firms that provide more equity-based compensation to employees perform better. Hanlon et al. (2003) find that executive stock option grants are positively associated with future operating earnings a dollar increase in executive stock option grant is associated with an increase of 3.97 dollars in future earnings. In Westphal (1999), the proportion of CEO long-term incentive plan compensation is positively associated with subsequent firm performance. Ittner et al. (2003) find that lower than predicted option grants to CEOs, directors, and technical employees are associated with lower subsequent ROA. Sanders (2001b) does not find evidence that firms gain any performance advantage from the use of equity-based pay. However, even by using accounting measures of return, the causal direction of the relation between equity incentives and performance still cannot be resolved entirely. Rather than higher equity incentives producing better future firm performance, it may be the case that firms expecting better future performance grant more equity (Yermack, 1997). Compensation mix has also been linked to alternative measures of performance. Holthausen et al. (1995a) investigate whether the compensation of business unit heads affects innovation. They find modest evidence that increases in the division executives proportion of total compensation which is based on long-term performance has a positive impact on the divisions future innovation. Berger et al. (1997) find that managers increase leverage permanently as a value-enhancing action after they receive large incentive compensation awards. McGuire et al. (2003) find that CEO incentive compensation is associated with lower corporate social performance. This suggests that

26

CHAPTER 3

paying CEOs with stock and options reduces their performance with regard to stakeholders other than shareholders. In addition to pay mix, researchers have investigated the effect of pay-for-performance sensitivity on future firm financial performance. Abowd (1990) finds that subsequent accounting performance is significantly and positively related to pay-performance sensitivity. Mishra et al. (2000) document a positive curvilinear effect of pay-forperformance sensitivity and subsequent firm performance except for very high levels of sensitivity.

3.1.7 The certainty-equivalent approach

Despite some empirical evidence of a relationship between pay and subsequent performance, stock options recently have been criticized as an efficient means of compensation. Researchers applying a certainty equivalent approach argue that stock options are worth (much) less to the executives than they cost to the firm. To be efficient, the incentive effects of stock options need to at least outweigh the discount that executives assign to options. Lambert et al. (1991) point out that the managers entire portfolio of wealth is important for contracting purposes. Their study models a firm that gives a risky contract to a manager who has initial wealth correlated with the stock price. The authors show that the risk-averse and undiversified manager has a certainty-equivalent value for the contract that is less than the risk-neutral firms value of the contract (or, equivalently, the cost of the firm of providing the contract). For power utility functions, Lambert et al. (1991) show that the managers valuation of an option can be less than 50 percent of the BlackScholes value when the manager is constrained to hold 50 percent of his wealth in firm stock. Generally, the valuation is lower for managers who are more risk-averse and less diversified. Replicating the analysis in Lambert et al. (1991), Hall and Murphy (2002) argue that stock options are an inefficient means of conveying compensation. The intuition is that paying compensation in stock or options to a risk-averse executive can be more costly to the firm than delivering to the executive the same value in cash. Their analysis is based on the assumption that the risk-averse executive cannot rebalance his portfolio following an increase in its value, and therefore will discount this value increase. Further, Jenter (2002) argues that the incentive effects of options are reduced substantially by the fact that the managers marginal utility covaries negatively with the pay-for-performance sensitivity. Tian (2004) develops a certainty equivalent approach to

RESEARCH REVIEW ON EXECUTIVE PAY

27

measure incentive effects of executive stock options. In his model, granting stock options over a certain threshold will actually decrease the average option delta, that is reduce rather than increase the incentive to increase the stock price. Despite these recent advancements, it is an unresolved question whether the benefits arising from option compensation outweigh its costs.

3.2 Other determinants of executive pay

Even after a decade of increasing equity-based compensation spurred by Jensen and Meckling (1990a, 1990b), agency costs are a salient and persistent problem. In a metaanalysis, Tosi et al. (2000) find that firm performance explains less than 5% of variation in CEO pay. These results seem inconsistent with the view that compensation contracts are set optimally and other explanations are necessary to explain executive pay and why it is not set optimally. In contrast to the economic perspective of an efficient contract, a number of scholars and practitioners either implicitly or explicitly take the view that contracting arrangements are largely inefficient and do not minimize agency costs (for example Morck et al., 1988; Crystal, 1991, and Jensen, 1983). Even strong proponents of optimal contracting acknowledge that contracts may temporarily be inefficient (Core et al., 2003). This subchapter discusses factors beyond performance, which have been put forward as explanations for CEO pay level and structure. The following subchapter 3.3 discusses alternative consequences of pay mix.

3.2.1 Economic explanations

Complexity Complexity forms a major part of this paper. I will review the literature on the complexity-pay relationship at length in chapter 5. The influence of risk in optimal contracting Principal-agent theory frames the design of incentive plans for corporate managers as a trade-off between incentives and risk-sharing. To induce managers to act in the interest of shareholders, managerial wealth is linked to measures of corporate performance. At the same time, it is unclear to what extent corporate performance is a good measure of a CEOs performance. Greater business risk makes it difficult to determine whether variations in organizational performance are due to inferior managerial performance or factors outside of managers control (Antle and Smith, 1986). This noise in the available measures of performance burdens managers with inefficient levels of risk.

28

CHAPTER 3

Therefore, the optimal contract balances the benefits of increased managerial effort with the deadweight loss from inefficient risk sharing. That is, although the corporation wants to provide the executive with incentives to increase firm value, the board would not want to put too much risk on the executive, as it may affect decision making in a suboptimal way (Mishra et al., 2000). In equilibrium, the corporation must increase remuneration to compensate the executive for bearing that risk. To reduce executive risk, the corporation may want to use multiple measures of firm performance. The measures used and the weights allocated to those measures should take into account the informativeness and noisiness of those measures in evaluating executive performance (Holmstrm, 1979). Empirical studies show that the weights placed on performance measures in contracts are consistent with this trade-off between noise and informativeness. The relative weight placed on a performance measure in a compensation contract is an increasing function of its informativeness with respect to the agents action (Lambert and Larcker, 1987; Aggrawal and Samwick, 1999; Sloan, 1993). Bloom and Milkovich (1998) show that business risk is negatively related to the use of bonus in managerial cash compensation contracts and that managers receive higher levels of base pay for accepting greater business risk. Consistent with efficient risk sharing, firms with higher risk tend to have lower pay-performance sensitivities in various occupations (Prendergast, 2002). Not all kind of risk is equally important in determining the optimal contract. Nonsystematic risk and systematic risk seem to matter differently to incentives. The standard agency theory assumption that principals are risk-neutral can be modified into that principals are also risk averse to the market risk, but that they are risk-neutral to firmspecific risk. First, it is costly for both shareholders and CEOs to bear the systematic risk, but shareholders have a clear cost advantage in bearing the non-systematic risk, because shareholders can diversify, while CEOs typically hold a large undiversified position in the firm. Second, CEOs can potentially trade the market portfolio to adjust their exposure to the market risk, but for incentive reasons they need to maintain the firm-specific risk. Lewellen et al. (1987) find that although the proportion of stock in the compensation package decreases with the systematic risk of the firm, the Beta7, it increases with the variance of stock returns or unsystematic, firm-specific risk, consistent

Beta indicates the sensitivity of a firms stock price to market-wide movements.

RESEARCH REVIEW ON EXECUTIVE PAY

29

with an efficient risk-sharing hypothesis. Jin (2002) also finds that firm-risk, but not market risk, is the driving force of the negative relation between risk and incentives. That is, compensation committees take into account firm-specific risk factors, but not market-wide risk factors, when deciding upon executive compensation contracts. In addition to the variance of stock return, capital structure is a commonly used measure of firm-specific risk. The higher the debt-assets ratio, the higher is the potential threat of bankruptcy for the firm. Ryan and Wiggins (2001) and Bryan et al. (2000a) find a negative relation between the use of stock options and the debt-to-equity ratio. Ittner et al. (2003) report a negative relationship between leverage and equity grants and CEO equity-pay mix. Frye (2004) finds leverage to be negatively related to the percentage of equity-based compensation granted to employees. Managerial labor market The existence and functioning of a managerial labor market is an important condition for the establishment of an optimal compensation contract (Fama, 1980). Central to the perspective that a managerial labor market exists is the idea that the external labor market establishes certain compensation ranges for top executives (Roberts, 1959) and that the marketplace can efficiently evaluate how well the executive is doing, so that his or her compensation cannot deviate too much from what the market considers appropriate. In case it does deviate, market forces will assure ex post settling up, that is, the manager will be commensurately compensated at his or her next job (Fama, 1980). An efficient working of the managerial labor market exerts pressure on firms to compensate their managers based on performance, because corporate performance provides important information for the managerial labor markets evaluation of the companys management (Fama, 1980). The performance of the firm thus enters calculations of each managers outside opportunity wage. As each manager depends on the performance of managers above and below him, he will exert internal monitoring of performance. The theory further contends that if performance is not rewarded with higher remuneration, the firm will lose the best talent. In short, the managerial labor market view states that agency problems attributed to the separation of ownership and control will be resolved, as managers compensation will automatically be revised based on their past performance. As the marketplace can efficiently evaluate how well the

30

CHAPTER 3

executive is doing, his or her compensation cannot deviate too much from what the market considers appropriate. The effects of the managerial labor market are relatively intractable to gauge because the markets boundaries are so diffuse. It is unclear how many candidates exist for available positions. Considerations of managerial labor markets have been limited to conceptual discussions of empirical phenomena and have not been the focus of empirical investigation themselves. For instance, Jensen and Zimmerman (1985) invoke a managerial labor market perspective to explain a positive relationship between compensation and firm stock price performance and between poor stock price performance and turnover. Based on the ideas of a managerial labor market, Fee and Hadlock (2003) show that superior stock price performance increases the demand for an executives services. New hiring grants are positively associated with prior performance. Agarwal and Walking (1994) find that acquisition attempts occur more frequently in industries where CEOs have positive abnormal return compensation. Consistent with Famas (1980) notion of ex post settling up, postbid compensation changes of managers retained after an acquisition attempt are negatively related to several measures of their prebid abnormal compensation in their studies. Also invoking a managerial labor market explanation, Agrawal and Knoeber (1998) find that CEOs coming from outside the focal firm are paid better than CEOs coming from inside the firm. Murphy and Zabojnik (2003, 2004) argue that the managerial labor market helps explain the increase in CEO pay over the last decades. In their view, general management skills, which are transferable across companies and industries, have become more important for the CEOs job. Furthermore, modern computer technologies have facilitated quicker acquisition of firm-specific skills. Together, these trends suggest that general management skills have become relatively more important compared with firm-specific knowledge. Under the assumption that transferable skills are priced in the managerial labor market while firm-specific skills are not, equilibrium CEO pay is expected to rise. Investment opportunities There is a relatively high degree of information asymmetry between managers and shareholders in growth firms. Information asymmetry arises because managers have private information about the value of future projects and also because it is more difficult to observe managerial effort in growth firms (Smith and Watts, 1992). As a result, greater potential for managerial opportunism and higher shareholder-manager agency

RESEARCH REVIEW ON EXECUTIVE PAY

31

costs exist in growth firms. Executive compensation arrangements in growth firms are therefore expected to differ systematically from those in non-growth firms. Because the selection of investment projects is assumed to command a higher equilibrium wage than the supervision of existing assets-in-place, executives of growth firms should get higher compensation. And due to monitoring difficulties associated with investment opportunities growth firms will select compensation packages that emphasize incentive (rather than fixed) compensation. Empirical research shows that the level and mix of executive compensation vary with the investment opportunity set of the firm. Smith and Watts (1992) and Gaver and Gaver (1993, 1995) show that firms with more growth options have higher executive compensation. Smith and Watts (1992) show that firms with more growth options have a greater use of bonus and stock option plans. Gaver and Gaver (1995) find that firms with more growth options pay a larger proportion of their total compensation in the form of long-term incentives. Baber et al. (1996) find stronger associations between pay and performance and more use of market, rather than accounting-based performance indicators, for high-growth firms. Managerial discretion Closely linked to the investment opportunities-hypothesis, but somewhat more comprehensive, is the managerial discretion view of executive compensation. Managerial discretion refers to the potential impact of managers on organizational outcomes (Hambrick and Finkelstein, 1987). Finkelstein and Boyd (1998: 179) define it as latitude of action in making strategic choices. The concept essentially defines to what extent organizational outcomes are within the control of its top managers. Highdiscretion contexts increase potential CEO impact on organizational outcomes because the constraints common to managing organizations are generally less severe under these conditions.8 Thus, high-discretion contexts increase both the potential marginal product of CEOs and the riskiness of the CEOs job. As a result, the absolute amount of CEO compensation is expected to be higher the greater the level of discretion. Further, the impact of a good manager will further differ more substantially from the impact of a

The concept of managerial discretion as defined here does not rely on any assumption whether CEO choices will result in positive or negative outcomes. In contrast, agency theory implies that the freedom of high-discretion environments will promote non-profit-maximizing choices by the CEO (Jensen and Meckling, 1976).

32

CHAPTER 3

bad manager, when there is more freedom in making strategic choices. Therefore, it is more important in high-discretion contexts to link compensation to performance. A frequent approach to measuring managerial discretion has been the extent of growth opportunities as a manifestation of a high-discretion environment. Empirical studies find support for a positive relationship between future growth opportunities and the level of equity-based compensation (Lewellen et al., 1987; Gaver and Gaver, 1993; Mehran, 1995) and stock options (Bryan et al., 2000a; Guay, 1999; Smith and Watts, 1992; Gaver and Gaver, 1995). Other proxies of discretion are R&D intensity and capital expenditure intensity. These proxies provide equally supportive results for the level of cash compensation (Eaton and Rosen, 1983) and stock option compensation mix (Ryan and Wiggins, 2001). Rajagopalan and Finkelstein (1992) find that firms with more discretionary strategic orientations offer greater compensation and that high discretion environmental periods are associated with greater pay. Mueller and Yun (1997) gauge the extension of managerial discretion by low return on investments and show that managers in high-discretion firms earn higher salaries. Inter-industry differences in compensation have also been explained by substantial industry-level differences in managerial discretion. Balkin and Gomez-Mejia (1987) find that the association between performance and compensation is higher in high-technology firms, which tend to have higher levels of discretion (Hambrick and Abrahamson, 1995). For selected years between 1966 and 1986, Smith and Watts (1992) find that US firms in regulated industries (insurance, gas and electrical utilities, and commercial banking) make less frequent use of executive stock option plans. Studying the deregulation of the banking industry Crawford et al. (1995) find an increase in the pay-performance sensitivity after deregulation of the banking industry, suggesting that as the CEOs input becomes more important to a firms success or as managerial discretion is increased, pay-performance sensitivity is expected to increase (Crawford et al., 1995). Bryan et al. (2005) find that electric utilities firms increased CEO cash pay-performance sensitivity, but not equity-based pay-performance sensitivity, after deregulation in 1992. Yermack (1995) and Bryan et al. (2000a) confirm that the regulated US electric utilities industry less heavily relies on stock option awards. In an important contribution, Finkelstein and Boyd (1998) operationalize the concept of discretion through various dimensions. These dimensions are level of industry regulation, investment opportunities, product differentiability, industry concentration and

RESEARCH REVIEW ON EXECUTIVE PAY

33

demand instability. Using a LISREL model to represent discretion as a latent construct, they find support for their hypothesis that managerial discretion is positively associated with CEO total compensation and long-term compensation. Human capital Executive compensation can also be determined by the executives accumulated knowledge and skills (Agrawal, 1981). Human capital derives from the experience and background of a manager and is an important source of compensation to the extent that it is recognized and valued in a firm. Human capital theory (Becker, 1975) identifies several employee attributes associated with lifetime income. Some of the most important of those attributes reflect investments in training, such as formal education and on-thejob training. The theory suggests that the number of years of education and labor market experience individuals have explain much of the variance in their pay levels. It also specifies that cognitive ability, a strong predictor of job performance, increases pay because the more capable acquire training at less cost. Although human capital is often invoked to explain compensation, it has seldom been formally operationalized in empirical research on executive pay. Examples of relevant human capital factors used in empirical research are managerial experience, education and tenure. Agarwal (1981) reports a significant association between job-related experience and executive compensation. Finkelstein and Hambrick (1989) find that a CEOs general management experience is related to his bonus (but not total cash compensation or salaries). Gerhard and Milkovich (1990) find that salary levels of topand middle-level-managers increase with human capital investments measured as years of education, years of labor market experience, length of firm tenure and length of job tenure. Harris and Helfat (1998) argue that managers are compensated for their accumulation of firm-specific skills and that they demand a compensation for the loss of return to this firm-specific skill when they change jobs and employers. Consistent with this theory, they report significant salary and bonus premiums for external CEO successors, especially when they lacked industry-specific skills. Several studies have investigated the effects of CEO tenure on compensation from a human capital perspective, typically yielding insignificant results (Deckop, 1988; OReilly et al., 1988, Hambrick and Finkelstein, 1995). Carpenter et al. (2001) find that CEO international experience is positively related to total pay levels. Fisher and Govindarajan (1992) report that the compensation of business unit heads is positively associated with years of

34

CHAPTER 3

education. Mayers and Smith (1992) use CEO age as an indicator of the level of specific human capital acquired by the executive, but they find no indication that this variable is related to CEO compensation. While a few variables yield significant associations, these studies indicate that human capital is unlikely to account for much variance in executive pay at least if measured with conventional measures of human capital. A potentially promising avenue for future research may be to use measures of interpersonal and social human capital, which is an equally important dimension of human capital. A general limitation of this type of research is that a human capital interpretation suffers from unresolved causality problems (Gomez-Mejia, 1994). To the extent that firms use measures of executive skills and knowledge to legitimize a given pay level, a relationship between human capital and compensation level may merely be a self-reinforcing effect between proxies of human capital and pay justifications. Market forces Market forces may also play an important role in compensation setting. One potential determinant is product market competition. Hermalin (1994) shows formally how increased competition affects the strength of the incentives provided in equilibrium. Beiner et al. (2005) and Cunat and Guadalupe (2005) find that intensive product market competition is associated with stronger incentive compensation for managers. Agrawal and Knoeber (1998) suggest the threat of takeover as a further market-based determinant of executive compensation. They argue that the threat of takeover has two opposing effects on executive pay, the competition effect in the market for managers, which reduces compensation, and the risk effect which increases compensation. Their results suggest that the net effect of a threat of takeover is an increase in executive pay. Executives personal characteristics Many studies have assumed that all executives will react equally to the compensation package. However, individual executive characteristics are likely to be important variables when designing a compensation package. For instance, Gomez-Mejia and Balkin (1989) show that individuals with a low risk propensity and a low tolerance for ambiguity do not react well to a pay mix that emphasizes variable compensation. Designing a riskier compensation package to an executive who is naturally inclined to be risk averse and who has a strong desire for certainty may induce this person to become very conservative when making strategic choices. These issues have received very little

RESEARCH REVIEW ON EXECUTIVE PAY

35

attention in the executive pay literature. One reason may be that data on executive characteristics such as risk aversion or total wealth are not readily available and difficult to obtain. However, including executive characteristics in studies of executive pay may provide useful avenues for future research. Stewardship theory Agency theory assumes that individuals are primarily interested in pursuing their selfinterest. Stewardship theory opposes this view of the individual. Under stewardship theory, the CEO essentially wants to do a good job and to be a good steward of corporate assets (Donaldson and Davis, 1991). It follows that there is no inherent, general problem of executive motivation, as suggested by agency theory. Stewardship theory and agency theory result in opposing recommendations in regards to corporate governance arrangements. For stewardship theory, organizational structure is supposed to facilitate effective action by the executive and to help him to formulate and implement plans for high corporate performance. For instance, this situation is attained more readily where the CEO is also the chairman of the board. While stewardship theory has been portrayed as the other side of the spectrum opposing agency theory, it has never been tested empirically to directly explain executive compensation or been used as an underlying theory. A promising point of departure is Kidder and Buchholtz (2002), who develop propositions between pay-performance, golden parachutes, the CEOs relational psychological contract and managerial stewardship.

3.2.2 Political explanations

Managerial power The relative power of the CEO and other executives over the board of directors is an important aspect in explaining executive compensation and a major focus of this study. I will review the literature on the power-pay relationship at length in chapter 6. Impression management Executive compensation may also be influenced by the degree to which the CEO, under particular performance conditions, successfully manages the expectations and reactions of investors, directors and other important stakeholders of the firm. Siegel and Brockner (2005: 2) investigate self-handicapping, an activity with the purpose to receive less unfavorable evaluations in the case of negative performance information and more favorable evaluations in the case of positive performance information. They find that

36

CHAPTER 3

CEOs who engage in claimed external handicapping are compensated more favorably than CEOs who did not engage in claimed external handicapping.

3.2.3 Social explanations

Social comparison Individuals have a need to evaluate their own opinions, attributes, and abilities, and select as comparison others who are similar in some way. OReilly et al. (1988) argue that this social comparison process leads compensation committee members to rely on their own experience and that of similar others. OReilly et al. (1988) and Main et al. (1995) find that CEO compensation is significantly higher when the outside salary of a firms compensation committee member or other non-executive directors was greater than the one of his CEO. OReilly et al. (1988) further document that compensation committee members salary in their own firms was highly related to executive salary levels for sample firms. A problem with social comparison theory is that it provides little guidance on which actor to emphasize as the relevant actor for comparison (Hambrick and Finkelstein, 1996: 278). A second limitation is that the results are also consistent with class hegemony theory (Gomez-Mejia, 1994), arguing that executives share a commonality of interests. Through interlocking directorates and homogeneous boards composed of CEOs, joint executive objectives and preferences are pursued as a class across different firms. Isomorphism Finkelstein and Hambrick (1996) suggest isomorphic pressures to conform to pay norms as a promising area of inquiry. They propose that board members who have affiliations with other firms may diffuse the adoption of certain pay practices to focal companies. Isomorphism also predicts differences across industries in compensation practices, which have been firmly established in the empirical literature (e.g. Rajagopalan and Prescott, 1990; Hambrick and Finkelstein, 1995). While industry effects are consistent with an isomorphism hypothesis, they do not give a full test of this view. There are other explanations than isomorphism managerial discretion, for instance for why compensation differs across industries. Other than providing an explanation for industry effects, isomorphic pressures have received little attention in executive pay research.

RESEARCH REVIEW ON EXECUTIVE PAY

37

3.3 Other effects of executive pay

The amount of research addressing the question of how executive compensation affects managerial decisions and firm outcomes is minimal in comparison with the large volume of work on pay-performance relations. More recently, however, academics have become more interested in studying the consequences of incentive compensation beyond firm performance, some of which are unintended and negative.

3.3.1 Stock options, dividend payments and share repurchases

The fact that the value of stock options decreases with the level of dividend payments suggests that option holders can have incentives to reduce dividend payments. Empirical research supports this view. The dividends pay-out ratio is negatively related to equitybased compensation (Lewellen et al., 1987) and stock options (Fenn and Liang, 2001). Lambert et al. (1989) document lower dividend payments following the initiation of option plans. Fenn and Liang (2001) further show that options lead to higher share repurchases. These results suggest that managerial option compensation and holdings are associated with a substitution of share repurchases for dividend payments.

3.3.2 Unvested stock options, restricted stock and manager retention

Unvested stock options and restricted stock are frequently put forward as a means of retention of valuable managers. The argument goes that executives have an incentive to stay with the firm in order to be able to exercise options or sell shares after the blocking period ends. So far, Fee and Hadlock (2003) is the only paper that investigates the retention hypothesis. However, they find no evidence of this contention.

3.3.3 Equity incentives and risk

Risk has been identified as a determinant of executive pay. However, the relationship between compensation and risk may be more complicated. It is possible that the two variables influence each other simultaneously. The value of a stock option is a convex function of the underlying share price. The higher the share price, the higher the value of the option. And the sensitivity of this relationship increases with a higher share price. Theory therefore posits that stock options offer incentives to risk-averse managers to invest in high-risk high-return projects on behalf of risk-neutral shareholders (Jensen and Meckling, 1976). Guay (1999) argues that in order to effectively control agency conflicts between shareholders and managers, it is important to consider not only the slope but also the convexity of the wealth-performance relation. Convexity refers to the sensitivity

38

CHAPTER 3

of managers wealth to the volatility of equity value. Since convexity in an incentive scheme generates a positive relation between an executives wealth and firm risk, compensation components with convex payoffs can induce risk-averse managers to invest in valuable risk-increasing projects. He finds that stock options, but not common stockholdings, significantly increase the convexity of the wealth-performance relation and thus are valuable instruments to inducing managerial risk-taking. Tian (2004) applies a certainty equivalent framework to valuing executive stock options, and shows that stock options create incentives to decrease idiosyncratic risk but to increase systematic risk. Several studies find a relationship between compensation and risk. Rajgopal and Shevlin (2002) examine whether executive stock options provide managers in the oil and gas industry with incentives to invest in risky projects. They find that stock option incentives lead to higher future exploration risk taking. A stream of research has investigated the effects of equity-based compensation on firm risk. Barron and Waddell (2003a) model the executives role in the principal-agent framework as two-fold, involving an effort choice and a decision-criterion choice. They show that stock options grants in place of restricted stock provide leverage to the firm in inducing executive effort. However, the potential cost of stock options grants in their model is that bad plans are adopted too frequently by the agent. They also find that stock options grants are associated with more extreme either very low or very high subsequent shareholder returns. Berger et al. (1997) find that leverage increases after managers receive large incentive compensation awards. Leverage not only is a measure of firm risk, it also influences return on existing equity capital (John and John, 1993) and the voting power of managers equity stakes. Therefore, managers may increase the debt-assets ratio in response to equitycompensation. Empirical results on the pay-performance relation and subsequent risk levels are ambiguous. While a few studies (e.g. Demsetz and Lehn, 1985; Core and Guay, 2001) find a positive relationship between firm risk and pay-performance sensitivity, other studies document a negative relationship (e.g. Garen, 1994, Aggarwal and Samwick, 1999; Himmelberg et al., 1999; Palia, 2001). Again, a reason for the ambiguous results may be the different properties of systematic and non-systematic measures of risk.

RESEARCH REVIEW ON EXECUTIVE PAY

39

3.3.4 Compensation, M&A and divestitures

Compensation strategy also has important effects on corporate strategic decisions such as acquisitions and divestments. First, compensation strategy influences the propensity of firms to engage in acquisition and divestiture activities. Sanders (2001b) finds that the level of option pay is strongly associated with the propensity of firms to engage in business portfolio churning. When executives are paid with options, their firms are much more likely to buy other companies and divest corporate divisions. Alternatively, executives paid primarily with cash are more likely to invest internally. These results are also consistent with options increasing risk taking. Bliss and Rosen (2001) document that CEOs with a higher proportion and level of stock-based compensation are less likely to make an acquisition. This stands in contrast to the result of Sanders (2001b). In addition to compensation structure, pay levels also appear to influence acquisitions activity. Agarwal and Walking (1994) find that acquisition attempts occur more frequently in industries where CEOs have positive abnormal compensation. Mergers and acquisitions may also be attractive to managers as prospects of higher pay. Bebchuk and Grinstein (2005a) report that firm expansions are generally associated with increases in CEO compensation. Kroll et al. (1990) report evidence that firms are more likely to engage in mergers and acquisitions if executive pay is more closely linked to firm size. Bliss and Rosen (2001) find that bank mergers increase net CEO compensation9, even if mergers cause the acquiring banks stock price or subsequent operating performance to decline, as is typical after a merger announcement. Harford and Li (2005) find that even in mergers where bidding shareholders are worse off, bidding CEOs are better off three-quarters of the time. This shows that boards tend to reward managers for growth and that a manager desiring a quick increase in compensation will have a strong incentive to make acquisitions. Second, compensation strategy has a bearing on the quality of mergers and divestitures undertaken. Datta et al. (2001) document a strong positive relation between acquiring managers equity-based compensation and stock price performance around and following acquisition announcements. These authors further show that managers with larger equity-based compensation pay lower acquisition premiums and acquire targets with higher growth opportunities. Tehranian et al. (1987) report favorable market
9

Increase in compensation minus value effects arising from stock and option holdings.

40

CHAPTER 3

reaction to voluntary sell-off announcements for firms with long-term performance plans as compared to stock market response for firms without such plans. Furthermore, compensation policy affects other strategic decisions such as liquidation or outsourcing decisions. Mehran et al. (1998) document that the likelihood of voluntary liquidation and the resulting enhancement in shareholder value increases with the extent of equity-based CEO compensation. Hall (2005) finds that both the level of stock option compensation and option pay mix influence decisions for large-scale information technology outsourcing.

3.3.5 Long-term compensation and capital investments

Two studies show that executive compensation packages designed to reward the executive for long-term accomplishments tend to be associated with investments that have a futuristic payoff. Larcker (1983) finds that companies that have adopted a performance plan tend to have higher subsequent capital investments. Hoskisson and Hitt (1988) report a positive association between long-term incentive plans and R&D spending.

3.3.6 Contingent pay and expectations management

Bolliger and Kast (2004) investigate the relationship between CEO pay and expectations management, which defines a state in which firm management manipulates analysts to issue relatively low forecasts. These authors find that CEO compensation components strongly influence the propensity of managers to engage in expectations management strategies. The sensitivity of the option portfolios to stock price movements, the value of in-the-money exercisable options, the value of shares held and the cash bonus are all associated with analyst guidance. In firms with fixed stock option awards schedules, there is evidence that CEOs make opportunistic voluntary disclosure decisions that maximize their stock option compensation. Aboody and Kasznik (2000) find that CEOs delay good news and rush forward bad news around option award dates. In firms without a fixed award schedule, CEOs receive stock option grants shortly before the release of favorable quarterly earnings news (Yermack, 1997) and followed by significant abnormal stock returns (Lie, 2005). This shows that CEOs opportunistically time stock option awards around news announcements in firms with unscheduled awards. There also exists evidence that

RESEARCH REVIEW ON EXECUTIVE PAY

41

executives use their inside information to time the exercise of stock options (Bartov and Mohanram, 2003).

3.3.7 Incentive compensation and earnings management10

A common form of executive bonus plans is to pay a linear piece rate for fiscal year profits above some pre-determined lower bound. Many plans also include a cap on bonus payments. Healy (1985) shows that managers manipulate earnings to maximize their bonuses. He finds that managers whose firms are above or below the bounds of the bonus program choose income-decreasing strategies that defer firm income to a time when it has more impact on their personal compensation. Managers, whose firms are within the bounds of the bonus program, however, choose income-increasing strategies. Holthausen et al. (1995b) show that income-increasing is much more prevalent than income-decreasing, as many plans do not include an upper bound. In addition, this study finds that CEOs may be trying to smooth their firms profits from one year to the next. There is evidence that earnings manipulation is related to executive share ownership and equity compensation (Peng and Rell, 2005; Bergstresser and Philippon, 2003). This result suggests that fast-vesting options that represent a substantial component of current executive pay may give managers perverse incentives to self-deal in pursuit of higher option payouts. A natural conclusion is that equity-based compensation contracts should focus more on restricted stock or options with a very long vesting period.

3.3.8 Stock options and litigation

Another unintended consequence of stock option compensation is that it increases partly through an increase in earnings manipulation the likelihood of shareholder class action litigation (Peng and Rell, 2005). Such lawsuits often cause large costs to the company.

3.3.9 Crowding out

Crowding out theory posits that to understand the concept of work in firms (and elsewhere) two types of motivation must be taken into account: extrinsic and intrinsic. Persons are intrinsically motivated if work is performed for works sake (Frey, 1997: 429). Extrinsic preferences are activated from outside the person concerned and may

Earnings management is the use of discretionary accruals in accounting to manipulate reported corporate earnings.

10

42

CHAPTER 3

either be positive (financial work incentives) or negative (threat of wage cuts or dismissal). Financial economists see the individual as a self-interest agent. As a result, principal-agent theory focuses on the principals need to monitor and control the agents. Crowding out theory argues that the use of extrinsic incentives, as proposed by principalagent theory, may crowd out or destroy intrinsic work motivation under identifiable conditions (Frey, 1997). High intrinsic motivation is especially present when agents get an interesting task, have a personal relationship with the principal and have extensive participation possibilities. Frey and Oberholzer-Gee (1997) provide evidence for the contention that monetary compensation tends to undermine an individuals sense of civic duty. Studying the reaction of Swiss residents to monetary compensation offered for the acceptance of a nuclear waste repository, a negative crowding-out effect reduces the effectiveness of introducing monetary compensation. Crowding out theory has been offered as an explanation for the weak observed link between pay and performance, as put forward by agency theory. Principals who are aware that payments contingent on performance tend to crowd out their agents work morale are reluctant to extensively use incentive pay. The discussion on mostly unintended and costly consequences of incentive compensation highlights the importance for compensation practitioners to carefully outweigh the potential costs and benefits in designing incentive compensation contracts. Much remains to be learned about the potential side effects of managerial compensation strategies. On one hand, it seems necessary to substantiate consequences of incentive suggested in the literature. For instance, it would be interesting to empirically test crowding out theory on executive pay. On the other hand, researchers may find other unintended consequences of incentive pay.

3.4 Determinants and consequences of pay differentials

3.4.1 Explaining pay differentials tournament theory
Tournament theory (Lazear and Rosen; 1981) is a branch of agency theory, wherein large differences in pay between the CEO and the next highest level serve to motivate the executives occupying that level by promoting competition among them. Therefore, the compensation of a corporations top executive may well exceed any measure of his marginal product and yet be economically efficient. Under tournament theory executive compensation is set to provide incentives, not to the executives themselves, but rather to

RESEARCH REVIEW ON EXECUTIVE PAY

43

their subordinates. The incentive is for lower level executives to work hard, win the tournament, and be promoted to receive that higher level of compensation. Rosen (1986: 714) claims payments at the top have indirect effects of increasing productivity of competitors further down the ladder. An underlying assumption of this theory is that top managers are motivated to work hard and perform well because they wish to win each successive tournament (for promotion to the next hierarchical position) and the prize in form of a greater level of compensation. Because, for a given firm, the tournament for the CEO position is the final one in which a top manager will be involved in, a particularly large pay increment is provided as incentive for that job. Empirical studies on the tournament model have provided inconsistent results. On the one hand, evidence exists that compensation differentials between hierarchical levels increase as one moves up the organization (Conyon et al., 2001; Lambert et al., 1993) and are negatively related to the promotion rate (Leonard, 1990). On the other hand, the effect of the number of contestants in the tournament as an indicator of the scale of the model remains unclear. While OReilly et al. (1988) report that the number of vicepresidents is negatively related to CEO compensation, Main et al. (1993) and Conyon et al. (2001) find a positive association between the number of contestants in the tournament and compensation.

3.4.2 Pay differentials and performance

As with compensation mix, a central issue in pay distribution models is understanding how differences in pay distributions influence individual and organizational performance. It is disputed whether a hierarchical model where pay is more widely dispersed and less equal across pay levels or a compressed pay model in which pay is less dispersed and is spread more equally across jobs or individuals induces higher performance. Proponents of more compressed pay distributions argue that such they are beneficial for group performance because they may trigger feelings of fairness and common purpose, foster cooperative, team-oriented behavior, and support common goal orientation. Some supportive empirical evidence can be found. An early contribution is Cowherd and Levine (1992). They find that when the gap in pay between top executives and those lower in the organization is particularly high, firm performance is lower. Studying baseball players and teams, Bloom (1999) shows that more compressed pay dispersions are positively related to multiple measures of organizational and individual performance.

44

CHAPTER 3

Further, pay inequality is positively associated with individual performance of highlypaid players, while it is negatively related to individual performance of lower-paid players. In a recent contribution, Siegel and Hambrick (2005) report a negative relationship between top management team pay inequality and performance. Tournament theory as discussed above provides the basis for the prediction that the less compressed pay distributions increase performance. Supporting this contention, Main et al. (1993) report a direct positive association between pay inequality in top management teams and firm performance. Clearly, more empirical research needs to be conducted to make a more definite statement of the effect of pay distribution on performance.

3.4.3 Other determinants and consequences of pay differentials

Little knowledge exists about determinants and consequences of pay differentials other than performance. Finkelstein and Hambrick (1996) suggest diversification strategy as a potential determinant of pay dispersion. In related diversifiers, there is a stronger need for collaboration and coordination, which suggests a more compressed pay structure. Bloom and Michel (2002) find supportive evidence that diversification positively influences pay dispersion with top management teams. A more dispersed pay structure may also be important for organizations subject to high levels of uncertainty. The expected result is that a more dispersed pay structure will ensure that high-performing managers with appropriate human capital will occupy the uppermost positions, which is particularly important in uncertain environments, where managerial discretion is substantial (Finkelstein and Boyd, 1998). Bloom and Michel (2002) find a significant and positive association between investment opportunities and TMT pay dispersion. Bloom and Michel (2002) further document evidence that managerial tenure is lower and turnover higher in organizations with higher pay dispersion. As is also posited by tournament theory (Lazear and Rosen, 1981), pay differentials promote the survival and retention of the most talented managers at the cost of decreasing workforce stability and increasing turnover among remaining mangers.

3.5 Stock option repricing

Finance scholars have developed an interest in the design features of equity-based compensation plans. The issue that has attracted most attention is the repricing of stock

RESEARCH REVIEW ON EXECUTIVE PAY

45

options. Other design features such as vesting and exercise periods or grant frequency have received very little attention. A reason for this may be that these design features exhibit little variation across firms. Resetting the exercise price on outstanding options following a decline in the companys stock price is one of the most controversial executive pay practices. The typical argument against repricing is that firms provide options to employees as a form of equity incentives, and these incentives are intended to encourage employees to take valuemaximizing actions. When the stock price rises, employees are rewarded through the increase in the value of their options. However, if options are repriced after the stock price falls, the repricing effectively removes the risk originally imposed on the executive for incentive purposes, and may be seen as a reward for poor performance. It has also been argued that repricing is unnecessary considering the long time to maturity of common stock options. Chance et al. (2000) find that over half of their sample firms would have stock options that were at the money within two years after the repricing event. Proponents of repricing offer two major explanations for such practice. One argument is that deep-out-of-the-money options no longer provide any meaningful incentives to executives (Hall and Murphy, 2000), so it is necessary to revise the strike price downwards. A second argument, often cited by high-technology companies, is that repricing is needed to retain key executives. Saly (1994) and Acharya et al. (2000) provide theoretical models of the incentive effects of repricing. Their models show that although the anticipation of resetting can negatively affect initial incentives, resetting can still be a value-enhancing practice. The advantages of resetting diminish, when managers have a relatively greater ability to influence the resetting process. The academic literature on repricing has returned inconclusive results. A major impediment to empirical research has long been the lack of adequate disclosure on repricing. Starting in 1992, the SEC has required firms that reprice executive stock options to disclose all instances of repricing over the preceding ten fiscal years in their annual financial statements. Chance et al. (2000) find that in 40% of the repricing announcements they studied, the CEO is not included in the list of executives, whose options have been repriced.

46

CHAPTER 3

The determinants of repricing are unclear. The typical repricing firm is young, rapidly growing and experiences an abrupt and relatively permanent shock to growth and profitability (Chidambaran and Prabhala, 2003). Gilson and Vetsuypens (1993) report frequent instances of repricing in a small sample of financially distressed firms that file for bankruptcy. Brenner et al. (2000) develop models for valuing options subject to repricing features, and, like Chance et al. (2000), find that besides poor share price performance firm size is about the only variable that explains the repricing decision. The existence of industry effects on repricing is unclear. While Brenner et al. (2000) and Chance et al. (2000) do not find evidence of industry effects, Chidambaran and Prabhala (2003) do report evidence of industry patterns in repricing. A few studies contribute to the discussion whether repricing is a sign of ineffective governance. Chidambaran and Prabhala (2003) find no evidence that repricing firms are associated with CEO tenure, more dispersed ownership, or low institutional ownership. Further, firms who engage in repricing tend to have smaller boards. These results do not suggest that ineffective governance arrangements drive the frequency of repricing. However, Pollock et al. (2002) report that large institutional investors negatively moderate the relationship between out-of-the-money options and option repricing and that CEO duality increases the likelihood of option repricing. These results are consistent with the idea that repricing is more frequent in firms with higher agency costs. Research is almost silent about the effects of repricing (Core et al., 2003). If repricing options reinstalls lost incentive effects as suggested by Hall and Murphy (2000) such firms are expected to outperform firms that do not reprice after the event occurs. However, Chidambaran and Prabhala (2003) observe that repricing firms never regain their historical profitability levels or growth rates.

3.6 Stock options versus stock grants

Often, stock options and stock grants are considered equally efficient in solving the agency problem. However, the recent economic literature has started investigating the relative efficiency of restricted stock and option grants. Stock options entitle the recipient to sell the underlying asset up to or at a certain date for a specified price. Restricted stock grants award a fixed number of equity shares to the executive with restrictions on the reselling of the asset. Opponents of stock options point to the low valuations by corporate managers (e.g. Hall and Murphy, 2002). Meulbroek (2001)

RESEARCH REVIEW ON EXECUTIVE PAY

47

argues that risk-averse and undiversified executives do not place enough value on the risky payout they will receive from an option to justify the cost given up by shareholders (and implicitly the incentives the option will provide).While Hall and Murphy (2002), Feltham and Wu (2001) and Jenter (2002) all find that restricted stock contracts dominate option-based contracts, Lambert and Larcker (2002) and Kadan and Swinkels (2005) favor option-based contracts in most situations.

3.7 International comparison of pay levels and mix

There exist considerable differences in the level and structure of executive compensation across the globe, partly driven by differences in regulation, culture and taxation (Cheffins, 2003; Tosi and Greckhamer, 2004). US CEOs are paid more than their international counterparts and receive a larger fraction of their pay in the form of stock options, and a lower fraction in the form of salaries, than any of their global counterparts (Murphy, 1999). Pay for US CEOs exceeds pay in other countries even after adjusting for tax rates, purchasing power, and public benefits (Abowd and Bognanno, 1995). More specifically, Conyon and Murphy (2000) report that CEOs in the US earn 45% higher cash compensation and 190% higher total compensation than CEOs in the UK and that US CEOs also have significantly higher ownership of their firms. They attribute many of the differences in pay packages to the greater use of stock options in the United States. However, the US premium is limited to the CEO. Within Europe, equity-based compensation is most common in the UK. Conyon and Schwalbach (1999, 2000) find that the use of long-term incentives is the most prevalent in the UK. A recent comprehensive international comparison is Ferrarini et al. (2003) who analyze comparative law and pay practices in 15 European Union member states. They find that reliance on high-powered equity-based contracts is more limited in continental Europe than in the UK and in the US and they attribute these differences mainly to differences in ownership structure. While large US and UK companies fit the Berle and Means (1932) description of the large corporation of dispersed equity ownership, firms in continental European countries are often controlled by large owners. These international observations support the monitoring substitution-hypothesis. Boards that supervise firms with dispersed ownership subject their CEO to equity pay to reduce equity costs.

48

CHAPTER 4

Research review on director pay

Members of the board of directors are charged with setting executive pay, and the way they compensate themselves may provide valuable information for understanding executive compensation (Yermack, 2004; Ryan and Wiggins, 2004) and other firm outcomes. Research on director pay is still in its infancy. Compared to the vast amount of research conducted on executive pay, academics have long shown relatively little interest in understanding how non-executive board members are getting paid and what effects their compensation has on the firm.

4.1 Directors as the shareholders agents

The lack of interest in director pay may partially be explained by the assumption of agency theory and optimal contracting theory that the board of directors and the compensation committee act in shareholders interest. Optimal contracting theory assumes that there is no principal-agent relationship between (outside) shareholders and the board of directors. Consequently, there are no agency costs between shareholders and board members and incentive compensation is irrelevant. The main research topic in executive pay the pay-performance relationship is not relevant in the area of director pay. Not surprisingly, Yermack (2004) documents that the pay-performance sensitivity of outside board members is several orders of magnitude lower than that of CEOs. Consistent with executive pay results (e.g. Jensen and Murphy, 1990a), most of this sensitivity stems from changes in the value of shares and option holdings. The payperformance sensitivity arising from annual changes in cash pay and equity-awards is extremely small. Yermack (2004) also shows that the incentives of each director will generally increase over his tenure due to an accumulation of equity-based retainer awards. Assuming that there is no principal-agent relationship between shareholders and board members is an unsatisfactory simplification. Directors are charged with the responsibility of managing and supervising the business and affairs of a corporation on behalf of shareholders. They are responsible for hiring, firing, evaluating, and monitoring the top management team. Thus, directors themselves are the shareholders agents and their conflicting preferences may give rise to agency costs. Incentive

RESEARCH REVIEW ON DIRECTOR PAY

49

compensation may also help align the interests of directors and shareholders. Paying directors with stock or options will force board members to recognize that their decisions will not only affect their own wealth but also the wealth of shareholders (Jensen, 1993; Shen, 2005). Director pay may have important effects on executive pay. Schmid (1997) suggests that outside directors without share ownership would not represent shareholders and their interests. Similarly, Hermalin and Weisbach (1998) determine a model of boards that suggests requiring incentive pay for directors could lead to greater monitoring. In situations of higher director equity pay, the monitoring-substitution perspective proposes a smaller need for executive incentive compensation.11 Acknowledging these issues, researchers now have increasingly become interested in the compensation of nonexecutive directors.

4.2 Overview of outside director compensation

Yermack (2004) provides an overview of director compensation in the US. Outside directors are usually paid a fixed annual retainer (and smaller additional fees for full board and committee meeting attendance). For the majority of directors, compensation is not tied to the success of the firm or their individual performance in the boardroom. This is consistent with a traditional agency theoretic view judging the role of director incentive pay as negligible. Nevertheless, since the late 1980s an increasing number of companies have paid annual retainers at least partly in equity. And while the grant values of equity awards still exhibit little variation, their subsequent fluctuation has had the effect of tying directors rewards more closely to firm performance than before.

4.3 Determinants and consequences of director compensation

Studies on the determinants of director compensation generally find that firm characteristics that explain executive compensation also explain director compensation. These findings are consistent with the notion that directors should be treated as agents of shareholders who need economic incentives to increase shareholder value. In an early

Monitoring by outside board members of executive management replaces outcome-based executive equity compensation as a monitoring mechanism.

11

50

CHAPTER 4

contribution, Schmid (1997) studies determinants of director and executive pay in 120 large German companies in 1991. He finds that the compensation of board members and that of the top management team are both related to firm size, performance and shareholder concentration. Brick et al. (2002) confirm some of the general results found for executive pay in their analysis of director compensation. They find that stock volatility and return on assets are significant predictors of total director compensation levels using pooled regression analysis. Bryan et al. (2000b) study economic determinants of outside director compensation. They find that the level of outside director stock option awards is positively related to the corporations growth opportunities and to the level of institutional shareholdings. The level of director stock grant is negatively associated with the corporations growth opportunities. There is also evidence that directors explicitly or implicitly negotiate their own pay and that the relative power of non-executive directors over the CEO affects director compensation. Ryan and Wiggins (2004) find that directors of firms with longer tenured CEOs and of firms where the CEO chairs the board an indication of less relative power of non-executive directors vis--vis the CEO receive smaller compensation packages. Further support for a relative power view is provided by Brick et al. (2002) who document lower director pay in firms where the CEO has higher stock ownership and the board has a larger proportion of inside directors. However, in contrast to the result of Ryan and Wiggins (2004), these authors find that CEO duality is positively associated with total director compensation. Director pay is further positively related to CEO overcompensation. After controlling for firm characteristics, CEO characteristics and other governance factors, Brick et al. (2002) find that director and CEO compensation are positively related. To investigate whether this compensation premium may represent excessive compensation, they regress excess compensation that is compensation not explained by standard factors on firm performance. Using changes in future Tobins Q, Brick et al. (2002) find that firm performance is negatively related to the excess compensation of directors. Additionally, they find evidence that average future return on assets is negatively related to the excess compensation of CEOs. This result suggests that directors profit from CEO pay premiums.

RESEARCH REVIEW ON DIRECTOR PAY

51

Adams (2003) uses total board compensation as a measure of board effort. Corroborating earlier findings on the determinants of executive compensation, she finds a positive relationship between firm size and board compensation and between stock price volatility and board compensation. In addition, she reports a weak negative relationship between diversification and director compensation as well as strong industry effects. However, her findings need to be interpreted with caution with respect to the determinants of non-executive director compensation. She does neither control for board size nor the proportion of executive directors on the board. Corporate financial decisions and strategies are also associated with director compensation structure. Yermack (2004) finds that in the presence of leverage, firms award less equity compensation. This result is consistent with the view that directors seek to mitigate the effects of the agency costs of debt associated with high leverage (John and John, 1993). Yermack (2004) further finds that when firms pay dividends they make less use of equity pay. This is consistent with the view that companies rely more heavily on equity-based compensation when they face a scarcity of cash. Brick et al. (2002) also provide first evidence on the determinants of director equity-pay mix. Stock volatility, firm size measured as the number of employees, R&D intensity and CEO duality are all positively related to the fraction of compensation that directors received in shares and options. And leverage is negatively associated with director equity-pay mix. The level and composition of director compensation as a determinant of executive compensation, firm performance or other outcomes has been a clearly under-researched area. A notable exception is Becher et al. (2005) who document a weak significant relationship between the use of equity-based compensation and related subsequent bank accounting performance. This indicates that director equity-based compensation induces higher corporate performance.

4.4 Summary of director pay studies

As the overview of studies on director compensation provided in Table 2 shows, the determinants and effects of outside director compensation are remarkably similar to those of executive pay. However, further studies are necessary to corroborate these early results.

52 Table 2: Overview of studies on directors pay.

STUDY SAMPLE (FIRMS) COUNTRY YEARS INDEPENDENT VARIABLES Performance Schmid (1997) 120 D 1991 Shareholder concentration Firm size Industry Growth opportunities Bryan et al. (2000b) 1788 US 19921997 Growth opportunities Institutional shareholdings Firm size R&D/assets Performance Brick et al. (2002) 740 US 19921999 Volatility CEO duality CEO ownership Internal directors Firm size Brick et al. (2002) 740 US 19921999 R&D/assets Volatility CEO duality Debt/assets Assets Adams (2003) 350 US 1998 Volatility Industry Ryan and Wiggins (2004) 1018 US 19951997 19941996 19921999 CEO tenure CEO duality Debt/assets Dividend payment % director equity-based compensation Total director compensation Director equity compensation Accounting performance Total board compensation % director equity-based compensation Total director compensation Option compensation Share compensation Option compensation Director cash compensation COMPENSATION VARIABLES

CHAPTER 4

RESULTS + + + + + + + + + + + + + + + + -

Yermack (2004)

529

US

Becher et al. (2005)

49

US

+ means a significant positive association; - means a significant negative association; 0 means no significant relationship

ARE EXECUTIVES PAID FOR THE COMPLEXITY OF THE JOB THEY HAVE?

53

5 Are executives paid for the complexity of the job they have?
5.1 Towards a framework of complexity and pay
5.1.1 Defining complexity
Complexity can be referred to as a state of being difficult to understand or explain because an entity has many different parts or sub-components. In the context of a corporation and its environment, complexity can arise from a variety of factors. Differing business segments and products, a firms organizational structure, various functional systems (e.g. HR systems, accounting systems), a multicultural workforce, different national regulations, the structure of product markets and supplier relationships, the differing demands of various stakeholder groups, a politicized environment or a high reliance on activities with an uncertain payoff such as research and development all are elements of a CEOs job complexity.

5.1.2 Complexity as a determinant of pay level and structure

There is reason to believe that executives are paid for the level of complexity they must manage. The allocation of more able CEOs to more complex firms through the managerial labor market should translate into higher pay as a compensation for a higher marginal product (Rosen, 1982). Furthermore, job complexity enters as a determinant of pay in compensation consultants work and is likely to translate into higher compensation levels for more complex positions. Marginal productivity theory holds that an individual should be paid commensurate to his or her net incremental value to an organization (Fama, 1980). If the firms marginal return to executive talent increases with complexity, an efficient market for managerial talent will generate higher compensation at more complex firms. These equilibrium effects of complexity on compensation can be illustrated in a simple matching model, where managers with higher ability are matched with positions in which the marginal return to ability is higher (Rosen, 1982; Himmelberg and Hubbard, 2000). In this model, the top manager contributes to the productivity of all other workers, and more-able managers make a larger contribution than less-able managers. The marginal return to

54

CHAPTER 5

ability realized by the firm is the increment to productivity caused by hiring a person of higher ability in a given position and is a characteristic of the firm. Characteristics that increase the marginal return to ability in the position will have two equilibrium effects: the ability of those selected to fill the position will be greater, and the compensation associated with the position will be higher. In Rosens (1982) model, this characteristic is firm size, but a similar argument can be made for other elements of complexity such as internationalization or diversification. Furthermore, job complexity plays a pivotal role in job evaluation concepts of major compensation consulting companies and recommendations of compensation consultants are likely to translate into managers compensation (Baker et al., 1988; Crystal, 1991). For instance, the job evaluation criteria used by the HR consultancy HAY Group to define the value of a particular job include factors such as the thinking challenge (problem-solving and creativity) or the thinking environment (operative and strategic thinking) inherent in a job. More complex jobs should therefore be valued more highly and be associated with higher levels of compensation. For CEO compensation, Hay Group uses company size measured in terms of employees and revenues, the technological complexity and the diversity of the market as an indication for higher pay (Hay Group, 2003). All of these variables have also been identified to represent the complexity of the managerial job by academic scholars (Henderson and Fredrickson, 1996; Sanders and Carpenter, 1998).

5.1.3 Prior operationalizations of complexity

Although complexity was early on suggested as an important determinant of CEO pay (Finkelstein and Hambrick, 1989), it has received very little theoretical and empirical attention in prior research on executive compensation (e.g. Gomez-Mejia, 1994). Few studies have explicitly invoked a complexity explanation for CEO pay. Moreover, operationalization of complexity has by and large been limited to firm diversification (Finkelstein and Hambrick, 1989), a firms degree of internationalization (Sanders and Carpenter, 1998) or both of them (Bushman et al., 2004). As much of the variation in executive pay remains unexplained (Tosi et al., 2000), a conceptualization of complexity and an empirical test of its influence on executive pay is warranted.

ARE EXECUTIVES PAID FOR THE COMPLEXITY OF THE JOB THEY HAVE?

55

5.1.4 Conceptual framework of complexity and executive pay

As I explicate in the subsequent review of the relevant literature, considerable support can be garnered for the idea that cross-sectional differences in complexity are related to executive compensation. Figure 3 depicts a conceptual framework of that relationship. The dimensions of complexity are internationalization, diversification, size, market uncertainty and a politicized environment.
Figure 3: Conceptual framework of complexity and executive compensation.

Firm

Complexity
Internationalization

Potential marginal product of CEO

Diversification Size Market uncertainty Politicized environment

Executive compensation level

Monitoring difficulty

Executive compensation mix

Environment

All dimensions of complexity are hypothesized to be associated with higher levels of compensation, as they increase the potential marginal product of an executive and thus his or her level of compensation. The higher level of complexity should match potential managers with a higher marginal productivity to such positions (Rosen, 1982). As a result, the managerial labor market assigns a higher level of compensation to more productive managers. Figure 3 also suggests that complexity leads to monitoring difficulties for the board of directors, aggravating agency costs, and that compensation committees will adapt executive pay structures to reflect their monitoring difficulties. If the CEOs job is characterized by high complexity, non-executive board members charged with monitoring the CEOs performance would find their task being made more difficult as

56

CHAPTER 5

well, as firm complexity increases the information gap between shareholders and managers (Bushman et al., 2004). In such a situation, it is difficult for board members to evaluate the appropriateness of managerial actions taken. Agency theory suggests that the board of directors awards shares and stock options as part of total executive pay to align managerial interest with that of shareholders (Eisenhardt, 1989). Each dimension of complexity will be discussed in detail in the following subchapters.

5.2 Firm internationalization and executive pay

The performance of global firms depends on their ability to cope with heterogeneous cultural, institutional, and competitive environments to coordinate geographically dispersed resources (Roth, 1995) and their ability to leverage innovation across national borders (Hitt et al., 1997). Consequently, todays multinational corporations often constitute the most complex managerial decision-making environment. Top executives of highly internationalized companies are faced with a set of more diverse markets with respect to state of development or currencies, diversified production sites with respect to culture, language and religion, resulting in a complex business environment and involving complex negotiations, or with differing national regulations and political systems (Redding and Fries, 1995). That is, top executives of multinational firms face greater market and company-related complexity. The complexity and subsequent information-processing demands12 (Tushman and Nadler, 1978; Henderson and Fredrickson, 1996) resulting from a firms degree of internationalization is likely to be reflected in the level and structure of executive compensation (Sanders and Carpenter, 1998). Information-processing demands are exacerbated as interdependencies among globally dispersed subunits become more complex through resource sharing, joint development, and value-chain integration (Egelhoff, 1982). Specifically, managerial complexity increases along with increases in the extent and dispersion of a firms dependence on its international operations. The higher a firms degree of internationalization, the greater is the level of complexity confronting its top

12

Information-processing theory holds that firms are open social systems that interface with internal

and environmental sources of complexity (Tushman and Nadler, 1978) and a firm must develop information processing mechanisms capable of dealing with the resulting complexity.

ARE EXECUTIVES PAID FOR THE COMPLEXITY OF THE JOB THEY HAVE?

57

management team. The higher level of complexity should match potential managers with a higher marginal productivity to such positions (Rosen, 1982). As a result, the managerial labor market assigns a higher level of compensation to more productive managers. Complexity arising from a firms international operations may not only be related to the level of compensation, but also to the mix of compensation between cash and equitybased components. In a complex international business environment, the task of nonexecutive directors of monitoring the performance of the CEO is made more difficult and agency costs are larger (Roth and ODonnell, 1996). Agency theory suggests that the compensation committee may therefore rely on outcome-based compensation (Eisenhardt, 1989) in form of stock options and share awards. The research record provides some support for these ideas. In the most comprehensive study on the link between firm internationalization and executive compensation, Sanders and Carpenter (1998) document a positive relationship between a composite measure of firm internationalization and total CEO compensation as well as internationalization and the proportion of long-term CEO pay. Roth and ODonnell (1996) find that the percentage of total incentive-based compensation paid to the foreign subsidiarys director increases with the foreign subsidiarys cultural distance. Recently, management researchers have started introducing firm internationalization as a standard control variable in regressions on executive compensation when studying other phenomena. Carpenter et al. (2001) investigate the impact of international assignment experience on multinational firm performance and CEO pay. Their results suggest that top managers in highly global firms are better paid and have higher proportions of performance-based pay in their compensation contract. Fatemi et al. (2003) study Economic Value Added and Market Value Added as performance determinants of CEO compensation for a large sample of US firms. Based on the ratio of overseas sales to total sales, they find weak evidence that executives of US firms with significant overseas operations enjoy somewhat higher total compensation. Their results need to be interpreted with caution, as they do not control for firm size. Miller et al. (2002) find that a composite measure of internationalization13 is significantly associated with CEO total

Their measure of internationalization includes the proportion of taxes paid to foreign governments, foreign sales out of total sales, foreign assets out of total assets, and the number of geographic segments in which the firm competes.

13

58

CHAPTER 5

pay levels. Gomez-Mejia et al. (2003) study executive pay in a sample of family-owned companies. They use a dummy variable to indicate the existence of a foreign subsidiary as a control variable. However, this variable has no effect on CEO total compensation. Carpenter and Wade (2002) study determinants of non-CEO executive compensation. Their control FSTS is positively related to top management cash compensation. The review of the relevant literature between compensation and firm internationalization can be summarized as follows. First, there is strong theoretical ground for a relationship between complexity arising from a firms degree of internationalization and the level and structure of compensation. Weak detected relationships may partly be due to poor operationalization of internationalization. Relying on a dummy variable indicating the existence of a foreign subsidiary (Gomez-Mejia et al., 2003) may not capture the entire internationalization dimension. Second, some datasets used in prior studies are outdated. The most recent observations date back to 1998. Assuming that internationalization has increased in recent years, more recent data may be especially warranted in the context of firm internationalization. Third, and linked to the point above, existing research using internationalization usually needs to rely on relatively low levels of internationalization due to the focus on US companies with a traditional reliance on a large home market. For instance, Carpenter and Wade (2002) use a sample of firms, none of which realized more than 30 percent of their revenues overseas. Studying firm internationalization for a sample of companies with a relatively small home market allowing for large variance in the internationalization variable appears to be especially warranted.

5.3 Firm diversification and executive pay

There seems to be consensus that managing a diversified firm well is a difficult task, as diversification increases the complexity of the CEOs job (Finkelstein and Hambrick, 1989; Nagar et al., 2003). The number of businesses and the relatedness of a firms businesses affect top management team coordination needs (Henderson and Fredrickson, 2001). Adding additional business segments may increase the complexity of the management task even if the new activities are small relative to the primary line of business (Rose and Shepard, 1997). At a minimum, they increase the complexity of the resource allocation decision, as multi-industry firms confront the possibility that capital will be inefficiently allocated within the firm (Stein, 1997). Managing diverse lines of businesses may require deploying a broad variety of resources and capabilities.

ARE EXECUTIVES PAID FOR THE COMPLEXITY OF THE JOB THEY HAVE?

59

Moreover, operating in more than one industry requires the CEO to understand several, potentially quite distinct, product markets. A CEO of a firm in multiple lines of business must evaluate competitive strategies for product lines that may have different suppliers, different customers, different industry structures, and different competitors. Finally, realizing potential synergies across business segments involves facilitating coordination and communication across business groups within the firm. These segments may have conflicting operational styles or corporate cultures. If the firms marginal return to executive talent increases with diversification, an efficient market for managerial talent will generate higher compensation at more diversified firms. Also, reiterating the agency-theoretic argument used above, complexity arising from firm diversification renders the board of directors job of monitoring CEO performance more difficult. In such circumstances, the compensation committee may tie a higher percentage of managerial reward to shareholder wealth creation. The empirical evidence on firm diversification and executive pay is consistent with these ideas. Rose and Shepard (1997) report that the CEO of a firm with two lines of business averages 14% more total compensation than the CEO of a similarly-sized but undiversified firm, ceteris paribus. Sanders and Carpenter (1998), Agrawal and Knoeber (1998) and Almazan et al. (2005) all show that diversified firms tend to have higher total CEO compensation. Finally, Anderson et al. (2000) report that CEOs of multi-segment firms earn higher cash compensation than CEOs of single-segment firms. Empirical evidence with regard to compensation mix is scarce. Sanders and Carpenter (1998) find no direct effect of diversification on long-term pay mix. Bushman et al. (2004) document a positive association between firm diversification and equity-based incentives out of total long-term and annual bonus incentives. Anderson et al. (2000) find that CEOs in multi-segment firms receive compensation that is less sensitive to firm performance.

5.4 Firm size and executive pay

Firm size has been established as the major determinant of CEO pay accounting for more than 40% of the variance in total CEO pay (Tosi et al., 2000). All studies I reviewed find an association between firm size and the level of compensation. However, the rationale for this association is disputed, ranging from greater responsibility associated with

60

CHAPTER 5

managing a larger firm, greater ability to pay and more hierarchical levels14 (GomezMejia and Wiseman, 1997). Early research on executive compensation attributed the relationship between firm size and CEO pay to a manifestation of managerialism. In this posture, firm size rather than performance, is the main predictor of executive pay, because greater size offers the executive several advantages: more power and prestige, less pay risk (because the incumbent has more control over firm size than performance), less employment risk (since firm size provides a buffer against business cycle effects), and a legitimate means to justify more pay at the top. However, Lambert et al. (1991) and later Gomez-Mejia and Wiseman (1997) argue that measuring the contribution of firm size to explain executive compensation is likely to suffer from an omitted variables problem. Firm size may pick up the effects of other excluded variables, which are highly correlated with firm size. Investigating changes in executive compensation related to changes in firm size, Lambert et al. (1991) find that firm size itself does not exhibit a high association with changes in compensation. Firm size may also be a reflection of the complexity of a firm (Ungson and Steers, 1984; Finkelstein and Hambrick, 1989; Gomez-Mejia and Wiseman, 1997). Larger firms are more structurally complex and culturally diverse than smaller ones and have subunits that are more specialized and differentiated. As firms become larger e.g. through mergers and acquisitions they naturally become more complex to manage unless constant efforts are undertaken to streamline divergent activities and processes. For instance, with firm size and acquisition activity, the likelihood increases that there are various human resources practices, financial reporting systems, public interest and exposure, or that the firm expands into international markets to name a few sources of complexity arising from larger companies. The prediction follows from the observation that the top position in a larger firm requires a greater marginal product because a larger firm is more complex and as such more difficult to manage (Baker and Hall, 2004). In other words, firm size may capture a sizeable amount of the complexity of the CEOs job, offering another viewpoint for the size-pay relation. This complexity explanation for the size-pay relationship challenges the use of cross-sectional studies to examine the relationship between size and compensation in tests of managerialism. Paying for size is

Simon (1957) draws attention to the influence of social norms for compensation and argues that pay differentials between hierarchical levels may account for differences in executive compensation across organizations.

14

ARE EXECUTIVES PAID FOR THE COMPLEXITY OF THE JOB THEY HAVE?

61

not a manifestation of managerial power, but a market-driven compensation for higher managerial complexity. Again, if complexity is indeed more prevalent in larger firms than in smaller ones, this should lead to greater reliance on stock-based compensation, reiterating the monitoringdifficulty argument. Relatively little attention has been given to this proposition. Firm size often does not enter the regression of compensation mix (e.g. Yermack, 1995; Bryan et al., 2000a). Nevertheless, some empirical evidence supportive of a complexity view of compensation mix can be garnered. Several studies document a positive relationship between firm size and various measures of compensation mix (Ang et al., 2002; Ryan and Wiggins, 2001; Daily et al., 1998; David et al., 1998; Barron and Waddell, 2003b).

5.5 Market uncertainty and executive pay

Market uncertainty determines the extent to which a CEO faces an environment that is predictable and stable or changing and uncertain, thereby influencing the complexity of the CEOs job (Mintzberg, 1973; Finkelstein and Boyd, 1998). Firms facing environments with volatile demand will require capable CEOs who can deal with this complexity. Large variations in demand lead to excess capacity at times. Managers may need to downsize and lay off people. This makes the managerial job very challenging. From a strategic viewpoint, up-markets may require different strategies as compared with down-markets. When market conditions are good and demand is high, a company may demand a higher price for its products. Similarly, when market conditions are poor, the company may sell its products or services at a discount. Nevertheless, market strategies still need to be consistent with the products and the firms image. For instance, selling premium products at a discount may distort the products image and may therefore be detrimental to the firm in the long-run. These challenges are likely to represent complexity. Again, an efficient managerial labor market will ensure that managers with a higher marginal product will be promoted to positions with higher market uncertainty and that this is reflected in higher compensation. Also, monitoring difficulty associated with the complexity arising from demand instability may lead to greater reliance on equity-based compensation. Market uncertainty has received little attention in prior research. Finkelstein and Boyd (1998) measure demand instability by firm-level volatility of the sales growth rate. They

62

CHAPTER 5

document a positive effect on the level of CEO compensation and the proportion of longterm compensation.

5.6 Politicized environment and executive pay

Finkelstein and Hambrick (1989) suggest a politicized environment as a type of complexity some managers may have to cope with. A politicized environment is characterized by high public scrutiny and interest and manifests itself in the media. Arguably, CEOs of politicized firms may have a more complex job, if only because they have another strong stakeholder, the public, to be satisfied. Decisions made by the CEO now need to consider the potential reaction of the public to a given decision in addition to managing for the value of shareholders and employees. In other words, the interests of different stakeholders need to be considered and this increases the complexity of a CEOs job. The Swiss national airline Swiss may well illustrate this argument. In early 2005, Swiss announced another restructuring resulting in about 1,000 job losses. After the restructuring, the new airline will employ roughly 6,000 employees and have sales of around 4 billion Swiss francs. Nevertheless, the special situation of Swiss as a national symbol of pride and the federal government and as such tax payers being the principal shareholder arguably makes the CEOs job highly complex and demanding. Again, the airline may need to attract a high-caliber executive to its top position capable to deal with public pressure and scrutiny, which is associated with a pay premium. Also, higher incentive pay may be needed because of monitoring difficulty. In contrast, Richemont, a company in the luxury industry totaling sales above 5 billion CHF in 2002 and employing 10,000 staff may get less public scrutiny, as their products are only available to a select group of people. If a similar reduction of jobs at Richemont would create less public outcry, managements decision would arguably be less complex and could be based more on economic grounds. In turn, CEO compensation may be lower and less variable. The average yearly number of newspaper articles published in NZZ about Swiss in 2002 and 2003 was 525. The corresponding number for Richemont was 160. In sum, the complexity arising from a politicized environment may be reflected in higher CEO compensation and higher reliance on equity-based compensation. To my knowledge, no empirical study exists that attempts to directly test Finkelstein and Hambricks proposition that a politicized environment may influence executive

ARE EXECUTIVES PAID FOR THE COMPLEXITY OF THE JOB THEY HAVE?

63

compensation. However, three papers provide results consistent with the idea that political pressures influence executive pay. Tosi et al. (2004) find weak evidence of a positive association between the political uncertainty perceived by top managers and CEO cash pay.15 Joskow et al. (1996) analyze the relation between CEO pay and firm characteristics in the electric utility industry and conclude that political pressures constrain CEO pay levels in that industry. Dial and Murphy (1995) describe how the political pressures at General Dynamics, a defense contractor, led the company to replace a controversial bonus plan with conventional stock options. Numerous newspaper articles had harshly criticized the payment of large bonuses to top executives for their downsizing and restructuring strategies.

5.7 Summary of the literature on complexity-pay relation

Table 3 summarizes prior empirical research linking complexity variables to executive pay level. Research has been mainly conducted in the US. The studies results show that considerable empirical support can be gathered for the hypothesis that executive compensation level is related to firm complexity. Diversification, size and internationalization are well established as predictors of executive and CEO total and cash compensation levels. Market uncertainty and a politicized environment have received little attention in prior research.

However, Tosi et al. (2004) focus on industry- and economy-wide levels of political uncertainty. Top managers were asked to assess the items Government agencies are implementing changes in laws or policies which are becoming more unpredictable and The political situation of countries in which you have operations is becoming increasingly uncertain.

15

64

CHAPTER 5

Table 3: Empirical studies on complexity-CEO pay level relation.

STUDY Finkelstein and Hambrick (1989) Kostiuk (1990) Rose and Shepard (1997) Agrawal and Knoeber (1998) Sanders and Carpenter (1998) SAMPLE (FIRMS) 68 COUNTRY YEARS 1971, 76, -82, -83 1980 19851990 1987 COMPLEXITY VARIABLES Diversification COMPENSATION VARIABLES CEO cash compensation CEO cash compensation Total CEO compensation CEO cash compensation Total CEO compensation Total CEO compensation Total CEO compensation Total CEO compensation Non-CEO executive cash compensation Total CEO compensation Total CEO compensation Total CEO compensation Executive cash compensation Total executive compensation RESULTS

US

258 397 446

US US US

Sales Diversification Diversification Internationalization Diversification Internationalization Diversification

+ + + + + + 0

258

US

1992

Carpenter et al. (2001)

245

US

19941996

Carpenter and Wade (2002) Miller et al. (2002) Fatemi et al. (2003) Gomez-Mejia et al. (2003) Almazan et al. (2005)

90

US

19811985 19941998 1995 19951998 19921997

FSTS Internationalization FSTS Foreign subsidiary Industry segments

423 119 253

US US US

+ + 0 + +

1500

US

Table 4 summarizes empirical studies on the complexity executive pay mix relation. The picture is less clear than for the complexity-pay level relationship. There is evidence that diversification, size and internationalization are predictors of equity-based compensation arrangements, but these results are not well established yet. Little is known whether and how complexity arising from market uncertainty and a politicized environment affect the structure of executive compensation.

ARE EXECUTIVES PAID FOR THE COMPLEXITY OF THE JOB THEY HAVE? Table 4: Empirical studies on complexity executive pay mix relation.
STUDY SAMPLE (FIRMS) COUNTRY YEARS COMPLEXITY VARIABLES Internationalization Diversification 125 423 US US 19901994 19941998 19941997 Size Internationalization Geographic concentration Diversification COMPENSATION VARIABLES Long-term CEO pay mix Long-term CEO pay mix % CEO equitypay % variable CEO pay % TMT equity pay % TMT equity pay RESULTS + +

65

Sanders and Carpenter (1998) David et al. (1998) Miller et al. (2002) Bushman et al. (2004)

258

US

1992

+ + 0 0

784

US

5.8 Literature gap and complexity hypotheses

In sum, there is considerable conceptual and to a lesser extent empirical support for the idea that each of the dimensions of corporate and environmental complexity I focus on here is related to executive compensation. A more complex managerial job will call for greater levels of compensation. With a few exceptions (Sanders and Carpenter, 1998; Finkelstein and Hambrick, 1989), the studies I reviewed did not formally refer to job complexity, yet the underlying idea behind much of this work is wholly consistent with the meaning of the construct. What is missing, however, is a formal test of the complexity hypothesis, one that incorporates multiple aspects of managerial complexity and allows for a more definitive statement about the effects of complexity on executive compensation. Thus, I hypothesize: Hypothesis 1a: The greater the complexity of a firm and its environment, the greater the level of executive compensation. Hypothesis 1b: The greater the complexity of a firm and its environment, the greater the proportion of equity-based executive compensation.

66

CHAPTER 6

Can powerful managers extract rents?

The extant financial literature has typically examined compensation decisions from the perspective of a board of directors that attempts to establish an optimal contract in order to mitigate agency conflicts. Recent research, however, suggests that the process of determining compensation is better described as a negotiation between the board and the CEO and that the power of CEOs to influence boards provides an explanation for the lack of pay-performance sensitivity. For instance, Hermalin and Weisbach (1998) model a bargaining game in which the CEOs compensation is negotiated between the two parties. Bebchuk et al. (2002) argue that the CEOs power over the board of directors distorts optimal compensation contracts and that the existing empirical evidence better supports the bargaining model than the optimal contracting paradigm. In a large review of the executive pay literature Gomez-Mejia and Wiseman (1997: 320) suggest that executive pay is a compromise between CEO power to inflate their compensation and societal pressures on boards to limit CEO pay and that the power of CEOs to influence boards provides a better explanation for the lack of pay-performance sensitivity than alternative explanations (Gomez and Wiseman, 1997: 321). The power view of executive pay is reflected in the widespread popular belief that CEOs and other top managers are overpaid (e.g. Schtz, 2004; Wittwer, 2005a). Academics such as Gomez-Mejia and Wiseman (1997: 329) support this view and argue that the reservation wage16 is often far lower than the payment received by CEOs. This suggests that CEOs are able to extract rents on top of their reservation wage. Today, high corporate profits coupled with low investment activity (Economist, 2005c) may have helped managers to deviate large parts of firm profits to their own books, as it seems the case with cash windfalls (Blanchard et al., 1994). Academic research adds that excessively paid managers may actually manage lower-performing firms (Brick et al., 2002) and that CEOs may be rewarded for luck (Bertrand and Mullainathan, 2001).

The reservation wage is a concept from labor economics and refers to the minimum amount a person is willing to accept to hold a job or perform certain services.

16

CAN POWERFUL MANAGERS EXTRACT RENTS?

67

6.1 Towards a framework of power and pay

From a financial economics perspective, managers pay arrangements are a (partial) solution to the agency problem. This optimal contracting approach assumes that compensation schemes determined by the board of directors or its compensation committee provide managers with efficient incentives to maximize shareholder value. Major conclusions of the optimal contracting approach for existing pay practices are that compensation schemes are not sufficiently high-powered (Jensen and Murphy, 1990a) and that divergence from optimal contracting can never be an equilibrium situation (Core and Guay, 2003). Under the managerial power approach (Bebchuk and Fried, 2003) executive compensation is viewed not only as a potential instrument for addressing the agency problem but also part of the agency problem itself: Some features of pay arrangements reflect managerial rent-seeking rather than the provision of efficient incentives. Support for a power explanation of executive pay is given by the fact that compensation practitioners report huge (unexplained) variance in salary, bonuses and long-term income of executives for firms of similar size, in the same industry, and performance at similar levels (Gomez-Mejia and Wiseman, 1997).

6.1.1 Defining executive power

There exists no agreement about the appropriate definition of power (Hardy and Clegg, 1999). Management researchers differentiate between legitimate and illegitimate power. Legitimate power is a reflection of an organizations hierarchical structure. It concerns the hierarchical structure of offices and their relation to each other (Hardy and Clegg, 1999). In the context of boards and the CEO, corporate structure clearly assigns higher formal power to the board over its CEO, with the ultimate power to hire and dismiss the CEO. Because such power has been viewed as a normal and inevitable consequence following from the structure of the organization, researchers instead focused on illegitimate power, i.e. power exercised outside formal hierarchical structures and the channels that they sanction. A consequence of this distinction has been that the use of power is being equated with dysfunctional and self-interested behavior. Two theories of power focus on intra-organizational dependencies. Strategic contingency theory of intra-organizational power (Hickson et al., 1971) builds on the idea that power is related to uncertainty or at least to its control. According to this model, strategically contingent sub-units are the most powerful, because they are the least

68

CHAPTER 6

dependent on other sub-units and can cope with the greatest systemic uncertainty, given that the sub-unit is central to the organization system and not easily substitutable. Similar to the strategic contingencies view of power is the resource dependency view (Hardy and Clegg, 1999). Information, uncertainty, expertise, credibility, stature and prestige, access to and contacts with higher echelon members and the control of money, rewards and sanctions have all been identified as resources of power (French and Raven, 1986; Pettigrew, 1973). All resource lists are infinite, however, since different phenomena become resources in different contexts. Without a theory of context one can never achieve closure on what the bases of power are. As opposed to the idea of legitimate power, these theories offer a better understanding for why and in what circumstances executives may actually be more powerful than their boards. CEOs may have a clear advantage with respect to information, uncertainty or prestige. On the other hand, board members depend directly on information provided by the CEO. Generally, Pfeffer, (1981: 32) refers to power as the capability of one social actor to overcome resistance in achieving a desired objective or result. Very similar, Finkelstein (1992: 506) defines power as the capacity of individual actors to exert their will. I use these definitions of power in the compensation setting process of my study. According to a political perspective, organizational outcomes are shaped by a struggle for dominance among coalitions with possibly opposing goals (Mintzberg, 1983; Pfeffer, 1981). Managers, owners and the board of directors represent major coalitions at the apex of a firm with opposing preferences as regards compensation policy, and the level and mix of compensation is likely to reflect the preferences of the group that gains the preponderance of power (OReilly et al., 1988; Tosi and Gomez-Mejia, 1989). In the case of a CEO, the assumed preferred result is higher pay and less compensation tied to shareholder wealth for a given pay level (Hill and Phan, 1991; Bebchuk and Fried, 2003). A powerful manager is able to influence or exert her will or desires on the remuneration decisions made by the board of directors or its compensation committee. Beyond pure economic utility of money a higher level of pay is associated with a symbolic value as a scorecard of managerial status and success (Ungson and Steers, 1984; Hambrick and Finkelstein, 1996: 287). And equity-based compensation subjects the CEO and top management to risk partly beyond their control, such as aggregate market demand and stock market fluctuations. Compensation components that depend on the firms stock price development also increase a CEOs firm-specific investment

CAN POWERFUL MANAGERS EXTRACT RENTS?

69

and, consequently, his or her associated risk (Beatty and Zajac, 1994). CEOs would presumably prefer to retain control over their pay by getting compensated through cash, thereby limiting the extent to which their income is exposed to risk. Conversely, the board of directors, as representatives of the companys shareholders should seek to pay their CEO as little as necessary to attract the desired talent and link his pay to the companys performance. Equity-based compensation and subsequent option and share holdings provide an explicit and direct link between executive compensation and firm performance (Jensen and Murphy, 1990a). Consistent with the view that more powerful executives earn higher pay, several scholars (e.g. Hambrick and DAveni, 1990; Finkelstein, 1992; Daily and Johnson, 1997) argue that compensation level is an important indicator of formal power.

6.1.2 Executive pay theories and executive power

Power has received attention from executive compensation researchers for a while (Allen, 1981a; Lambert et al., 1993). Agency theory implicitly acknowledges the existence of power in the relationship between executives and shareholders (Gomez, 2004). As Berle and Means (1932: 25) point out the separation of ownership from control produces a condition where the interests of owners and of ultimate managers may, and often do, diverge, and where many of the checks which formerly operated to limit the use of power disappear. The principal (i.e. stockholders represented by the board of directors) designs an optimal contract for the compensation of the agent (i.e. CEO or TMT) that maximizes the principals return accounting for optimal risk sharing between the principal and the agent (Grossmann and Hart, 1983). This may have a moderating effect on the level of compensation contracts and lead to tying compensation to firm performance. However, if the agent gains increasing influence over the body responsible for setting his compensation (i.e. the board of directors as the agents of shareholders), the design of the compensation contract may reflect executives preferences, that is, a higher level of total compensation and a lower portion of pay linked to the creation of shareholder wealth for a given level of pay (Grabke-Rundell and Gomez-Mejia, 2002; Zajac and Westphal, 1996a). Proponents of agency theory include power to explain the observed phenomena. Jensen and Murphy (1990a) explicitly refer to power, within and outside corporations, as a determinant of CEO pay. They argue, for example, that the public outcry over large bonuses and other financial rewards for CEOs has prevented efficient contracts that

70

CHAPTER 6

would ensure strong pay-performance relationships. Milgrom (1988) and Milgrom and Roberts (1992) suggest that executives make various efforts to influence their pay, which, in fact, might weaken the actual relationship between pay and corporate performance. Other agency theorists observe that excessive compensation does not seem to be a general phenomenon. As Holmstrm and Kaplan (2003: 27) note the main problem with executive pay levels is not the overall level, but the extreme skew in the awards. In other words, a few executives are able to extract substantial rents. Management researchers have also long been aware of the power of the CEO vis--vis the board of directors. As the firms central decision maker, the CEO represents the nerve centre where many communication lines intersect. He connects many networks with information and therefore enjoys an information advantage with respect to his associates with whom he shares strategic and operational responsibility (Barkema and Pennings, 1998). A complementary analytical position for studying board/CEO power relationships is frequently labeled managerial hegemony theory. Also inspired by Berle and Means (1932) work, this tradition of work argues that in spite of their legal responsibilities to safeguard shareholder interests, many boards are effectively controlled by the full-time, better informed, and more experienced corporate management and acquiesce into a rather passive, rubber-stamping role (Pettigrew and McNulty, 1998). This theory is consistent with the strategic contingency theory of power and resource dependence theory of power discussed above.

6.1.3 Prior studies on the link between executive power and pay
Although power has often been implied as a major determinant of executive compensation, it has seldom been carefully operationalized or tested in empirical research for its effects on the level and mix of executive pay. In fact, as Grabke-Rundell and Gomez-Mejia (2002: 13) state the negligence by agency theorists to explicitly17 consider power as a predictor of executive compensation levels may explain why strong empirical support for agency theory formulations is lacking. One reason for the lack of studies on the power-pay relationship may be that management power is hard to prove (Bratton, 2005: 18). Although CEOs can acquire power from a variety of sources (Daily

17

Emphasis added.

CAN POWERFUL MANAGERS EXTRACT RENTS?

71

and Johnson, 1997) prior research has usually focused on a single or a few archival variables such as CEO stock ownership (Allen, 1981a; Geletkanycz et al., 2001), CEO duality (Finkelstein and DAveni, 1994), CEO tenure (Hill and Phan, 1991; Ryan and Wiggins, 2004) or CEO duality and insider ratio (Sridharan, 1996) to proxy for the relative power of CEOs over the board of directors. Reliance on these individual proxies of executive power is further problematic, as alternative explanations exist for the influence of these variables on executive pay. For instance, CEO tenure has been used as a human capital explanation of executive pay (Harris and Helfat, 1997). Or CEO stock ownership has usually been considered to be a proxy for incentive alignment by financial economists (e.g. Jensen and Murphy, 1990a). Studies applying a more sophisticated measure of executive power include Combs and Skill (2003) who use founder status and CEO board tenure to develop a single measure of executive power, and Grinstein and Hribar (2004) who use three dichotomous variables (CEO duality, CEO on nominating committee and large boards) to develop a managerial power construct. Lambert et al. (1993) study the influence of power in 303 organizations between 1982 and 1984 on the compensation of top managers. Beyond organizational level they conceptualize power as equity ownership and selection of board members. Boyd (1994) examines the relationship between board control and CEO compensation. The five measures tested conceptually using LISREL VII methods sort into two factors with CEO duality, ratio of insiders and level of director compensation sorting into a structural power category, and board stock and institutional ownerships sorting into an ownership power category. Yermack (1997) studies the hypothesis that managers influence the terms of their own compensation by analyzing the timing of CEO stock option awards. He uses three individual variables to study CEO power over the compensation committee: the presence of a non-executive committee chairman, the presence of a non-executive director who is a major shareholder, and whether all committee members were appointed to the board by previous CEOs. The most explicit study on CEO power and compensation is Barkema and Pennings (1998). These authors study the impact of overt and covert power on CEO cash compensation in 143 medium-sized firms in the Netherlands. Their measure of overt power is shareholdings; covert power is proxied by CEO tenure, founder status and firm diversification. They argue that CEO compensation reflects the joint effects of his overt

72

CHAPTER 6

and covert sources of power: power from shareholdings is conditional on whether it is backed up by other sources of covert power such as tenure, diversity, and founder status. Bebchuk and Fried (2003, 2004) have re-launched the discussion of the impact of management power on executive compensation. They argue that the more power a manager possesses, the greater the rents benefits greater than those obtainable under true arms-length bargaining in the pay package. In their view, four factors contribute to the preponderance of managerial power over their pay. First, the board itself is weak because outside directors tend to be loyal to or dominated by the CEO due to process infirmities like large boards, CEO duality, interlocks and financial dependence. Second, most firms lack a large outside shareholder, the financial interest of whom would influence bargaining over pay. Third, oversight by large institutional shareholders tends to lead to more sensitive pay arrangements and some firms have fewer large institutional shareholders than others. Fourth, anti-takeover arrangements insulate most managers from the discipline otherwise imposed by the market for corporate control. However, while Bebchuk and Fried (2004) argue that managerial power is responsible for the large increase in executive pay in the 1990s, they do not test their propositions crosssectionally. There has been little consistency in results, although sound theoretical ground can be gathered, which may suggest that operationalization has been flawed and that a full test of the executive power hypothesis is warranted. The reliance on single indicators of power is surprising, as executive-level power has been shown to be multidimensional in character (Finkelstein, 1992). Furthermore, some potentially important sources of power such as interlocking directorates have received very little attention. Two reasons may have contributed to the insufficient operationalization of executive power in pay investigations. First, much of the data is not readily available from standard databases and therefore cumbersome to collect. Second, researchers have typically assumed that most sources of power increase with CEO tenure (Hambrick et al., 2005). As a result, CEO tenure was used as the main proxy for CEO power thereby neglecting the nuances provided by other sources of power. An integrative approach is further warranted, as individual variables cannot capture the effect of conflicting hypotheses.

6.1.4 Related studies on CEO power

As outlined above, researchers on power theory agree that there is no single definition or conceptual approach to power that is universally appropriate and that the sources and

CAN POWERFUL MANAGERS EXTRACT RENTS?

73

uses of power are fundamentally shaped by the general context and particular features of the setting in which those power sources accumulate (Pettigrew and McNulty, 1998). The conceptualization of power therefore needs to be tailored to the specific situation of executive/board power to meet the specific needs of this investigation. Adams et al. (2004) study whether the CEOs power to influence decision-making is associated with future firm performance. Their influence power index is the sum of five indicator variables, depending on whether the CEO is also the chairman, whether the CEO also holds the title President, whether the firm has no President or Chief Operating Officer, if the CEO is the founder of the company, and insiders on the board. Zajac and Westphal (1996a) study the influence of relative CEO/board power on desired characteristics of a new CEO. They measure the relative CEO/board power in four ways: CEO duality, board tenure relative to CEO, interdependent outside directors and outsider stock ownership. Powerful CEOs are shown to influence boards to appoint demographically similar successors, thus inhibiting change in CEO characteristics (Zajac and Westphal, 1996a). Finance scholars have investigated the link between managerial entrenchment and pay. Managerial entrenchment is close to the concept of power. Berger et al. (1997: 1411) define managerial entrenchment as the extent to which managers fail to experience discipline from the full range of governance and control mechanisms including monitoring by the board, the threat of dismissal or takeover, and stock- or compensationbased performance incentives. Characteristics of entrenchment include long tenure in office, compensation that has low sensitivity to performance (Berger et al., 1997) and stock ownership (Morck et al., 1988). What is different in this viewpoint is that executive pay itself is used as a governance mechanism determining the level of entrenchment. In other words, while executive power is a construct hypothesized to influence compensation level and structure, managerial entrenchment is a construct that includes compensation structure to explain other firm-level outcomes. Finkelstein (1992) develops a number of objective measures of power that were thought to cumulatively explain overall power within a firms top management team. These measures were compensation and formal titles, corporate board memberships, memberships on non-profit boards, the rating of these boards, elite education and expertise. Grabke-Rundell and Gomez-Mejia (2002) elaborated on Finkelsteins model

74

CHAPTER 6

of power to categorize the variables into dimensions of power and state propositions on the relationship between power and the level of executive compensation. These elements of power are structural power, ownership power, expert power, and prestige power. Several subsequent studies employed variations of Finkelsteins (1992) TMT power scale to examine elements of managerial power. Haleblian and Finkelstein (1993) utilize a derivative version of the scale for a variable labeled CEO dominance with ten of the thirteen Finkelstein (1992) scale components comprising this variable. CEO dominance is found to be a significant predictor of firm performance. Finkelstein and DAveni (1994) include a construct labeled informal CEO power in their research model. Informal CEO power is measured by utilizing seven of the thirteen measures validated in Finkelstein (1992) with the ownership factor measure excluded. The study finds that higher informal CEO power tends to increase the vigilance of the board of directors, therefore decreasing the occurrence of CEO duality. Daily and Johnson (1997) include eight of the thirteen composite measures tested in Finkelstein (1992) and use at least one measure from each of the four construct factors. Their results indicate that firm performance is both an antecedent condition and outcome of CEO power. However, no study has embraced Finkelsteins (1992) model of power to study executive compensation. As Grabke-Rundell and Gomez-Mejia (2002: 19) stated a major contribution to the understanding of executive compensation may be obtained by incorporating power measures such as those developed by Finkelstein (1992).18

6.1.5 Conceptual framework of executive power and compensation

The managerial power hypothesis posits that senior executives have substantial influence over their pay. As a result, at least some executives receive pay in excess of the level that would be optimal for shareholders, and this excess pay constitutes rent. Agency theorys basic risk aversion assumption is that agents do not like variability (risk) in their compensation (Eisenhardt, 1989). Powerful managers may therefore decouple their pay from firm performance and shift the mix of compensation towards more stable elements such as salary (Dyl, 1988).

I use the model as part of the process of conceptualizing CEO power, as it is based on the relative power of the members of a top management team - which is not the same as the relative power of the CEO versus the board.

18

CAN POWERFUL MANAGERS EXTRACT RENTS?

75

This study develops a broad concept of executive power along various dimensions and tests its impact on executive compensation level and mix applying principal component analysis.
Figure 4: Conceptual framework of executive power and compensation.

Executive ownership Non-executive ownership Shareholder concentration Non-executive directors

Ownership power

Structural power

Independent directors CEO duality Compensation committee Board size

CEO tenure Interdependent directors

Tenure power

Cash compensation Total compensation % equity-based compensation

Executive compensation

Interlocking directors

Network power

Outside board memberships

Credibility power
Prior performance CEO celebrity status Education

Figure 4 above depicts the conceptual model of executive power and compensation. The model is based on the earlier conceptualization of executive power developed by Finkelstein (1992) and a review of the relevant literature on executive pay and power. First, executive power can arise from an ownership dimension. Depending on how much equity ownership of the firm executives and non-executives have, their relative power will be different. Also, the concentration of outside shareholdings will affect the relative power of incumbent management. Second, executive power depends on the boards structural set-up. Formal structures such as the combination of the board chairman and CEO positions (CEO duality), the proportion of non-executive and independent directors, the existence of a specialized compensation committee and the size of the board all influence the relative power of executives. Third, it has been advocated that CEOs are able to develop more power over the course of their tenure. Fourth, networks may play a crucial role to develop executive power. And fifth, executive power may stem from credibility achieved through good performance, higher education or an aura of

76

CHAPTER 6

celebrity that the CEO is able to create. All five dimensions of power are hypothesized to affect the level and structure of executive compensation in the manner outlined above. In the following, the five dimensions of power will be investigated in more detail.

6.2 Ownership power and executive compensation

6.2.1 Executive ownership
Economic and finance research originally used the distribution of stock ownership among the members of the board of directors as the specification of power or control (Allen, 1981a, b; Salancik and Pfeffer, 1980; Jensen and Warner, 1988; Toyne et al., 2000). In this view, a CEO is more powerful whenever he is a principal stockholder and less powerful whenever there are principal stockholders among the other directors (Allen, 1981a: 1115). As legal owners of the firm, shareholders are given the rights to vote and influence strategic decisions, power that increases along with the percentage of shares owned. An executive who is also a shareholder has the same privileges. Moreover, the executive shareholder is in the position to influence board decisions and his or her performance criteria. Research has shown that CEOs with greater share ownership are more likely to be retained (Denis et al., 1997; Milbourn, 2003). Not only does stock ownership grant privileges to the executive, but it also reduces uncertainty for the firms owners by guaranteeing that the executives interests are aligned with theirs.19 Prior empirical research mainly supports the notion that powerful executives in terms of their equity ownership are able to extract a pay premium. CEO equity ownership is positively related to salary (Khan et al., 2005), cash compensation (Allen, 1981a20; Lambert et al., 1993; Dyl, 1988, Sanders, 2001b), fixed compensation (Cyert et al., 2002) and total compensation (Core et al., 1999; Bliss and Rosen, 2001). However, other
Managerial ownership has also been advocated as a means of reducing agency costs by aligning the interests of managers and with those of shareholders (Jensen and Meckling, 1976). This suggests an opposite effect of managerial ownership on agency-related problems. However, with respect to the size of the executives pay package, it may be argued that the incentive alignment hypothesis has less merit. An example illustrates this view. Assuming the CEO were to give up 500,000 CHF in cash compensation. This would result in a 1-to-1 increase in before-tax profits of the firm, thus increasing shareholder wealth by 500,000 CHF. However, the CEO is unlikely to be indifferent, as he only will regain these 500,000 CHF if he is the single owner of the firm. As long as his ownership is less than 100%, the CEO will prefer a higher salary, and share ownership will not influence this decision. Similarly, if executives prefer cash salary to equity-based compensation, other things being equal, their preference will not be reversed at higher levels of managerial ownership. On the contrary, higher levels of managerial ownership may reinforce the preference for fixed cash compensation.
20 19

This result does not hold when the CEO is the principal shareholder.

CAN POWERFUL MANAGERS EXTRACT RENTS?

77

studies (e.g. Sanders and Carpenter, 1998; Coombs and Gilley, 2005) report no significant association between CEO stock ownership and total compensation level. Also, more powerful CEOs measured as their equity ownership stake may be able to shield their compensation from too much dependence on a firms share price. CEO stock ownership has a significant negative association with the level of stock option compensation (Khan et al., 2005) and restricted stock awards (Ryan and Wiggins, 2001) and the proportion of equity-based (Zajac and Westphal, 1994; Brick et al., 2002), stock option (Bryan et al., 2000a; Khan et al., 2005) and long-term compensation (Sanders and Carpenter, 1998). Top management team stockholdings is also negatively associated with the use of equity-based pay (Barron and Waddell, 2003b). In Yermack (1995), stock option pay mix is not related to CEO stock ownership, and in Lewellen et al. (1987) there is no relationship between executive shareholdings and equity-based compensation. More recently, Toyne et al. (2000) point out that the relationship between executive ownership and equity-based compensation may be non-monotonic. Executives may need a certain threshold of ownership in order to have enough voting power to influence the compensation setting process. They find that across a range of ownership between 13% and 22%, increases in management ownership results in decreases in the proportion of equity-based compensation. However, up to 13% and above 22%, additional management stock ownership increases the equity pay mix. Their results are also consistent with the view that incentive alignment of executive ownership may dominate the power effect at very high ownership levels. These results are mostly consistent with the ownership power hypothesis and the assumption of a preference for non-variable pay by the CEO (Grabke-Rundell and Gomez-Mejia, 2002), although these studies usually do not explicitly refer to managerial power as an explanation. Instead, these studies apply an optimal contracting view, which emphasizes the alignment between managers and shareholders through an optimal level stock ownership, balancing incentive alignment and risk aversion. For instance, Bryan et al. (2000a: 668) note when CEOs hold a large fraction of their firms equity, the demand for further stock-based compensation is likely to be reduced, since the interests of CEOs and shareholders are relatively aligned already.

78

CHAPTER 6

Three related studies are consistent with a power explanation of share ownership. Boeker (1992) studies managerial power and its relation to CEO dismissal through scapegoating and find that CEO ownership is negatively associated with CEO dismissal. Shen and Cannella (2002) show evidence that CEOs with a higher ownership stake are less likely to be involved in power contests resulting in CEO dismissals suggesting that stock ownership increases a CEOs influence. Bertrand and Mullainathan (2001) show that CEOs with higher ownership tend to be rewarded more heavily for luck consistent with a managerial power view.

6.2.2 Non-executive directors ownership

The relative power of non-executive directors can also be viewed in terms of their equity holdings. Essentially, the power of any director depends on the extent to which that individual controls a significant block of stock in the corporation (Allen, 1981a). It seems that investigating CEO power over the board of directors arising from CEO equity ownership without considering non-executive directors ownership is flawed. Finkelstein and Hambrick (1996: 226) propose that non-executive board directors with a higher equity ownership stake would be more effective monitors of top management. Following the argument above, powerful non-executive directors will set executive pay according to shareholder preferences lower total compensation and more equity-based, other things remaining equal (Hill and Phan, 1991). However, empirical evidence is scarce and does not support these propositions. Finkelstein and Hambrick (1989) and Lambert et al. (1993) document no significant relationship between outside directors holdings and executive compensation. Core et al. (1999) and Daily et al. (1998) find no relationship between outside director ownership and total CEO compensation. Cyert et al. (2002), however, document a negative relationship between the equity holdings of the compensation committee and total CEO compensation. And Cheng and Firth (2005) find that director ownership moderates top management pay in Hong Kong. Other studies that use CEO equity ownership as a proxy for CEO power do not control for outside director holdings (e.g. Gray and Cannella, 1997; Finkelstein and Boyd, 1998; Gomez-Mejia et al., 2003; Ryan and Wiggins, 2001). To my knowledge, no study has investigated the relationship between non-executive directors ownership and executive pay mix. Beyond a direct effect on executive pay levels, non-executive stock ownership further appears to affect special compensation arrangements. Kosnik (1990) finds that when

CAN POWERFUL MANAGERS EXTRACT RENTS?

79

non-executive directors have a small ownership interest relative to management then the company is more likely to pay greenmail.21 Yermack (1997), in studying the timing of CEO stock option awards, finds that the presence of a non-executive director with at least 5 percent of the firms shares decreases the likelihood of favorably timed stock option awards. In short, there are potentially important effects of non-executive shareholdings on executive pay, but these potential effects have been largely neglected by prior research.

6.2.3 Shareholder concentration

Another source of the relative ownership power of top management is the concentration of outside equity ownership. The agency problem associated with the separation of ownership and control (Fama and Jensen, 1983) can be overcome by effective monitoring of large shareholders (Shleifer and Vishny, 1986; Prowse, 1992). Large shareholders have the power to exert control over management either through their voting power or through representation on the board, or both. They should also have enough incentives to do so, as selling stock when dissatisfied with management has become prohibitively expensive. With regard to executive compensation, pay levels may be more moderate. Also, strong shareholders may be able to set compensation according to their preferences, that is, tying pay to shareholder wealth creation (Blazekovic, 2004). Early empirical studies have used a 5% threshold in operationalizing ownership concentration when investigating the actions and compensation of agents (McEachern, 1975; Dyl, 1988; Gomez-Mejia et al., 1987; OReilly et al., 1988; Tosi and GomezMejia, 1989; Porac et al., 1999). This standard operationalization of ownership structure gives rise to two types of firms: owner-controlled and management-controlled firms. Owner-controlled firms are defined as those that have at least one shareholder, other than a manager of the firm, who owns 5% or more of the companys stock. In managementcontrolled firms, there are no such major shareholders. In general, CEOs in managementcontrolled firms get more generous cash (Allen, 1981a) and total compensation (Core et al., 1999), more stock options (Ryan and Wiggins, 2001) and have higher pay raises (Hambrick and Finkelstein, 1995) than CEOs of owner-controlled firms. Further, these

Greenmail refers to the agreement between a large shareholder and a company in which the shareholder agrees to sell his stock back to the company, usually at a premium, in exchange for the promise not to seek control of the company for a specified period of time.

21

80

CHAPTER 6

managers have a lower stock option compensation mix, are rewarded relatively abundantly for profit increases, but are not penalized for profit declines (Ryan and Wiggins, 2001). In a similar vein, lower-level managers earn lower base pay and have their pay more closely tied to performance in owner-controlled firms than in management-controlled firms (Werner and Tosi, 1995). On the other hand, Mehran (1995) finds that firms in which a higher percentage of shares are held by outside blockholders use less equity-based compensation. This result is inconsistent with a managerial power view, but consistent with the view that different corporate governance arrangements substitute one another (Rediker and Seth, 1995). Studies using other operationalizations of shareholder concentration find results consistent with the ones reported above. Using a Herfindahl-Index, Schmid (1997) also documents a negative relation between total compensation and ownership concentration. Beiner et al. (2005) measure the cumulated voting rights exercised by large outside investors with voting rights exceeding 5% in Switzerland and find that this measure is negatively related to the percentage of executive share compensation. Finally, some studies focus on special types of ownership. Kato (1997) studies CEO compensation of 154 large Japanese firms with and without financial keiretsu affiliation.22 He finds that CEOs of financial keiretsu earn 21% less than those of independent firms. This study shows the importance of banks as lenders and shareholders as monitoring agents. Examining CEO pay in almost 2,000 US firms during the 1991-1997 period, Hartzell and Starks (2003) find that the more concentrated institutional ownership, the lower is executive compensation and the higher is the payperformance sensitivity. Khan et al. (2005) find that large institutional investors are associated with lower levels of salary, options and total CEO compensation. They also find that a larger institutional presence results in more performance-sensitive compensation. David et al. (1998) find that both large non-institutional shareholders and pressure-resistant institutional shareholders reduce the level of CEO compensation and increase the proportion of long-term incentives to total compensation. Higher levels of compensation and lower proportions of equity-based pay in management-controlled firms are consistent with the pay preferences of CEOs and the finding that the CEO is
A financial keiretsu is a group of firms linked by their relationships to a main bank and by crossholdings of equity. The main bank is the principal lender to member firms. It also owns a significant number of shares in member firms and often sends a representative to the board of directors.
22

CAN POWERFUL MANAGERS EXTRACT RENTS?

81

most influential in such firms in setting CEO compensation (Tosi and Gomez-Mejia, 1989). Gomez-Mejia et al. (2003) study a sample of family-controlled US firms. Their results suggest that the determinants of CEO compensation of family-controlled firms may be different from CEO compensation of firms without a large family ownership. For instance, the authors find that executives with family ties to owners receive lower total pay than professional managers. Kole (1997) finds that participation in corporate management by a founding family significantly reduces the likelihood of establishing most types of equity-awarding plans.

6.3 Structural power and compensation

I build the dimension of structural power23 on the variables non-executive directors, independent directors, CEO duality, presence of a compensation committee and board size.

6.3.1 Non-executive and independent directors

Past economic research has investigated board structure as indicative of the relative power of the CEO over the board or to what extent boards use their monitoring role (e.g. Boeker, 1992). This stream of research mainly follows the typology of the proactive board (Pearce and Zahra, 1991). For proactive boards to exist they must usually be composed primarily of independent outside (non-executive) directors. Non-executive and especially independent directors are viewed as better monitors of a CEOs performance. The higher the number of independent outside directors on the board, the higher the overall power of non-executives as opposed to executive directors. Fulfilling their duty as shareholders representatives, boards with a high proportion of independent outside directors will restrict pay excesses to a minimum and tie compensation to shareholder wealth creation. This line of argument forms the basis for the recent listing requirement at the New York Stock Exchange and the recommendations of the Swiss Code of Best Practice that the board be composed of a majority of independent directors with no material relationships with the company (Holmstrom and Kaplan, 2003; Bebchuk and Fried, 2004: 202).

Finkelstein (1992: 512) refers to structural power as the distribution of formal positions within an organization.

23

82

CHAPTER 6

A vast body of literature has tested the effect of outside directors on executive compensation, although little supporting evidence could be found. Contrary to the theoretical arguments of outside directors as being more powerful to defend shareholders preferences, researchers found non-executive directors to be positively associated with cash pay (Main, 1991; Boyd, 1994), the value of restricted stock grants (Ryan and Wiggins, 2001) and total executive compensation (David et al., 1998; Core et al., 1999). Conyon and Peck (1998) report no significant effects of non-executive directors on cash compensation and Benz et al. (2001) report a positive relationship between the number of executive managers on board and total option grants. Further, Grinstein and Hribar (2004) find no effect of board composition on CEO M&A bonus. However, Bertrand and Mullainathan (2001) find that insider presence on the board increases the pay for luck for CEOs consistent with a power perspective.24 Less research has been conducted with regard to compensation mix, but existing evidence is more supportive. Non-executive directors are found to be positively associated with the proportion of long-term incentives (David et al., 1998) and equitybased compensation (Mehran, 1995). However, non-executive directors are negatively associated with the likelihood of stock option repricing25, a behavior that may be considered as beneficial to managers at the potential detriment of shareholders (Chidambaran and Prabhala, 2003; Chance et al., 2000; Economiesuisse, 2002; Bebchuk and Fried, 2004: 165).26 Recently, Beiner et al. (2005) measured the compensation for all top management team and board of director members for a sample of 156 Swiss firms in 2002 and found a significant positive relationship between the ratio of outside directors and the proportion of equity-based compensation.

Pay for luck refers to pay that is not directly attributable to the performance of the firm, but rather to overall industry and market developments. Formally, repricing of stock options is executed either by canceling the old options and replacing them with new grants at more favorable terms, or by simply rewriting the terms of the existing option contracts. Proponents of repricing offer two major explanations for repricing. One argument is that deep-outof the-money options no longer provide any meaningful incentives to executives (Hall and Murphy, 2000), so it is necessary to revise the strike price downward. A second argument is that repricing is needed to retain key executives. However, the resistance of stockholders to repricing suggests that that option repricing nonetheless represents a breakdown in the agency relationship between managers and stockholders (Pollock et al., 2002).
26 25

24

CAN POWERFUL MANAGERS EXTRACT RENTS?

83

At best, there is mixed evidence on whether more power in the hands of non-executives as opposed to executives in terms of board structure leads to pay practices consistent with shareholder preferences, as it has been suggested by the Swiss Code of Best Practice and other governance reports. This indicates that the specification of relying on this variable only may not be very helpful. One problem with many of these studies is that these researchers use the incentive alignment hypothesis for absolute values of option grants. However, as options now constitute a large part of total compensation in the US (Murphy, 1999), researchers actually apply a crude measure of total compensation. To illustrate this case, consider the following example. In firm A, the CEO receives 1 MUSD in cash compensation and 10 MUSD of stock options. In firm B, the CEO receives 4 MUSD in cash and 10 MUSD worth of stock options. Arguably, while the CEOs of both firms score similarly on incentive alignment in such studies, the CEO of firm B gets a lower proportion of his total pay in options and may therefore be less aligned with shareholders interests. Moreover, this CEO seems to be able to extract a rent from shareholders, all else being equal.

6.3.2 CEO duality

CEO duality is a very important structural measure of relative power of the CEO over the board of directors and other top executives (Brickley et al., 1997; Finkelstein and DAveni, 1994; Boyd, 1994). If the CEO is also the chairman of the board of directors, the ability of the board to exert control over the compensation setting process may be compromised, increasing the potential for agency conflicts. The CEO may exert influence by setting the agenda, controlling information flows and creating dependence for non-executive directors. A CEO always has an information advantage over nonexecutive board members. Formally, the CEO bargains with his subordinates for his compensation. Several studies find that CEO duality is a significant predictor of executive compensation levels (e.g. Main et al., 1994; Westphal and Zajac, 1994; Core et al., 1999; Main 1991; Brickley et al., 1997) and M&A cash bonus (Grinstein and Hribar, 2004). This can be viewed as a reward for the extra workload involved. It does not, however, rule out the possibility that executives in such dominant positions are able to garner more generous pay packages, which remains a question of proportion. While a modest pay premium may suggest a reward for the extra workload, more pronounced differentials

84

CHAPTER 6

may support a self-serving management perspective. There is indication that the pay premium in M&A bonuses associated with CEO duality is not entirely attributable to increased responsibility (Grinstein and Hribar, 2004).27 And Core et al. (1999) find that CEO excess compensation resulting from weak governance structure such as CEO duality is negatively associated with subsequent accounting performance. This provides support for a managerial self-serving view. Little and mixed evidence exists with regard to compensation mix. Westphal and Zajac (1994) document that CEO duality is negatively associated with long-term incentive grants, but Beatty and Zajac (1994) report that CEO duality is associated with higher performance-based compensation in a sample of initial public offerings. This issue awaits further empirical testing. Finally, CEOs who also hold the position of board chairman appear to influence the terms of compensation. Pollock et al. (2002) document that CEO duality increases the likelihood of option repricing. And CEOs occupying both positions are less likely to be fired by the board for poor performance (Goyal and Park, 2002).

6.3.3 Presence of a compensation committee

In powerful proactive boards directors responsibilities are also divided among established committees (Pearce and Zahra, 1991). There is widespread belief that the existence of committees may enhance the boards power position, as long as the CEO is not a member of these committees (e.g. Hambrick and Fukutomi, 1991; Ferrarini et al., 2003; Andjelkovic et al., 2002). The presence of a compensation committee institutionalizes the boards monitoring task. Detailed work is undertaken in such committees and suggestions are prepared and presented to the full board. The compensation committee is responsible for setting executive compensation and therefore seen as the powerful monitoring mechanism with regard to setting optimal compensation packages (Daily et al., 1998). Compensation committees are viewed as forums within which directors determine the appropriate design of reward structures for management and align management and shareholders interests (Conyon and Peck, 1998: 148). In their absence, there exists an opportunity for top executives to award themselves excessive pay packages. For instance, ABB established a compensation committee after

Bebchuk and Fried (2004: 129) argue that the payment of acquisition bonus itself is a manifestation of executive power.

27

CAN POWERFUL MANAGERS EXTRACT RENTS?

85

excessive pension payments to former CEOs became public with the purpose to prevent such excesses in the future (Catrina, 2003: 203). Furthermore, with the existence of a compensation committee, responsibility lies with a smaller group of people. As high compensation packages are likely to draw attention to a firm (Pollock et al., 2002), directors may be unwilling to provide generous pay packages, as the media may pinpoint such bad behavior and a directors reputation and further presence as a director may be at stake (Schildknecht, 2004: 138). According to agency theory, directors may have a fundamental interest that the public views them as critical and confident board members (Fama and Jensen, 1993). Despite strong theoretical focus on committees, little evidence can be found for their effect on executive compensation. Given that literally all stock-listed companies in the US have had such a committee for a long time28, no empirical US research exists to test the hypothesis that established committees limit executive power and thus align managerial and shareholder interests with regard to compensation. This governance structure is relatively new to Switzerland (Stern and Peck, 2003). The recently published Swiss Code of Best Practice in Corporate Governance (2002) prescribes that companies establish a compensation committee. In the UK, almost all stock-listed companies had adopted remuneration committees by the mid-1990s (Conyon, 1997). Four studies have investigated the relationship between compensation committee existence and executive compensation in the UK based on early data sets. Main and Johnston (1993) find a positive association between the existence of a remuneration committee and pay levels. Conyon (1997) examines 213 large UK companies between 1988 and 1993 and documents that companies, which adopt remuneration committees, have lower growth rates in top director cash compensation. In a study of 94 UK firms between 1991 and 1994, Conyon and Peck (1998) find that the existence of compensation committees was positively associated with top management cash pay. Finally, Johnston (2002) studies the salary of 220 the highest-paid directors in the UK in 1995 and finds no significant impact of the presence of a remuneration committee on remuneration. With the exception of Conyon (1997), these results are inconsistent with a power explanation of compensation committee

The New York Stock Exchange requires all listed companies to have a compensation committee (NYSE Manual 303A.04-.06.).

28

86

CHAPTER 6

existence, as advocated by Pearce and Zahra (1991). To my knowledge, no empirical evidence exists on the relationship between compensation committee existence and the mix of compensation.

6.3.4 Composition of the compensation committee

The composition of the compensation committee may also be an indication of executive power (Conyon and Peck, 1998). If the CEO sits on this committee or if directors on the compensation committee are not independent, the CEO may be in a powerful position to influence his pay according to his preferences. It may be enough for a CEO to serve on the compensation committee for one term only. Once established, compensation levels are unlikely to decrease again and a one-time presence on the compensation committee may therefore be sufficient to fix higher compensation levels for several years (Bertrand and Mullainathan, 2001). Reports and Codes on Corporate Governance in the UK (e.g. the Greenbury Committee (1995)) or the Swiss Code of Best Practice in Corporate Governance (2002) recommend that compensation committees be comprised entirely of non-executive (independent) directors. In the US, recent regulations essentially preclude executive directors from serving on a firms compensation committee (Bebchuk and Fried, 2004). Prior empirical evidence on the effect of compensation committee composition on CEO remuneration has been mixed at best. Daily et al. (1998) find no evidence that the composition of the compensation committee has an effect on CEO compensation, although they apply a wide variety of measures such as affiliated directors, interdependent directors and outside CEOs. Similarly, Newman and Mozes (1999) find no support for the hypothesis that total CEO compensation is greater in firms with an insider on its compensation committee than in firms without an insider on the committee. However, they report evidence that while firms with an insider on the compensation committee compensate their CEOs similarly for favorable performance, they shield their CEOs more effectively from the impact of unfavorable performance on their compensation. Conyon and Peck (1998) document a negative association between the proportion of executives on the compensation committee and cash compensation. Main et al. (1995) show that a CEOs pay is higher when the chair of the compensation committee has been appointed during the term of the CEO. Anderson and Bizjak (2003) examine whether greater compensation committee independence promotes shareholder interests and whether the CEOs presence on the compensation committee leads to

CAN POWERFUL MANAGERS EXTRACT RENTS?

87

opportunistic pay structure. Their results provide no evidence that committees consisting of insiders or the CEO do award excessive pay or lower overall incentives. Consistent evidence exists on the relationship between compensation committee composition and pay mix. Anderson and Bizjak (2003) document a negative relationship between CEO on the compensation committee and stock options pay mix. Newman and Mozes (1999) find that when at least one member of the compensation committee is an insider and therefore subordinate to the CEO, the sensitivity of pay to performance is lower. While the relationship between committee composition and pay level remains unclear, there is evidence that executives sitting on compensation committees influence the terms of option compensation. The finance literature tried to investigate whether committee composition is related to agency problems in stock option compensation. Brenner et al. (2000) and Callaghan et al. (2004) document that the likelihood of stock option repricing increases when executive officers are present on the firms compensation committee. Yermack (1997) documents that, when the CEO is a member of the compensation committee, the firm is more likely to time option grants prior to the release of good news. The latter practice leads to an underestimation of option grants made, as the good news immediately after the award pushes options in the money. Overall, the results suggest that while any direct effect may be weak, compensation committee composition influences pay in a more subtle way.

6.3.5 Board size

Finally, board size may be an important structural determinant in the distribution of power between the CEO and the board of directors (Yermack, 1996). Jensen (1993) proposes that as boards of directors become larger, they are less capable of holding frank discussions and are less effective as monitors of top management. From an accountability perspective, members of larger boards may feel more anonymous and less responsible, as the individual responsibility of each director (Andjelkovic et al., 2002) and resulting risk-aversion (Bebchuck and Fried, 2004) are lower in larger boards. Awarding larger pay may go unnoticed more easily. From a free-riding perspective, the larger the board the less time there is for individual contributions in a board meeting. This makes free-riding behavior by individual directors more likely in larger boards (Golden and Zajac, 2001). It follows that larger boards are less likely to function

88

CHAPTER 6

effectively and are easier for the CEO to control (Jensen, 1993). These arguments imply that the CEOs power increases with the number of directors on the board. Board size has often been omitted in studies of executive compensation. Those studies that included board size are consistent with the managerial power hypothesis. There is a positive association between board size and cash compensation of the highest paid director in the UK (Main et al., 1991; Conyon and Peck, 1998) and total CEO compensation (Core et al., 1999; Gosh and Sirmans, 2005). Yermack (1996) reports that smaller boards are more likely to dismiss the CEO in cases of weak performance and smaller boards provide stronger incentive compensation to the CEO. However, recent research suggests that the relationship between board size and executive pay is more sophisticated. Grinstein and Hribar (2004) suggest that when the CEO is more involved in choosing board members (i.e. he sits on the Nomination Committee) a smaller board might actually mean that the CEO has more managerial power.

6.4 Tenure power and executive compensation

6.4.1 CEO tenure
Allen (1981b) finds a positive link between CEO power and CEO tenure, measuring power as the level of stock ownership by the CEO. He concludes that CEOs manage to accumulate power over the length of their tenure. Ever since, the notion of power has provided the main theoretical basis for studies on the influence of CEO tenure on compensation (e.g. Hill and Phan, 1991) and board composition (Hermalin and Weisbach, 1998; Arthur, 2001). Over the time of his office, the CEO is likely to become more powerful stemming from a variety of means. For example, board independence declines over the course of a CEOs tenure as the CEO acquires influence through his control over the nomination of new board members (Hermalin and Weisbach, 1998) and the removal of troublesome ones (Finkelstein and Hambrick, 1989). In addition, CEOs may gain power due to an increasing level of firm-specific human capital (Arthur, 2001) or the firms investment decisions being biased towards the skills and knowledge of the manager (Shleifer and Vishny, 1989). Also, lengthening tenure may allow the CEOs power to become institutionalized through the creation of a personal mystique or patriarchal aura, including unquestioned respect or loyalty (Hambrick and Fukutomi, 1991). Tenure signifies cumulative success, insofar as CEOs who perform well are retained (Hambrick and Fukutomi, 1991; Shen, 2003). Finally, CEOs are more exposed

CAN POWERFUL MANAGERS EXTRACT RENTS?

89

to power struggles with other top managers in the early years of their tenures (Shen and Cannella, 2002). This cumulative success, or substantial track record, then, is also a basis for power. It will be more difficult or costly for the firm to fire such entrenched managers. Adams (2004) documents a positive relationship between CEO tenure and Finkelsteins (1992) four-dimensional power measure. Prior empirical research has found mixed results on the influence of CEO tenure on the level of CEO compensation. While numerous studies find a positive linear or log-linear relationship between tenure and various measures of compensation (e.g. Mayers and Smith, 1992; Sanders and Carpenter, 1998; Hallock, 1997; Brickley et al., 1997; Kato, 1997; Brick et al., 2002; Sanders, 2001b; Johnston, 2002), other studies document no significant relationships (Finkelstein and Boyd, 1998; Henderson and Fredrickson, 1996; Miller et al., 2002) or even a significant negative relationship (David et al., 1998). Hambrick and Finkelstein (1995) find no relationship between CEO tenure and the change in cash-plus-option compensation. Singh and Harrianto (1989) document a positive relationship between TMT tenure and the size of golden parachutes.29 The mixed results may be partially due to misspecification. Few studies attempted to test for a non-monotonic link between tenure and compensation. Finkelstein and Hambrick (1989) observe that extremely long CEO tenure (over 18 years) had a negative effect on compensation. This final result is consistent with the lame duck hypothesis (Hambrick and Fukutomi, 1992), which states that the CEO who is approaching mandatory retirement may experience a diminution of power. Furthermore, if CEOs get golden hellos a large initial payment on top of the annual compensation package (Bebchuk and Fried, 2004: 130) a linear relationship may be hard to detect. Despite inconsistent empirical evidence CEO tenure continues to be used as a single control variable for CEO power in more recent pay studies (e.g. Sanders, 2001b). Empirical results on the relationship between tenure and compensation mix are more coherent, and there is evidence that equity-based pay may decrease over a CEOs tenure (Westphal and Zajac, 1994; Bushman et al., 2004; Ryan and Wiggins, 2004; Brick et al., 2002). Newly hired executives may receive especially large option awards (Yermack, 1995) and equity grants (Ittner et al., 2003). Hubbard and Palia (1995) differentiate a

Golden parachutes involve a negotiation by top management with the board for the inclusion of sizeable payments made in the event of takeover of the firm (Singh and Harianto, 1989: 143).

29

90

CHAPTER 6

sample of banks into a sub-sample with CEO changes and one without CEO changes. They observe that the increase in the sensitivity in the value of options granted is higher in the sample that includes CEO changes than in the sample without CEO change. Anderson and Bizjak (2003) document a negative, but statistically insignificant association between new option grants and CEO tenure and a weak negative association between CEO tenure and stock option pay mix. Under the assumption that CEOs prefer cash pay to equity-based pay, this suggests that CEOs are able to shift their compensation packages away from long-term compensation and toward cash-based pay.30 Finally, Westphal (1998) shows that CEO tenure has a negative influence on the subsequent change in compensation contingency. Bushman et al. (1996) find that for longer-tenured CEOs a higher percentage of the annual bonus is based on individual performance. This result is consistent with the notion that a powerful CEO may use his influence to procure individual performance evaluation which may be more prone to manipulation. Finally, it is interesting to note that longer-tenured CEOs are associated with smaller compensation packages for outside directors (Ryan and Wiggins, 2004).

6.4.2 Interdependent directors

A more refined way of measuring the influence of tenure power on pay is through the nomination of new board members (Hill and Phan, 1991). In practice, CEOs have great influence in the selection of directors (Hermalin and Weisbach, 1998; Shivdasani and Yermack, 1999; Hengartner and Rick, 2003; Charkham, 1995) and this influence increases over time (Hermalin and Weisbach, 1998). Directors appointed during the tenure of an incumbent CEO have been termed interdependent directors (Daily et al., 1998). The timing of directors appointments to a board may provide some insight into a CEOs ability to exert influence over compensation committee members and the board of directors in general in order to garner a more attractive compensation package. Also, board appointments confer prestige and status, as well as financial rewards and

The shift away from long-term pay to cash pay can also be a result of long-tenured CEO having accumulated already a significant ownership leading to sufficient incentives from existing stockholdings and reducing incentive pay (Jensen and Meckling, 1976). Indeed, Milbourns (2003) results indicate that overall stock-based pay sensitivity (including changes in wealth arising from stock option holdings) is positively related to CEO tenure.

30

CAN POWERFUL MANAGERS EXTRACT RENTS?

91

perquisites. Thus, given norms of reciprocity, non-executive directors may feel socially obligated to support CEOs who favored their appointment (Wade et al., 1990). Empirical evidence has been predominately supportive. Several studies (Lambert et al., 1993; Main et al., 1994; Westphal and Zajac, 1994; Core et al., 1999) find a positive relationship between the percentage of outside directors appointed during the current CEOs period in office and executive compensation. Main et al. (1995) find that a CEOs pay is higher when the chair of the compensation committee has been appointed during the term of the CEO. Yermack (1997) documents that compensation committee members appointed to the board before the current CEO took office are negatively associated with favorable timing of CEO stock option awards. However, Daily et al. (1998) find no evidence that the proportion of interdependent directors on a compensation committee lead to higher executive compensation. Overall, empirical evidence from the US is indicative for the notion that the distribution of power between boards and CEOs expressed as the proportion of interdependent directors is an important determinant of the level of executive compensation. Studies on the relationship between interdependent directors and the structure of CEO compensation are scarce. Westphal and Zajac (1994) document that interdependent directors are positively associated with the adoption of long-term incentive plans, but they do not influence the actual use of incentive plans in terms of their size. Daily et al. (1998) report no significant relationship between interdependent directors on the compensation committee and the CEO contingent pay ratio.

6.5 Network power and executive compensation

A stream of research on managerial elites has focused on interlocking directorates and their importance as arenas of power. However, this stream of research has been largely focusing on the structural anatomy of director ties, neglecting the meaning and significance of interlocking directorates (Pettigrew, 1992). Network theorists have suggested that social ties between individuals that extend across different contexts (e.g. on different boards) will lead to particularly high levels of social cohesion (Emirbayer and Goodwin, 1994). Managers who have entrenched themselves in a self-compiled network of people and groups face numerous obligations, but likewise, can count on reciprocity in matters that are important to them, including their compensation package. When individuals have common memberships on multiple other

92

CHAPTER 6

boards with each other, they have more shared experiences and a stronger basis for mutual identification (Belliveau et al., 1996). It is frequently claimed that the pool of suitable candidates for top positions on the board and the top management team in Switzerland is limited. This suggests that top executives and board members build an exclusive elite group that knows each other from shared board memberships (Keller, 2003), exclusive network groups or events such as the American Chamber of Commerce and the World Economic Forum, or alumni clubs such as the one from Harvard Business School.

6.5.1 Executive board memberships

Outside board memberships provide prestigious contacts with other organizations (Zajac, 1988; Fama and Jensen, 1983; Finkelstein, 1992; Belliveau et al., 1996). This is especially true for board memberships on large and successful companies (Useem, 1979). Service on other boards may indicate an increased ability to manage interorganizational dependencies through increased access to information. Executives serving on other firms board have the opportunity to interact with other prestigious individuals several times a year. This contact with other elite individuals may enhance the focal organizations legitimacy. On a personal basis, non-executive directors interested in further board memberships may see the CEO who serves on other firms boards potentially on the nominating committee as an important link. Overall, outside board membership gives executives more power vis--vis their boards. Only two empirical studies investigate the relationship between outside board memberships and compensation in the US. Geletkanycz et al. (2001) find weak evidence that CEO external ties are a determinant of CEO compensation.31 Wade et al. (1990) document that CEOs with a high number of outside board seats are more likely to receive golden parachutes agreements. In addition, Belliveau et al. (1996) use the number of board memberships as part of a wider construct of CEO status that is positively associated with CEO cash compensation. In these papers, the primary reason for higher compensation levels is that CEO external directorate networks are strategically valuable to firms and should therefore be reflected in higher compensation.

Geletkanycz et al. (2001) do not use a power perspective to explain this result. They argue that boards compensate their CEOs for their external directorate network, because it is strategically valuable to firms.

31

CAN POWERFUL MANAGERS EXTRACT RENTS?

93

Conyon et al. (2001) provide supportive evidence from the UK. They document that CEOs and other executives with more outside board mandates receive higher cash and total compensation.

6.5.2 Reciprocal and social interlocks

An interlock exists if a member of the top management team of firm A also sits on another firm Bs board of directors or top management team. Reciprocal interlocks are interlocks in which a top management team member of firm A sits on another firm Bs board of directors, and a member of firm Bs top management team sits on firm As board of directors. In addition to reciprocal interlocks, non-executive and executive directors may also meet non-executive directors on the board of a third company (Belliveau et al., 1996), a situation that I term social interlock. The press has criticized interlocking directorates as intolerable practice in Swiss corporate governance (Wittwer, 2005b: 27). The basic hypothesis that reciprocal interlocks can lead to higher levels of compensation for both CEOs can be formalized by thinking of a simple model in which each board has only one director (Hallock, 1999). In the absence of interlock, the CEO of firm A is an agent for the owners of the firm represented by the board. Standard principal-agent models suggest that the board will try to design a compensation package to maximize the value of the firm, after paying the CEO. Suppose now that CEO1 is the director of firm B and CEO2 is the director of firm A. Then these two CEOs could be thought of as a pair of agents who ideally have to negotiate with the collective set of owners of firm A and firm B. If the compensation of CEO1 is left in the hands of CEO2 and vice versa, it is difficult to imagine that their compensation will be set at the optimal levels from the perspective of the firms. Interlocking directorates can potentially threaten the boards ability to objectively set compensation levels on behalf of the firms owners. Little empirical evidence exists on the influence of interlocking directorates on executive compensation. Hallock (1997, 1999) finds that CEOs heading reciprocally interlocked firms earn substantially more than CEOs who are not interlocked, measured as cash compensation. However, his results are not robust to the inclusion of commonly used control variables. Core et al. (1999) find no relationship between the percentage of

94

CHAPTER 6

interlocked outside directors32 and the level of CEO compensation. Hanlon et al. (2003) document that the presence of interlocked directors on the compensation committee appears to increase stock option grants. Fich and White (2005) study determinants of CEO reciprocal interlocks for a cross-section of firms in 1991. Their results show that CEO reciprocal interlocks and the proportion of stock option compensation are negatively associated consistent with the power hypothesis. To my knowledge, no test has been undertaken on total compensation levels comprised of both cash compensation and equity-based compensation in the form of shares or share options.

6.5.3 Non-executive board memberships

Similarly to top managers, non-executive board members may gain power from external contacts (Pettigrew and McNulty, 1995). Relationships outside the firm can help nonexecutive board members to build a power-base. Such relationships suggest that the individual has high qualifications and powerful friends.

6.6 Credibility power and executive compensation

6.6.1 CEO celebrity status
Celebrity is another source of CEO power that may influence compensation (Elson, 2003). A firms legitimacy depends in part on the prestige of its managers (DAveni, 1990). Serving as a figurehead, the CEO constitutes the liaison between the firm and its environment (Mintzberg, 1973). As the symbolic head of the company, the CEO is responsible for defining and conveying organizational meaning. Chief executive officers increasingly represent the face of their company to the outside world, which may occasionally turn into a kind of personality cult. As legal authorities of their firms, managers act as symbols and are obliged to perform symbolic activities. In a broader context, top managers often act as boundary-spanners to owners, government, employee groups and the general public. A large part of power derives from the managers ability to manage symbolic activity (Ungson and Steers, 1984). Examples may include former CEO and Chairman of ABB, Percy Barnevik, or the CEO and Chairman of Novartis, Daniel Vasella. As the company and its CEO are intertwined in public opinion, the departure of a CEO of high personal standing would result in a significant negative share

Core et al. (1999) define an outside director to be interlocked if an inside director of the firm serves on the board of that outside directors firm.

32

CAN POWERFUL MANAGERS EXTRACT RENTS?

95

price reaction. CEOs aware of their power arising from being the face of a company may be able to use this power to increase their compensation. Schildknecht (2004) suggests that CEOs through public relations consultants try to create a successful image. Nonexecutive directors charged with negotiating the CEOs compensation package and suffering from information asymmetry with respect to the CEOs performance will be influenced by such a creation of a successful leader. The proposition that CEO celebrity status may influence CEO compensation practices has not been directly studied yet. However, three related studies suggest a relation between celebrity status and compensation. Using survey data, Tosi and Gomez-Mejia (1989) suggest that executives can take advantage of the board, particularly when ownership is highly dispersed, by using personal charisma to extract higher pay without performance contingencies. Tosi et al. (2004) find a positive association between perceived charisma and total compensation. Milbourn (2003) investigates the relationship between CEO reputation and equity-based compensation and documents that CEOs who are more present in the business press have higher overall stock-based pay sensitivities. This last result is not consistent with the notion of managerial preferences for non-contingent pay.33

6.6.2 Prior firm performance

Prior firm performance may also be a source of managerial power in future negotiations as it helps building a reputation as a successful manager and thus credibility (Fama, 1980; Johnson and Daily, 1997; Adams, 2004). Zajac and Westphal (1996a) show that, similar to their measure of relative board power, poor prior performance leads to change in CEO characteristics in the direction of board member characteristics. That is, the CEO becomes less powerful after poor prior performance.34 In an examination of CEO succession, Ocasio (1994) finds that CEO power diminishes under conditions of economic adversity. When firm performance is poor, the balance of power shifts to the board. Siegel and Brockner (2005) posit that observers will be more positively disposed towards CEOs, whose firm has performed well in the past. If prior performance is

Milbourn (2003) asserts that high-reputed CEOs are more likely to be retained. As a result, stock prices reflect a larger part of the incumbents CEOs performance and equity-based compensation is more efficient. Zajac and Westphal (1996b) do not model prior poor firm performance as a source of relative board power over the CEO but rather view it as a reinforcing economic rationale for change (1996: 81).
34

33

96

CHAPTER 6

favorable, the CEO will have even greater credibility. In a laboratory experiment, they find that more favorable prior firm performance is associated with greater subsequent raises in CEO salaries. In their study, external claimed handicaps had a positive effect on the CEOs compensation when prior performance was good, but a negative effect on the CEOs pay when prior performance was bad. This shows the importance of prior firm performance when observers interpret claimed handicapping. They interpret the results that when prior firm performance is negative, the CEO who claims an external handicap may simply be viewed as someone who identifies obstacles to performance but does nothing to ameliorate these problems. The combination of bad prior performance and external claimed handicapping may reinforce stakeholders (investors or directors) negative impression of the CEO and the possibility that he may not have sufficient control over the firms operations. On the other hand, a strong reputation (in this case built from a good track record) provides leaders with leverage for defending against possible challenges to legitimacy. Ryan and Wiggins (2004) find that non-executive director equity-based compensation is negatively related to past accounting performance and interpret this result as an indication that directors of firms in which the CEO achieves superior accounting performance are given fewer incentives to monitor (Ryan and Wiggins, 2004: 510). This interpretation suggests again a power shift towards the CEO after good past performance. Another indication that performance increases CEO power is found by Fee and Hadlock (2003). They observe a positive relationship between firm performance and the likelihood that a manager moves to a superior position at another firm. There exists some empirical evidence that prior performance as a power basis influences executive pay.35 Fee and Hadlock (2003) show that superior performance at the prior firm is associated with a more valuable hiring grant, when managers get promoted to CEO positions of other large companies. Brickley et al. (1997) show that prior firm performance during a CEOs tenure is positively associated with both salary and total compensation.

Generally speaking, all pay-for-performance studies can be viewed in a power context. However, the dominant research design in pay-for-performance studies of tracking current changes in performance with changes in pay seems ill-specified for a power explanation.

35

CAN POWERFUL MANAGERS EXTRACT RENTS?

97

6.6.3 Expertise
The board of directors must engage a top management team who has been highly trained to do complex work (Mintzberg, 1973). As a signaling device, the board may take top managers experience as an indicator of whether they will be successful at their job. In addition to that, the board may also take a good education as evidence for managers to be able to do a good job. Therefore, a reputable education may also provide some indication of a CEOs level of credibility. There is empirical evidence that higher education is associated with higher pay (Carpenter and Wade, 2002; Angel and Fumas, 1997).

6.7 Summary of the literature on power pay relation

Table 5 provides an overview of prior empirical research linking executive power variables to level of executive compensation.36 It becomes clear that for all variables, conflicting evidence can be found. CEO duality is the clearest predictor of higher compensation levels, but at least two studies cannot find a significant relationship. Other measures such as CEO tenure, board size or outsider ratio have returned ambiguous results.
Table 5: Major empirical studies on the CEO/board power variables and CEO pay level.
STUDY SAMPLE (FIRMS) COUNTRY YEARS COMPENSATION VARIABLES POWER VARIABLES CEO tenure CEO tenure squared Finkelstein and Hambrick (1989) 68 US 1971, 76, -82, -83 CEO cash compensation CEO ownership CEOs family holdings Non-executive holdings Main (1991) 241 UK 1985 Cash of highest paid director CEO cash compensation CEO duality Board size Outsider ratio CEO duality RESULTS + 0 0 + + + +

Boyd (1994)

193

US

1980

Not all these studies use managerial power as their research hypotheses, but they study powerrelated variables.

36

98

CHAPTER 6

Henderson and Fredrickson (1996) Sridharan (1996) Cosh and Hughes (1997) Hallock (1997) Gray and Canella (1997) Conyon and Peck (1998)

131

US

1980, 1990 19891991 19891994 1992 19801989 19911994

Total CEO compensation CEO cash compensation CEO cash compensation CEO cash compensation Total CEO compensation Cash of highest paid director

CEO tenure

+ 0 0 0 0 + + + 0 + 0 0 0 0 + + 0 0

167

US

CEO duality Outsider ratio Outsider ratio Reciprocal interlocks CEO duality Outsider ratio CEO duality CEO duality Board size Outsider ratio Interlocked outsiders Interdependent ratio Outside ownership

64 602 100

UK US US

94

UK

Core et al. (1999)

205

US

19821984

Total CEO compensation

Eriksson (2000) Carpenter et al. (2001) Miller et al. (2002)

317

DK

19921995 19941996 19941998

CEO cash compensation Total CEO compensation Total CEO compensation

CEO tenure CEO ownership CEO tenure CEO tenure CEO ownership CC ownership

245

US

423

US

Cyert et al. (2002)

1648

US

19921993

Total CEO compensation

CEO tenure CEO duality Largest shareholder

Coombs and Gilley (2005)

406

US

19952001

Total CEO compensation

CEO ownership CEO tenure

+ means a significant positive association; - means a significant negative association; 0 means no significant relationship

Table 6 shows major empirical studies on the CEO power pay mix relation. It shows that we know relatively little about how managerial power affects pay structure. The two power variables CEO tenure and CEO stock ownership have been investigated in several

CAN POWERFUL MANAGERS EXTRACT RENTS?

99

empirical studies. There is evidence that power in terms of ownership is negatively associated with equity pay mix, while the relationship between CEO tenure and equity based compensation remains unclear.
Table 6: Empirical studies on CEO power pay mix relation.
STUDY Westphal and Zajac (1994) Zajac and Westphal (1994) Mehran (1995) SAMPLE (FIRMS) 669 COUNTRY US YEARS 1972-1990 COMPENSATION VARIABLES % of CEO equity comp % of CEO equity comp % of CEO equity comp % of CEO equity comp % of CEO long-term comp % of CEO options comp % of CEO options comp; % of CEO stock comp % contingent pay % of CEO options comp % of CEO equity comp % of CEO options comp % of CEO options comp POWER VARIABLES CEO tenure CEO ownership Largest shareholder Non-executive directors CEO ownership Non-executive directors CEO ownership CEO tenure Outsider ratio CEO duality CEO tenure CEO on compensation committee CEO tenure Interlocks CEO ownership RESULTS -

405

US

1987-1991

153

US

1979-1980

Sanders and Carpenter (1998) David et al. (1998) Bryan et al. (2000a) Ryan and Wiggins (2001) Miller et al. (2002) Anderson and Bizjak (2003) Bushman et al. (2004) Fich and White (2005) Khan et al. (2005)

258

US

1992

125 1788

US US

1990-1994 1992-1997

+ 0 0 0 0 -

1095

US

1997

423 110 784 576 224

US US US US US

1994-1998 1985-1998 1994-1997 1991 1992-2000

+ means a significant positive association; - means a significant negative association; 0 means no significant relationship

100

CHAPTER 6

6.8 Literature gap and power hypotheses

The literature review on ownership, structural and tenure power in relation to executive compensation shows that while theoretical grounds for the proposed relationships are strong, empirical evidence is often ambiguous and contradictory, and sometimes lacking. Several explanations are possible for these inconsistent results. First, using single variables to proxy for executive power may not capture the full influence of power on compensation, although the individual variable may contribute to the construct of power. Each of the indicators by itself may be an insufficient measure of executive power, because each addresses a relatively narrow slice of the construct. Any single power measure may be unlikely to capture the breadth and qualitative nature of managerial power as a determinant of executive pay. Second, inconsistent results may be due to conflicting hypothesis, which makes it difficult to discern the explanatory effect of power. For instance, while the power hypothesis suggests a negative relationship between compensation committee existence and compensation, it may be argued that compensation committees will be associated with higher managerial pay through a bidding-up process (Ezzamel and Watson, 1998; Conyon and Peck, 1998). As another example, the structural power hypothesis suggests that non-executive board members are associated with higher reliance on equity-based compensation. It may, however be argued that the proportion of non-executive directors be negatively related to equitybased pay consistent with a monitoring-substitution hypothesis (Beatty and Zajac, 1994). Depending on the dominant influence, the resulting relationship may come out differently. Third, in addition to opposing hypotheses, some variables, which are hypothesized to indicate managerial power, have also been used to explain other constructs. For instance, CEO tenure was used as an indicator of human capital accumulation (Mayers and Smith, 1992). It is unclear whether the effect of CEO tenure is attributable to CEO human capital or CEO power. It is equally unclear whether the negative relationship between CEO stock ownership and equity-based pay established in the literature is attributable to CEO power and preferences (Allen, 1981a; Hill and Phan, 1991) or to a substitution mechanism between incentives arising from stock ownership and equity-pay (Jensen and Meckling, 1976). Such caveats may be overcome by developing and testing a comprehensive framework of power that captures the power nuances of individual variables (Finkelstein and Hambrick, 1996; Finkelstein and Boyd, 1998). Finally, it is worth noting that many of the cited studies do not explicitly use a

CAN POWERFUL MANAGERS EXTRACT RENTS?

101

power hypothesis in their analysis, but that their variables are important pieces to the construct of power. In sum, there is strong conceptual support for the idea that managerial power is related to executive compensation. However, as the literature review has revealed, traditional measures of managerial power have failed to provide consistent evidence. One reason for this may be that many variables individually may contribute only marginally to explaining a link between power and pay, but in sum they may explain significant variance. Further, in some dimensions of power, empirical work is scarce. As with complexity, what is missing is a formal test of the power hypothesis, one that incorporates multiple aspects of executive power and allows for a more definitive statement about the effects of power on executive compensation. Thus, I hypothesize: Hypothesis 2a: The greater the power of the CEO and other members of the top management team, the greater the level of executive compensation. Hypothesis 2b: The greater the power of the CEO and other members of the top management team, the lower the proportion of equity-based compensation.

102

CHAPTER 7

Research methods

7.1 Development of hypotheses

The hypotheses have emerged out of a large-scale review of around 450 articles on executive compensation and related fields, review meetings with the referees of this dissertation and my work at the multinational corporation ABB. Working for the head of executive remuneration, I was involved in the design and development of a group-wide Employee Share Acquisition Plan (ABB ESAP) at ABB Global Headquarters and the project manager for the implementation of this plan within ABB Australia. Among others, this job assignment enabled me to conduct interviews with experts in the field of executive remuneration, thus gaining a deeper understanding of the top management pay scene. I conducted interviews with Gary Steel, Head of HR, member of the Executive Committee and secretary to the Compensation Committee of ABB, David Tankel, Managing Director of London-based New Bridge Street Consultants and expert on executive remuneration, Hanspeter Fssler, member of the compensation committee of Daetwyler, and Prof. Dr. Rolf Dubs, chairman of the compensation committee of Schindler and member of the board of directors of Julius Br. With the exception of the consultancy, all of these firms are included in my sample. Furthermore, I have gained insight into the processes and workings of top management teams in my current position as an assistant to the Executive Committee of ABB Switzerland Ltd, the ABB Groups Swiss country organization employing 5, 000 employees. In this position, I have regularly attended top management and board of directors meetings.

7.2 Sample and data

7.2.1 Archival data
As the starting point of my analysis, I use data from companies listed at the Swiss stock exchange SWX in 2002 and 2003. I target all 289 firms quoted at the Swiss Exchange (SWX) by the end of 2003. The exclusion of investment companies leaves me with 235 firms. I further delete companies that have not been quoted over the sample period or

RESEARCH METHODS

103

disclose insufficient information. The final sample consists of 199 firms and 398 firmyear observations. The sample has a few note-worthy characteristics. Swiss stock-listed companies are characterized by large variations in terms of compensation structure, degree of managerial power or degree of complexity (Hengartner, 2003b). Such observable differences will make statistical tests more powerful. Second, prior research has focused on very large companies. This study comprises both larger and smaller firms providing a more balanced picture of the executive compensation landscape. It further provides a rather complete picture of executive compensation in Switzerland by potentially including all stock-listed companies.37 Third, I study companies from a wide variety of industries. Prior studies have usually either excluded banks or other financial services companies from their sample, or they exclusively focused on banks. The rationale for including firms from a broad range of industries was that it will enhance the external generalizability of the studys results. I use two years of data, although I am not especially concerned with controlling for firmspecific fixed effects over time. Rather, as Murphy (1999) proposes and the result of Balkin et al. (2000) show, the variance explained by commonly used variables may fluctuate significantly across time. Also, estimators based on panel data are quite often more accurate. Even with identical sample sizes, the use of a panel data set will often yield more efficient estimators than a series of independent cross-sections (Verbeek, 2004). The panel data approach further helps ameliorating the problem of omitted variables associated with a cross-sectional sample (Murphy, 1985). I obtained data on the corporate governance variables executive remuneration, shareholdings, board composition and shareholder structure from companies annual reports. Data on firm complexity variables and control variables firm internationalization, diversification, dividend yield and volatility are from the Share Guide 2002/2003, Share Guide 2003/2004, the Share Guide 2004/2005 all published by Finanz und Wirtschaft, and from Top 2002, Top 2003 and Top 2004 all published by
37

I have not included privately held firms for two reasons. First, data availability is largely restricted in such

firms. Second, there is evidence that privately held firms differ substantially in the compensation of their managers due to differences in corporate governance (e.g. Ke et al., 1999).

104

CHAPTER 7

Handelszeitung. I collected missing data through an email and telephone follow-up to ensure completeness of data.

7.2.2 Survey
In addition to archival data collection, I measured executive power and complexity attributes through a questionnaire in June 2005 to Chairmen of the board or the compensation committee of Swiss stock-listed companies, excluding investment companies. The questionnaire is reproduced in the appendix. It attempts to capture more subtle and perceptual elements of complexity and managerial power. Applying both objective as well as perceptual measures of complexity is a generally preferred approach over using either objective or perceptual measures only (Finkelstein, 1992). For instance, it has been argued that expertise and knowledge are a source of power (Abernethy and Vagnoni, 2004). Therefore, I measured expertise power by the subjective evaluation of the chairman of the board or the chairman of the compensation committee of the CEOs expertise and market knowledge. To reduce response bias, items measuring each construct were scattered throughout the survey. Moreover, the title of the questionnaire was kept neutral and questions worded carefully to minimize the likelihood of a social desirability bias. Because surveys of top managers and members of board of directors have often had low response rates, I took several steps to increase the response rate: (1) I used a pretest involving an in-depth pilot interview with a senior executive of a large company to revise the survey, making it easier to complete. I also used this interview to ensure that questions were interpreted as expected. (2) a second wave of questionnaires was sent to nonrespondents three weeks after the initial mailings. (3) Participants were invited to either fill in a German or a English version of the survey. I received 50 responses, representing a response rate of 21%, and was able to use 47 of the responses in the statistical analysis.

RESEARCH METHODS

105

7.3 Model specification

7.3.1 Regression analysis
The model regresses the natural logarithms of total compensation and the proportion of equity-based compensation on factor-analyzed managerial power and complexity measures. The total compensation model looks as follows: Yit = a0 + a1X1it + a2X2it + a3 X3it + e1it In this regression, Yit is the dependent variable of interest (e.g. log of CEO total compensation, total TMT compensation, equity-pay mix) for firm i in year t and Xit are explanatory and control variables. X1 is a vector containing executive power variables, X2 a vector containing variables measuring the complexity of a firm and its environment, and X3 other determinants of executive compensation. When single variables are used, all variables are expressed in natural logs with the exception of dummy variables and proportions. To reduce the number of variables and to detect structure in the relationship of variables, I apply principal component analysis to the concepts of complexity and executive power. The use of principal component analysis also helps alleviate the problem associated with multicollinearity. Using the measured proportion of equity-based compensation as the dependent variable introduces potential econometric problems as this variable is bounded in the unit interval. Such boundedness implies that the assumption of a normally distributed error term is not tenable (Verbeek, 2004). Further, as indicated in the descriptive statistics section, the equity proportion of total compensation is frequently at the lower bound. Recognizing these issues and following Barrell and Waddell (2003a), I rephrase my question concerning the proportion of compensation awarded in a particular fashion to take the following form: What determines the likelihood that a given Swiss Franc of compensation is equity-based? thereby handling both the unit interval and the lumpiness of the proportional data. Following prior research (e.g. Ryan and Wiggins, 2004; Yermack, 1995; Anderson and Bizjack, 2003) I use Tobit panel data regressions

106

CHAPTER 7

instead of regular OLS techniques in regression of equity-pay-mix, as Tobit estimation better accommodates censored variables38 (Greene, 2000).

7.3.2 Unit of analysis

Under the assumption that the findings applicable to CEO pay can be generalized to top management team pay, research on executive compensation typically focuses on the CEO (Gray and Cannella, 1997). Listed companies at the Swiss stock exchange are required to disclose the compensation of the highest paid member of the board of directors, but they do not need to disclose the name of this individual. In many cases, companies explicitly disclose CEO compensation. The general rule is to keep the highest paid director variable in the regression of CEO compensation if total compensation for this executive is higher than the average total compensation for top management team members. If this value is smaller, I use average total TMT compensation. This rule may include a few executive board chairmen in regression of CEO compensation. To the extent that this is the case, results may not fully represent and perhaps slightly underestimate the true effects on CEO pay. I also collect data on the overall compensation for executive board members and top executive management as a whole. It has been suggested that consideration of the top management group provides better predictions of organizational outcomes than those attributed to the CEO alone (Cannella and Monroe, 1997: 222). Recent research on executive compensation has moved beyond the focus on the CEO to study a wider number of top executives (Carpenter and Sanders, 2004; Schaefer, 1998; Main et al., 1996; Murphy, 2003; Bushman et al., 2004; Carpenter and Sanders, 2002) indicating that firm performance is more greatly enhanced by high TMT compensation than by high CEO compensation and that the benefits of long-term forms of pay appear to be more pronounced at the level of the TMT than at the level of the CEO (Carpenter and Sanders, 2004; Carpenter and Sanders, 2002). Also, the analysis of compensation for executives below the very top of the corporate hierarchy is warranted because many strategic decisions (including the setting of objectives or investment decisions) are made by top management as a team (Lambert et al., 1991; Main et al., 1996; Coles et al., 2002). As researchers are increasingly acknowledging the importance of the TMT in influencing

Censored variables have a large fraction of observations at the minimum or maximum. In the case of equity-pay structure, a large number of sample companies will have zero values.

38

RESEARCH METHODS

107

organizational outcomes, understanding the determinants of TMT compensation is crucial. The team-approach is arguably even more pronounced in Switzerland with its consensus-oriented culture compared with US companies (Ang et al., 2002; Hofstede, 1980). With respect to managerial power Finkelstein (1992: 505) notes that in an organization in which the chief executive officer (CEO) yields dominant power, studying only the CEO may provide sufficient information with which to test propositions. However, in organizations in which power is less polarized, consideration of a coalition of top managers is necessary to fully capture the range of managerial orientations prevailing. Further, while total CEO compensation may occasionally still be a relatively small part of corporate earnings TMT compensation levels more truly reflect the companys costs incurred for paying their executives (Crystal, 1991: 264). For instance, UBS paid its three executive board members and seven TMT members approximately 140 MCHF in total compensation in 2002, which represented 4.8% of the companys profits. I aggregate the value of all compensation components for all TMT members of each firm. Implicit in this aggregation is the assumption that the firms compensation policy (e.g. the use/non-use of stock options) applies uniformly to top executives of different ranks. For firms where individual compensation data is available, the data support this assumption. Furthermore, prior research has shown that this hierarchical assumption is reasonable (Werner and Tosi, 1995; Fatemi et al., 2003). CEO compensation level and structure is a strong albeit imperfect predictor of TMT compensation (Carpenter and Sanders, 2002). I use average compensation values to control for TMT size.39

7.4 Measuring executive compensation

The expansion of corporate governance disclosure requirements in 2002 significantly increased public information about top managers pay in Switzerland. I use several measures of compensation in my analysis. The first is the total cash compensation paid to the executive during the year. The second compensation measure total compensation is the sum of all cash (salary plus bonus), options and shares awards made to an

Using average values instead of total TMT compensation is especially relevant in the context of my data, as TMT size differs across companies.

39

108

CHAPTER 7

executive during the year. The third measure equity-based compensation mix40 is the proportion of the sum of options and shares awards to total compensation. This operationalization of compensation measures has become standard practice in this area of research (see Murphy, 1999). Compensation amounts indicated for executives serving less than the full year are annualized. In the analyses, the natural logarithms of all compensation measures are used to reduce heteroskedasticity. The measurements of the individual components of the executive compensation package are discussed below. All compensation measures are based on data published in companies annual reports. I checked the accuracy of my numbers by cross-checking with compensation data published by the Handelszeitung (Speck, 2003: 12, 13; Handelszeitung 2004: 87, 88; Speck, 2005: 8).

7.4.1 Cash compensation

Cash compensation represents the Swiss Franc value of the base salary and the cash bonus received during the year including employers contributions to pension plans.

7.4.2 Share awards

Shares are valued by multiplying the number of shares awarded by the share price on the date of the grant. If the grant date was not disclosed, the share price at the end of the fiscal year was applied. If shares are not awarded free of charge, but granted at a discount, the difference between the prevalent share price and the purchase price was multiplied by the number of shares awarded to indicate the monetary benefit arising from the award.

7.4.3 Stock options

Stock options are rights to purchase company stock at a specified exercise price over a stated option term. The value of stock option grants is calculated using the BlackScholes (1973) option valuation model adjusted for continuous dividends (Merton, 1973), which has become the standard practice in the executive compensation literature to estimate the value of executive options (Conyon and Sadler, 2001).41

I use the terms pay mix, pay structure, compensation mix and compensation structure interchangeably. It has recently been suggested that executives value their stock options at a discount as compared to standard Black-Scholes values (Meulbroek, 2001; Hall and Murphy, 2002). However, this value depends both on the proportion of the executives wealth tied to firm value and his risk-aversion
41

40

RESEARCH METHODS

109

The information reported on the exercise price varies across firms and I made reasonable assumptions in case of missing information. If the firm disclosed the exercise price, I used the actual price. In the rare cases where the firm disclosed a range of prices, I used the mean price. Following Yermack (1995), Hall and Liebman (1998) and Fee and Hadlock (2003), I assumed that options are granted at the money, if not otherwise indicated. A few companies disclose the tax value of their options. In these cases, I also value the options granted based on the Black-Scholes value. This is especially important, as the Black-Scholes value increases with the time to maturity of the option, while the tax value is a decreasing function of the options vesting period and as such also time to maturity. That is, options with longer blocking periods are subject to a lower tax value. Generally, Conyon and Sadler (2001) argue that few economic biases arise from less than complete information disclosure of stock option grants. Furthermore, alternative and less sophisticated approaches to valuing stock options have provided similar results and appear to be valuable (Lambert et al., 1991b; Finkelstein and Boyd, 1998; Balkin et al., 2000). For instance, Lambert et al. (1993) and Henderson and Fredrickson (1996, 2001) all value stock options at 25% of their exercise price and conclude that this method produces values in the same range as more sophisticated option-pricing methods such as the Black-Scholes model (1996: 585). Kerr and Kren (1992) obtain similar results for different option valuation methods. The discussion suggests that valuing stock options is not as straightforward a process as valuing cash compensation and the results will always reflect the assumptions used to value that pay. Nevertheless, the discussion also suggests that few biases should arise from my approach to option valuation.

7.5 Measuring complexity

7.5.1 Archival data
Firm internationalization is measured as the proportion of foreign employees to total employees (FETE). This measure captures the structural attribute of internationalization. Sullivan (1994) suggests that the degree of internationalization (DOI) of a firm has three

data that is not available. On the other hand, to the extent that managers are able to influence option grant dates (Yermack, 1997), information flows around grant dates (Chauvin and Shenoy, 2001) and use inside information to exercise options (Carpenter and Remmers, 2001), the Black-Scholes value may actually underestimate the value of executive stock options.

110

CHAPTER 7

attributes: structural, attitudinal and performance. To capture the complexity arising from a firms DOI, the structural attribute appears most relevant, as it proxies well for the description of complexity in multinational corporations (Bartlett and Ghoshal, 2002). Sullivans attitudinal attribute relates to top managements international orientation and experience. However, rather than reflecting a dimension of international complexity itself, this attribute measures top managements ability to cope with such complexity (Carpenter et al., 2001). In the sensitivity analysis, I also include the performance attribute of internationalization, measured as the proportion of foreign sales to total sales (see 8.6). Firm size is measured as the logarithm of sales revenues. Sales revenues for financial firms are defined as gross income (banks) and gross premium income (insurance companies). Firm diversification is measured as the number of different business lines. As the FASB notes in its statement No. 14 paragraph 12 determination of an enterprises industry segments must depend to a considerable extent on the judgment of the enterprise. I gauge the number of different business lines by the number of segments reported in the companies annual reports, the Standard Industry Classification (SIC) and the company descriptions in the Share Guides. Following previous studies on firm diversification (e.g. Rose and Shepard, 1997) I use a dummy variable to differentiate between companies with only one major business line and all other companies. I also measure firm diversification as the number of different business lines at the three-digits SIC level in which the firm operates, following Agrawal and Knoeber (1998). I calculated firm-level sales volatility measured over the period 19982003 to indicate market uncertainty. This long period is likely to capture a large part of the business cycle. I proxy for a politicized environment by the number of newspaper articles published in the NZZ. Under the assumption that newspapers continuously strive to publishing articles that the public is interested in, the number of published articles may serve as a good proxy for the politicized environment of a particular firm. Also, the media have a high potential influence on public opinion and the opinion of employees, which managers may have to cope with.

7.5.2 Survey data

Strategic thinking environment. The questionnaire sent out to chairmen of the boards or the compensation committee respectively asked participants to estimate how complex the operative / strategic thinking environment (thinking challenge, problem solving challenge) of their CEOs job was. Answers were coded into a scale between 1 (not

RESEARCH METHODS

111

complex at all) and 4 (very complex). Perceived complexity. The questionnaire contained three questions estimating the technical, management and human relations skills necessary for the CEOs job. The three answers were coded into scales of 1-4 and added. Thus, perceived complexity ranges from a minimum of 3 to a maximum of 12. Subsequently, I created a dummy variable of high degrees of job complexity, if the total of the perceived complexity was 9 or higher. Number of direct reports. Complexity was also gauged by the number of people that directly report to the CEO.

7.6 Measuring executive power

7.6.1 Ownership power
Following Habib and Ljungqvist (2003) the executive ownership and non-executive ownership variables are measured as the number of shares and options owned by the top management team and the board of directors as a proportion of total shares outstanding, indicating the relative power of the top management team (fractional ownership). CEO equity ownership is not routinely disclosed for Swiss stock-listed companies. As I am interested in measuring the power of executives and not incentives, I do not use total equity ownership but rather ownership percentage. Conversely, measuring a managers dollar ownership of the firm as used in Core and Guay (1999) is more appropriate in measuring equity incentives (Core et al., 2003). Because a considerable fraction of Swiss firms deviate from the one vote per share rule (Economist, 2005b), I use the percentage of voting rights as in Barkema and Pennings (1998). In 20 cases for 2002 and 12 cases for 2003, executive shares and option holdings cannot be determined. In these cases, I use average values. In 24 cases for 2002 and 11 cases for 2003, non-executive shares and option holdings cannot be determined. In these cases, I also use average values (Greene, 2000). Largest shareholder measures the percentage of equity owned by the largest outside shareholder. Following Conyon and Peck (1998), this refined measure capitalizes on the variation in ownership by treating the ownership of the largest shareholder as a continuous variable. This treatment stands in contrast to considering blockholders categorically by indicating the mere presence or absence of an owner with at least 5% of

112

CHAPTER 7

the organizations outstanding shares as ownership concentration.42 The latter approach, used by most prior empirical studies (e.g. OReilly et al., 1988; Hambrick and Finkelstein, 1995) does not test for the idea that a shareholder may have increased incentive to monitor management with an increased ownership stake. International comparisons suggest that large shareholders are especially prevalent in Switzerland as opposed to the US (La Porta et al., 1999). In Switzerland, the one share one vote-rule is not uniformly applied in all stock listed companies (Economist, 2005b). As I am interested in measuring the power of the largest shareholder, I used the voting percentage rather than the ownership percentage.

7.6.2 Structural power

Non-executive directors. I measure the proportion of non-executive directors out of total directors on a firms board. Independent directors. I measure the proportion of independent directors out of total directors on a firms board. Independence is defined following the definition used in the Swiss Code of Best Practice (2002: 8). These directors are non-executive, have not been a member of the executive management for at least three years and do not have any material business relationship with the company. Not all companies disclose explicitly whether their directors are independent or not in the sense of the Swiss Code of Best Practice. The determination of a board members independence may occasionally require the researchers judgment based on the study of the curriculum and other pieces of information disclosed in the annual report, such as compensation paid by the company to services firms such as law firms, at which nonexecutive board members are employed. CEO duality is measured as a dummy variable 1, if the CEO is also the chairman of the Board of Directors, and 0 otherwise. Compensation committee existence. I code companies that have a compensation committee with 1, and companies without such a committee with 0. If a company just has one board committee and if this one committee also determines the compensation of the CEO and other executives, it is coded as a compensation committee, as it supports to notion of power through the establishment of specialized committees (Pearce and Zahra, 1991). Compensation committee composition is measured as the proportion of executive directors out of total directors sitting on this committee. This variable can have

Hambrick and Finkelstein (1995: 183) argue 5 per cent is an appropriate cutoff for attributing a strong, vigilant owner.

42

RESEARCH METHODS

113

considerable lag, as compensation is unlikely to be decreased once it has reached a certain level (e.g. Bertrand and Mullainathan, 2001). Schtz (2005) argues that Daniel Vasella managed to make a higher salary in his time as chairman of the compensation committee in 2000 and 2001. His total compensation increased from an estimated 5 MCHF in 1999 to disclosed 20 MCHF in 2001, where it remained ever since.43 Vasella resigned from the compensation committee in 2001. I measure board size as the number of members sitting on a firms board of directors.

7.6.3 Tenure power

CEO tenure is measured as the log of the number of years the CEO has been in office. The log is used because it has been suggested that CEO power resulting from his tenure reaches an upper limit over time (Finkelstein and Hambrick, 1988). Following Daily and Johnson (1997), the proportion of interdependent directors is calculated as the proportion of non-executive board members appointed during an incumbent CEOs tenure to total board size.

7.6.4 Network power

Interlocks are a count measure of total interlocks on the board. They are either reciprocal or social in nature. I count a non-executive board member as reciprocally interlocked, if he is an executive of another company, on which board a top management team member is a nonexecutive director. This is consistent with prior research (Hallock, 1997, 1999; Core et al., 1999), except for the fact that prior studies only investigated CEO reciprocal interlocks. Social interlock refers to a situation, where the CEO meets a non-executive director of his firms board on the board of another firm.

Outside executives board memberships. I measure the number of board memberships on other stock-listed companies in Switzerland for all executive members of the board and members of the top management team. The count measure used in the analyses is the number of such memberships. Outside non-executives board memberships. I count the number of board membership on other stock-listed companies in Switzerland for all nonexecutive members of the board of directors.

43

Based on values disclosed in the annual report.

114

CHAPTER 7

7.6.5 Credibility power

CEO celebrity status. I measure the effect of CEO celebrity status with a dummy variable of 1, if the particular CEO appears in the top ten ranking of the best entrepreneur of the Swiss weekly business publication Handelszeitung and 0 otherwise. The ranking is determined by the members of the Club of Zurich Business journalists and should therefore provide a good proxy of how the media portray a particular CEO. Expertise. Having received a doctoral degree is generally regarded as prestigious in Switzerland and is likely to provide the CEO with credibility in his or her analytical skills. Doctoral degree is coded as a dummy variable with 1 for all CEOs who either have a doctoral degree and 0 otherwise. As I am interested in capturing credibility power arising from a doctoral degree, honorary doctorates are included.

7.6.6 Survey data

As in the case of complexity, archival data on power is augmented with more subtle information gathered through a survey with Chairmen of the board or the compensation committee.44 Discussion is a dummy variable and indicates whether the CEO has discussed issues of top management pay with the board chairman or the chairman of the compensation committee outside official board meetings. It is an indication of persuasion behavior. Meeting CEO is a count variable between 1 and 4 and indicates the frequency of meetings between the CEO and survey participants in 2004. These two variables indicate the extent to which CEOs have the opportunity to engage in persuasion and ingratiation behavior towards the chairman of the board or the compensation committee (Westphal, 1998). Good relationship. A dummy variable indicates that the relationship between the board chairman or the chairman of the compensation committee and the CEO is perceived as friendship. Friendship may be an important source of power (Pettigrew and McNulty, 1998) that has been neglected previously. Performance satisfaction. CEO performance is measured by how satisfied a board chairman or chairman of the compensation committee has been with the performance of

44

The full questionnaire is presented in the appendix.

RESEARCH METHODS

115

the current CEO since his or her appointment. Applying a subjective performance measure in regressions of executive pay may yield additional information, as objective performance measures typically cannot be used to create ideal incentives (Gibbons, 1998) and, as a consequence, subjective performance assessment may be better predictors of executive pay (Gomez-Mejia, 1994). Furthermore, as dynamic and lag effects may obscure the observed relationship between pay and performance (Boschen and Smith (1995), this variable accounts for Lamberts (1983: 443) pledge that the CEOs compensation may be based on the entire history of his employment to diversify away some of the uncertainty surrounding the CEO actions. Judging the performance since the appointment of the CEO is the appropriate time frame from a power perspective (Brickley et al., 1997). Expertise is a dummy variable and indicates if survey respondents estimate the current CEOs expertise of the firm, its industry and customers to be very high. Expertise has been suggested as an important source of power by Finkelstein (1992) and GrabkeRundell and Gomez-Mejia (2002). First proposal is a dummy variable and indicates whether the CEO, another executive manager or the compensation consulting firm makes the first proposal for the compensation contract. Agenda is a dummy variable and indicates whether the CEO sets the agenda for board meetings. Process power is a count measure between 0 and 4 and indicates the degree of influence of the CEO in the negotiation process of determining his compensation.

7.7 Other variables

7.7.1 Firm financial performance
Firm financial performance has been the determinant of executive compensation that agency theorists have been mainly interested in (Jensen and Murphy, 1990a). I measure performance as the percentage increase in the total shareholder return (share price increase plus dividend yield).

116

CHAPTER 7

7.7.2 Firm risk

Various forms of firm risk have been related to the level and mix of managerial pay (Gray and Cannella, 1997; Miller et al., 2002; Hermalin and Wallace, 2001). I measure firm risk as the ratio of debt to total assets.

7.7.3 CEO age

Studies based on human capital (Harris and Helfat, 1997; Combs and Skill, 2003) and risk-aversion theories (Eaton and Rosen, 1983) suggest that CEO age is related to the level and structure of CEO compensation. Several studies link CEO age to the level of compensation (David et al., 1998; Conyon and Murphy, 2000; Hallock, 1997). There also exists empirical evidence that older CEOs have a lower proportion of their total compensation tied in equity-based components (Zajac and Westphal, 1994) and stock option compensation (Bryan et al., 2000a; Eaton and Rosen, 1983), and that the compensation of older CEOs has a shorter time horizon (Gray and Cannella, 1997).

7.7.4 Anglo-American board members and top managers

I control for international isomorphism as an explanation for executive compensation, as pay policies that include equity participation practices pioneered in the United States are becoming common throughout the global business community (Kramarsch, 2000: 190; Richard, 2000; Fehr and Biswas, 2005; Cheffins, 2003) It may be argued that the award of equity-based pay and thus the structure and level of executive remuneration may be a reflection of influence from the Anglo-American world, which has traditionally had more reliance on stock options and stock awards and higher levels of executive compensation (Murphy, 1999; Conyon and Schwalbach, 2000; HayGroup, 2003; Kramarsch, 2000: 194; Conyon and Schwalbach, 1999; Abowd and Kaplan, 1999). For instance, Main (1999: 95) concludes that for UK firms the award of options to top executives was almost a matter of routine in many companies. While Anglo-American board members granting compensation may rely on their own experience when setting the pay of their executives (OReilly et al., 1988), American or UK executives may also demand compensation levels commensurate with what they could expect in their home country (Cheffins, 2003). I therefore control for international isomorphism by including the proportion of Anglo-Americans sitting on a firms board of directors, top management team and compensation committee.

RESEARCH METHODS

117

7.7.5 Disclosure effects

An obvious difficulty in analyzing executive compensation in nondisclosure jurisdictions is that, by definition, compensation data is generally unavailable to the researcher. However, when a switch from nondisclosure to disclosure occurs, compensation structures in the new regime may initially be very similar to those prevailing prior to disclosure. The long-term nature of many CEOs employment contracts means that there may be limited opportunity to make significant changes to compensation structures in the short run (Andjelkovic et al., 2002). On April 17, 2002, the Swiss Stock Exchange issued Guidelines for Corporate Governance, which requires stock-listed companies to disclose information on executive remuneration starting with the annual reports of 2002.45 At that time, it is likely that compensation decisions for 2002 were already set in most companies. In 2003, however, companies may have started to adapt their compensation packages to make it more aligned with shareholder wealth through outside scrutiny (Andjelkovic, 2002), the compensation practices of peer firms (Finkelstein and Hambrick, 1988; Bebchuk and Fried, 2003), or public opinion (Murphy, 1999). Ezzamel and Watson (1998) and Johnson et al. (1997) contend that executives paid below the going market rate are more likely to have their compensation increased. Conversely, those above the prevailing market rate are less likely to see their compensation reduced. Therefore, we may expect some kind of disclosure effect for 2003. I control for a potential effect of disclosure by applying a year dummy variable.46

7.7.6 Industry effects

CEO compensation practices vary significantly among various industries (e.g. Yermack, 1995; Ittner et al., 2003; Murphy, 2003; Stern and Peck, 2003), as managerial discretion differs widely across industries (Crawford et al., 1995; Hubbard and Palia, 1995), isomorphic pressure may be particularly strong within the same industry (Hambrick and Finkelstein, 1995; Porac et al., 1999), and managers are most likely to be incentivized away from companies within related business segments (Hill and Phan, 1991). I use the

The commentary to the corporate governance guidelines of the Swiss Exchange was issued on November 18, 2002. The yearly dummy variable is never significant in regression results and I therefore dropped it from my analysis. To make a more definite statement about a disclosure effect, a more detailed analysis may have to be used (see, for instance, Andjelkovic et al., 2002; Ezzamel and Watson, 1998), which is beyond the scope of this paper.
46

45

118

CHAPTER 7

industry classification of the Swiss stock exchange SWX: basic material, consumer cyclical, consumer non-cyclical, financials, healthcare, industrial, technology and utilities. I did not use a finer industry breakdown in order to maintain an adequate ratio of cases to variables.

7.7.7 Consultants
There is indication that compensation consultants influence the terms of executive pay (Baker et al., 1988; Conyon and Peck, 2004). I include the existence of compensation consultants with a dummy variable in the survey.

RESEARCH RESULTS

119

Research results

8.1 Descriptive statistics

8.1.1 Overview of the Swiss pay scene
Before turning to a detailed analysis of the influence of managerial power and complexity on executive compensation, I first examine the descriptive properties of executive compensation in Switzerland. CEO compensation I present summary statistics on CEO compensation in Table 7. On average, the CEO received total compensation of 2.0 MCHF in 2002, 2.1 MCHF in 2003 and 2.5 MCHF in 2004. The corresponding median values are 700,000 CHF, 800,000 CHF and 1.0 MCHF. This shows a clear trend of increasing CEO compensation between 2002 and 2004. Total compensation values for my sample are smaller than those usually found in US samples. For instance, Ryan and Wiggins (2001) report mean and median CEO compensation of 3.6 MUSD and 1.9 MUSD respectively in 1997. Large standard deviations indicate that there is large variability in the compensation of CEOs in Switzerland. The minimum total CEO compensation in my sample in 2002 was 180,000 CHF and the highest was 23.3 MCHF. CEOs received average (median) cash compensation of 1.3 (0.6) MCHF in 2002, 1.2 (0.7) MCHF in 2003 and 1.6 (0.9) MCHF in 2004. On average, CEOs received company stock options with a Black-Scholes value of 520,000 CHF in 2002, 440,000 CHF in 2003 and 450,000 CHF in 2004. However, these values are extremely skewed, and the median CEO does not receive any company stock options. The average CEO further received shares with a market value of 260,000 CHF, 400,000 CHF and 550,000 CHF in the years investigated. Again, the median CEO was not awarded company shares in any of these years. This suggests that equity-based pay is used for only a relatively small group of highly paid CEOs. The numbers further suggest that boards replaced optionbased compensation with direct shares awards and higher cash compensation between 2002 and 2004. The increase in share compensation thereby more than offset the decrease in option-based compensation.

120 Table 7: Descriptive statistics on CEO compensation.

Variable Year 2002 Cash compensation 2003 2004 Stock-option compensation 2002 2003 2004 2002 Shares compensation 2003 2004 2002 Total compensation 2003 2004 Equity-based compensation as percent of total compensation
a

CHAPTER 8

Na 121 129 109 121 129 109 121 129 109 121 129 109 121 129 109

Mean 1,260,044 1,227,617 1,607,924 520,298 436,159 448,758 258,061 396,260 549,032 2,037,228 2,059,523 2,507,946 13.2 13.9 13.9

Median 637,617 741,000 890,000 0 0 0 0 0 0 736,114 804,000 971,104 2.0 0 2.7

Standard deviation 1,609,458 1,525,951 1,981,428 1,974,959 2,125,366 2,104,171 1,204,300 1,757,818 2,155,043 3,724,888 4,209,201 4,207,720 21.4 21.0 21.4

2002 2003 2004

This table includes observations where the highest paid director variable is higher than the average compensation for TMT members.

The average proportion of equity-based pay out of total pay was 13.2% in 2002 and slightly increased to 13.9% in 2004. Reliance on equity-based compensation is substantially lower than in the UK, where cash compensation totals around 65% of total compensation, with option grants and Long-term Incentive Plans (LTIPs) amounting to 18% each on average (Conyon et al., 2000). The median value is almost 100% cash compensation. In the US, salary only constitutes one-third of average total CEO compensation (Khan et al., 2005). Top management compensation Table 8 shows the compensation received by the top executives of my sample companies. On average, an executive sitting on a companys top management team received a total compensation of 828,000 CHF in 2002 increasing to 842,000 CHF in 2003 and 980,000 CHF in 2004. Average cash compensation amounted to 610,000 CHF in 2002, 630,000 CHF in 2003 and 720,000 CHF in 2004. The figures further show a clear trend for companies to replace stock option compensation with direct share grants in 2003. While the average Black-Scholes value of stock options granted in 2002 was 150,000 CHF, it sharply decreased to 85,000 CHF in 2004. However, companies simultaneously increased the value of shares granted from an average of 72,000 CHF in

RESEARCH RESULTS

121

2002 to 161,000 CHF in 2003, thus more than offsetting the decrease in option-based compensation. There exists unfortunately only relatively old evidence that pay differentials between executives of Swiss and American top management teams are less substantial than those for CEOs. For instance, Ang et al., (2002) report that average (median) total compensation for a top executive of a US bank amounted to 800,000 USD (400,000 USD) in the mid 1990s. In a sample of 224 US multinationals in 1992, the average top management team pay was 750,000 USD (Carpenter and Sanders, 2004).
Table 8: Descriptive statistics on top management compensation.
Variable Cash compensation Year 2002 2003 2004 Stock-option compensation 2002 2003 2004 2002 Shares compensation 2003 2004 2002 Total compensation 2003 2004 Equity-based compensation as percent of total compensation
a

Na 199 199 184 199 199 184 199 199 184 199 199 184 199 199 184

Mean 609,386 628,371 724,043 149,598 76,238 84,644 72,143 141,730 161,169 827,806 842,108 980,435 10.0 10.5 11.3

Median 400,881 426,442 500,550 0 0 0 0 0 0 446,205 465,035 557,747 0.8 1.8 2.7

Standard deviation 730,001 696,664 852,380 561,699 270,495 239,330 325,184 777,159 681,245 1,410,887 1,463,529 1,638,058 16.2 15.6 17.1

2002 2003 2004

There are only 184 observations in 2004, as some of the companies published their annual report after the data collection was finished. In regressions of compensation on archival data, the year 2004 is excluded.

Total awards made in equity-based elements such as options and shares awards was, on average, 10.0% in 2002, 10.5% in 2003 and 11.3% in 2004. Thus, average TMT compensation is somewhat less incentivized than CEO compensation. The median value for each option and share awards individually is zero. As a proportion, the median executive received a modest 0.8% of total compensation in equity in 2002. Those executives receiving higher total compensation also tend to receive a larger part of their remuneration in equity instruments. In the US, equity-pay mix is substantially larger. For instance, Carpenter and Sanders (2002) document that the average level of TMT member equity-pay structure was 28% for a sample of S&P 500 firms in 1992.

122

CHAPTER 8

Non-executive compensation Table 9 presents summary statistics on non-executive directors compensation. The average non-executive director received total compensation of 111,000 CHF in 2002, 123,000 CHF in 2003 and 135,000 CHF in 2004. This corresponds to an increase of 12% in 2003 and 10% in 2004. Total median non-executive director compensation was 66,000 CHF in 2002, 72,000 CHF in 2003 and 83,000 in CHF. These values are broadly similar to the total fees received by non-executive directors in the US in the mid-1990s (Ryan and Wiggins, 2004; Yermack, 2004). Compared to other European countries, where non-executive board members earned, on average, 63,500 EUR in 2004, the numbers of my sample companies are substantially higher (Pfiffner, 2005).
Table 9: Descriptive statistics on non-executive director compensation.
Variable Year 2002 Cash compensation 2003 2004 Stock-option compensation 2002 2003 2004 2002 Shares compensation 2003 2004 2002 Total compensation 2003 2004 Equity-based compensation as percent of total compensation
a

Na 199 199 184 199 199 184 199 199 184 199 199 184 199 199 184

Mean 82,266 91,802 103,736 15,250 9,645 8,730 9,989 21,397 22,828 110,617 122,735 135,304 13.6 14.0 14.2

Median 56,686 63,221 71,636 0 0 0 0 0 0 65,794 71,557 82,933 0 0 0

Standard deviation 87,662 106,653 105,375 56,312 40,471 27,725 32,972 79,801 67,916 139,028 170,767 158,879 23.9 23.8 23.5

2002 2003 2004

There are only 184 observations in 2004, as some of the companies published their annual report after the data collection was finished. In regressions of compensation on archival data, the year 2004 is excluded.

Mean (median) cash compensation for non-executive directors was 82,000 CHF (57,000 CHF) in 2002. It increased to 92,000 CHF (63,000 CHF) in 2003 and 104,000 CHF (72,000 CHF) in 2004 respectively. While a majority of boards does not grant stock options to its non-executive members, the average Black-Scholes value granted in 2002 is 15,000 CHF, decreasing to 10,000 CHF in 2003 and 9,000 CHF in 2004. In a similar vein, the median board does not grant any shares to its non-executives. On average non-

RESEARCH RESULTS

123

executive directors received company shares with a value of 10,000 CHF in 2002, 21,000 CHF in 2003 and 23,000 CHF in 2004. As it seems the case with executive pay, non-executive board members replaced much of the stock option grant with stock awards in their own compensation contracts, and the additional value of shares granted more than offset the reduction in the value of stock option pay. Non-executives received approximately 14% of their total compensation in equity-based pay. Again, the median value is zero. Compared to US non-executive directors, Swiss directors receive a considerably lower proportion of compensation in the form of stock option and share awards. Yermack (2004) reports that average US non-executive directors received 50% of their pay in equity-based instruments in 1997. Overall, the results on top management and board compensation show that stock option and share plans are much less prevalent and less substantial in Swiss companies than in the US or the UK.

8.1.2 Evidence on complexity variables

Table 10 presents descriptive statistics on complexity variables. The average (median) sample firm has businesses in 3.52 (3) three-digits codes. Swiss firms have a high degree of internationalization. On average, they employ 42% of their workforce and achieve 54% of their revenues outside Switzerland. The high standard deviations reveal substantial within-sample variation. Fatemi et al. (2003) measure the percentage of foreign sales for a sample of 119 US firms in 1995. The mean value of their sample is 14.1% with a standard deviation of 13.4. Sanders and Carpenter (1998) measure a firms degree of internationalization along the three dimensions percentage of foreign sales, percentage of foreign assets and number of countries with subsidiaries as a percentage of the highest country counts of an individual firm in the sample. With a potential value of 3, their reported mean of 0.61 is comparatively low. In sum, existing studies on firm internationalization and executive compensation report considerably lower degrees of internationalization. Average (median) firm revenues are 3.1 billion CHF for both years (430,000 TCHF in 2002 and 480,000 TCHF in 2003 respectively). Sample firms had, on average, 8,000 employees in 2002 and 7,000 in 2003. The median firm has 1,300 employees. The NZZ published 80 articles about the average sample firm. The average (median) revenues standard deviation is 500 (2.0) MCHF.

124 Table 10: Descriptive statistics on complexity variables.

Variable Diversification Internationalization FETE (%) Internationalization FSTS (%) Revenues (in TCHF) Year 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 19982003 N 199 199 199 199 199 199 199 199 199 199 199 199 Mean 3.52 42.0 42.4 54.2 54.1 3,081,964 3,123,938 8,025 7,043 79.6 75 482.2 Median 3 41 40 69 69.6 429,100 479,000 1,286 1,278 18 17 2.0

CHAPTER 8

Standard deviation 1.85 36.47 36.83 40.44 40.71 9,503,184 9,552,383 25,385 22,544 151.3 140.6 18.6

Employees

Politicized environment Revenues standard deviation

8.1.3 Evidence on managerial power variables

Table 11 shows descriptive statistics for the managerial power variables. Mean and median percentage executive ownership is 1.6% and 0.2% respectively. Mean and median non-executive ownership is 1.8% and 0.2% respectively. From a power perspective, this suggests that non-executive directors tend to have slightly more power than executives. The largest outside shareholder holds on average 33% of the Swiss stock-listed company and the median largest outside shareholder owns 25% of a companys voting rights. This shows the prevalence of one major owner in Swiss corporate ownership. In comparison, most stock-listed US firms are characterized by dispersed ownership. Prowse (1992) show that in the US, the largest five shareholders together hold on average between a fifth and a quarter of the outstanding shares of the firm. For instance, Khan et al. (2005) report that the largest institutional shareholder owns, on average, 8% of equity. The median and average board in my sample has 7 members, out of which 86% are nonexecutive board members. A large majority of boards only have one executive director. This typically is the CEO in the function of the delegate director. On average, 78% of board members in 2002 and 75% in 2003 were independent. The CEO also held the chairman position in 23% and 18% of my sample companies in the years analyzed and is a member of the board of directors in about every other company. This decrease in the

RESEARCH RESULTS

125

frequency of CEO duality may be a response to the increased public and shareholder criticism to combining the roles of board chairman and CEO.
Table 11: Descriptive statistics of executive power variables.
Variable Executive ownership (%) Non-executive ownership (%) Largest shareholder Year 2002 2003 2002 2003 2002 2003 Board size 2002 2003 Non-executive directors (%) 2002 2003 Independent directors 2002 2003 CEO duality 2002 2003 CEO board member 2002 2003 Compensation committee 2002 2003 Executives on CC (%) 2002 2003 CEO tenure 2002 2003 Interdependent directors (%) 2002 2003 Interlocks Executives outside board memberships Non-executives outside board memberships Celebrity status 2002 2003 2002 2003 2002 2003 2002 2003 N 199 199 199 199 199 199 199 199 199 199 199 199 199 199 199 199 199 199 116 125 199 199 199 199 199 199 199 199 199 199 199 199 Mean 1.53 1.55 1.78 1.62 32.61 32.24 7.17 7.06 86.0 87.3 78.4 75.4 0.23 0.18 0.54 0.48 0.58 0.67 8.1 6.9 6.1 5.7 49.1 47.5 0.075 0.055 0.29 0.29 2.53 2.58 0.03 0.04 Median 0.17 0.12 0.18 0.10 24.7 22.2 7 7 87.5 87.5 80.0 80.0 0 0 1 0 1 1 0 0 4 4 50.0 50.0 0 0 0 0 1 2 0 0 Standard deviation 3.54 4.37 4.69 3.89 25.90 26.92 2.7 2.6 18.13 15.80 42.8 21.32 0.42 0.39 0.50 0.50 0.49 0.82 16.6 16.3 5.60 5.33 33.03 34.23 0.38 0.35 0.72 0.66 2.94 2.91 0.17 0.19

126

CHAPTER 8

An increasing proportion of Swiss stock-listed companies adopt compensation committees. In 2002, 58% of the companies had a compensation committee. By 2003, this number increased to 69%. Keller (2003) reports that back in 1999 out of 217 Swiss stock-listed companies 41% had a remuneration committee. These compensation committees usually entirely consist of non-executive board members: Only 8% in 2002 and 7% in 2003 of all compensation committee members were at the same time executives of the focal firm. Average and median CEO tenure is 6 and 4 years respectively. About 50% of board members were interdependent, that is they were elected to the board within the current CEOs tenure. The average top management team has 0.3 outside board memberships. The group of non-executive members of the average board of directors has 2.6 outside board memberships. There are very few interlocks in my sample. Furthermore, the data collection revealed that interlocks are highly concentrated within a few individuals and companies. Two explanations arise for the low frequency of interlocks in my sample. First, as it is at the companies discretion which of the outside board memberships of their executives to disclose the reported incidences may underestimate the true number of interlocks. However, most companies appear to fully report the activities of their nonexecutive board members. As, by definition, interlocking directors need to appear in both a top management team and another companys board of directors, the number of lost interlocks should be very low. Second, and more importantly, there is reason to believe that companies have reacted to negative publicity arising from interlocking directorates. A few companies such as UBS even prohibit their top management team members to sit on boards of other listed companies. In a study of 217 Swiss companies in 1999, Keller (2003) still finds substantial interlocking activity. US Studies investigating determinants and effects of reciprocal CEO interlocks use very old data sets. Hallock (1997) uses a cross-section of firms in 1992, Core et al. (1999) panel data from 1982 to 1984, and Fich and White (2005) a cross-section in 1991. This indicates a similar decline in interlocks in the US.

8.1.4 Other variables

Table 12 presents summary statistics of control variables in non-transformed values. Average total shareholder return (share price appreciation plus dividend yield) was -21.1% in 2002 and +39.8% in 2003. Average (median) return on sales was -2.8% (3.5%) in 2002 and -7.8% (3.0%) in 2003. These performance figures show the special

RESEARCH RESULTS

127

market environment during the investigation period. Companies had, on average, a debtassets ratio of 59%.
Table 12: Summary statistics of other variables.
Variable Total shareholder return (%) Return on sales (%) Year 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 2002 2003 N 199 199 199 199 199 199 199 199 199 199 199 199 Mean -21.1 39.8 -2.8 -7.8 58.6 58.6 4.6 7.8 4.8 7.7 52 52 Median -19.7 25.8 3.5 3.0 58.8 57.6 0 0 0 0 53 53 Standard deviation 34.4 64.9 62.0 85.9 21.7 23.1 11.7 15.7 11.9 15.6 7.9 7.8

Capital structure Anglo-American Board members (%) Anglo-American TMT members (%) CEO age

The median board and TMT of my sample do not have an Anglo-American member. On average, the board of director had 4.6% Anglo-American members in 2002 and 7.8% in 2003. Similarly, the average TMT had 4.8% Anglo-American members in 2002 and 7.7% in 2003. This is consistent with an overall increasing trend for internationalization of Swiss boards and top management teams. The CEOs in my sample average 52 years of age. The median CEO is 53 years of age.

8.2 Basic determinants of executive pay

In this section I provide results of OLS regressions on executive compensation of commonly used determinants in the literature to compare the results with previous studies.

8.2.1 The determinants of the level of executive pay

In Table 13, I show the correlation matrix of basic executive pay determinants and total TMT compensation. Ln revenues and FETE are strongly correlated with each other and with Ln TMT total compensation. Equity ownership of the largest shareholder is negatively correlated with Ln TMT total compensation.

128 Table 13: Correlation matrix of basic determinants.

Shareholder return Shareholder return Ln revenues FETE Largest shareholder Ln total TMT compensation 1 -0.025 0.045 -0.019 0.032 1 0.496 -0.075 0.715 1 -0.214 0.427 1 -0.206 Ln revenues FETE Largest shareholder

CHAPTER 8

Ln total exec compensation

Table 14 shows OLS estimation results of determinants commonly specified variables in earlier studies on total TMT and CEO compensation. Consistent with prior literature (Tosi et al., 2000), the main determinant of compensation levels is firm size, which I measured as the logarithm of revenues. Although it has a positive sign in the pooled regression (model 1), current-year total shareholder return is not a significant predictor of TMT compensation levels. The corresponding p-value is 0.21. When regressions are run for the years 2002 and 2003 separately, shareholder return is positively related to total TMT compensation in 2003 (model 3) at the 5% significance level and negatively related, but not significant, in 2002 (model 2). As average stock market returns were positive in 2003 and negative in 2002, this suggests that managers are rewarded for increased shareholder returns but not punished for weak performance. This result is consistent with prior US research. Hambrick and Finkelstein (1995) find that CEOs are rewarded for profit increases but not penalized for profit declines in firms with dispersed ownership. This result is also consistent with a disclosure effect (Andjelkovic et al., 2002). Public compensation disclosure after pay was set in 2002 resulted in smaller agency problems and more efficient compensation policies. Total shareholder return is not significantly related with CEO compensation. However, it shows the same pattern as for TMT compensation. Shareholder return has a negative sign for the 2002 regression, but a positive sign for the 2003 regression. Firm internationalization, measured as the proportion of foreign employees out of total employees, is positively related to total TMT compensation, but not significant. Internationalization is significantly associated with total CEO pay in the full regression and the 2002 regression, consistent with prior studies (e.g. Sanders and Carpenter, 1998). Shareholder concentration, measured as the ownership stake of the largest outside

RESEARCH RESULTS

129

shareholder, is negatively associated with TMT and CEO compensation levels. This supports prior studies (e.g. Werner and Tosi, 1995; Khan et al., 2005).
Table 14: OLS estimation of common determinants on total executive compensation.
Dependent variable Independent variables Intercept Ln Revenues Total shareholder return FETE Largest shareholder CEO duality CEO dummy Adjusted R2 F-statistics Observations Year 0.53 115.35*** (0.00) 398 2002/2003 0.51 50.22*** (0.00) 199 2002 0.58 68.13*** (0.00) 199 2003 Ln Total TMT Compensation (1) 11.39*** (106.37) 0.297*** (17.14) 0.001 (1.26) 0.001 (1.47) -0.004*** (-4.01) (2) 11.36*** (69.42) 0.258*** (11.33) -0.002 (-1.30) 0.001 (0.81) -0.004** (-2.16) (3) 11.30*** (75.49) 0.31*** (13.21) 0.001** (2.07) 0.001 (0.69) -0.004*** (-2.85) Ln Total CEO Compensationa (4) 11.21*** (86.04) 0.328*** (16.30) 0.000 (0.89) 0.003** (2.40) -0.005*** (-4.30) 0.341*** (4.04) 0.404*** (5.70) 0.59 96.74*** (0.00) 398 2002/2003 (5) 11.27*** (57.90) 0.316*** (10.65) -0.001 (-0.43) 0.003* (1.87) -0.005*** (-2.68) 0.302** (2.47) 0.333*** (3.15) 0.56 42.96*** (0.00) 199 2002 (6) 11.10*** (59.24) 0.344*** (12.33) 0.001 (0.75) 0.002 (1.26) -0.005*** (-3.17) 0.405*** (3.38) 0.468*** (4.82) 0.62 53.86*** (0.00) 199 2003

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses. a In cases where the compensation of the highest-paid board member was lower than the average TMT compensation, I used the latter number. The CEO dummy variable controls for this effect.

In regressions on CEO compensation (model 4-6) I also include CEO duality as a firmly established determinant of compensation (e.g. Conyon, 1997). The coefficient of CEO duality measures the relative difference in total compensation for CEOs who are also the chairman of the board and CEOs who do not combine these two positions. The coefficient of 0.341 for CEO duality means that there is a wage differential of approximately 40% (e0.341-1). Overall, the results confirm the common determinants of CEO compensation levels established in the executive pay literature: shareholder concentration, firm size, internationalization, and CEO duality. They are also in line with the weak and sometimes non-existent relationship found between current performance and CEO pay. The adjusted R-squared indicates that these basic determinants help explain more than half of the variation in executive pay.

130

CHAPTER 8

8.2.2 The determinants of equity-compensation mix

Table 15 shows Tobit estimates of determinants established by prior studies on the proportion of equity-based compensation. Larger firms tend to award a larger proportion of equity-based compensation to the TMT and the CEOs. This is consistent with prior literature (e.g. Zajac and Westphal, 1994). Firm internationalization is a significant positive determinant of CEO equity-compensation mix. The relationship is also positive, but not statistically significant for TMT members. This relationship was first found by Sanders and Carpenter (1998).
Table 15: Tobit model estimates of common determinants on equity-pay mix.
Dependent variable TMT equity compensation% Independent variables Intercept Ln Revenues FETE Largest shareholder CEO age Executive ownership Capital structure LR chi2 Pseudo R2 Observations -0.404 (-1.10) -0.001 (-1.07) 92.55*** (0.00) 0.31 398 -0.001 (-1.49) 117.6*** (0.00) 0.30 398 (1) -0.203*** (-3.70) 0.049*** (5.83) 0.001 (1.46) -0.002*** (-4.38) CEO equity compensation %a (2) 0.071 (0.63) 0.073*** (7.04) 0.001* (1.68) -0.002*** (-3.56) -0.008*** (-4.22)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses. a In cases where the compensation of the highest-paid board member was lower than the average TMT compensation, I used the latter number. To test whether this procedure affects the results, I use a dummy variable for cases where I replace the highest board compensation with average TMT compensation. The dummy variable is statistically insignificant and does not alter the results.

Firms with more concentrated ownership tend to have lower reliance on equity-based compensation for both the average TMT member and the CEO. This is also in line with prior studies (e.g. Mehran, 1995). CEO age is negatively related to CEO equity compensation mix, in line with Zajac and Westphals (1994) results. This is consistent

RESEARCH RESULTS

131

with the idea that an older CEO might have different preferences for contingent and noncontingent forms of pay than younger executives. Executive ownership is negatively, but not significantly, related with TMT equity-pay mix. This result is not consistent with Mehran (1995), who finds a significant negative association between these two variables. Capital structure, measured as the debt-assets ratio, is not significantly related to compensation structure. This is inconsistent with prior empirical research and the idea of efficient risk-sharing between shareholders and management (Zajac and Westphal, 1994). In short, these results support prior empirical findings from the US that firm size, shareholder concentration, CEO age and to a lesser degree internationalization are significant determinants of executive compensation structure. Further, the results do not support the hypothesis that capital structure and executive ownership are determinants of executive pay mix.

8.3 The influence of complexity on executive pay

8.3.1 The influence of individual complexity variables on pay levels
Table 16 depicts the correlation matrix based on the individual complexity variables and TMT total compensation. As the last row shows, the complexity variables have high correlations with TMT compensation levels. Only the diversification variables business lines and ln 3 digits codes have relatively low correlations with compensation. Furthermore, the complexity variables exhibit large correlation coefficients among themselves.
Table 16: Correlation matrix based on individual complexity variables.
Ln revenues Ln revenues FETE Ln news Business lines Ln 3 digits codes Ln std revenues Ln TMT total compensation 1 0.496 0.680 0.329 0.422 0.907 0.716 1 0.404 0.108 0.358 0.496 0.427 1 0.307 0.318 0.656 0.724 1 0.309 0.288 0.219 1 0.366 0.228 1 0.691 1 FETE Ln news Business lines Ln 3 digits codes Ln rev stdev Ln TMT total

132

CHAPTER 8

Table 17 shows OLS estimations of individual complexity variables on TMT compensation. As the high adjusted R-squared shows, these variables explain a large proportion of executive compensation. The most important complexity variable is firm size. Top executives earn higher total compensation when they are employed by larger organizations. For instance, 10% higher revenues are associated with 1.9% higher total TMT compensation in model 1. Similarly, 10% higher revenues are associated with 1.7% higher CEO cash compensation (model 4) and 2.7% higher total CEO compensation (model 5). The association between firm size and pay is stronger and larger for total compensation levels than cash compensation both for TMT and CEO pay. These elasticities are somewhat lower compared to evidence from the US. Belliveau et al. (1996) report a cross-sectional elasticity of CEO cash compensation to sales of 0.20 and Henderson and Fredrickson (1996) find an elasticity of CEO total compensation to sales of 0.31. This indicates that in the US, 10% higher firm sales are associated with 2.0% (3.1%) higher CEO cash (total) compensation.47 Executives also appear to get compensated for the complexity reflected in a firms degree of internationalization. The proportion of foreign employees to total employees is significantly related to compensation in all models. There is indication that the hypothesized association between internationalization and pay is more robust and larger for total compensation than for cash compensation. These findings corroborate earlier studies (Sanders and Carpenter, 1998; Fatemi et al., 2003).48 There is also strong support that a politicized environment explains variations in executive compensation. The number of newspaper articles is significantly positively associated with all measures of compensation applied in the analysis. This has not been tested earlier. Based on individual variables regressions, the other two dimensions of complexity examined in this work, firm diversification and market uncertainty, do not appear to help explain variations in executive pay. For firm diversification, neither the number of three digits SIC codes the firm operates in (model 1) nor the number of different business lines (model 2-5) is significantly associated with compensation levels. These results are not

The difference in firm size pay elasticities may also be due to the inclusion of other complexity values. In a regression not reported here, I find similar elasticities to those found in studies on US data after excluding the other complexity variables. Interestingly, this association between FETE and pay is stronger, after including other complexity variables (see also Table 14)
48

47

RESEARCH RESULTS

133

consistent with prior studies that find a link between diversification and compensation (e.g. Rose and Shepard; 1997). Market uncertainty, measured as the logarithm of the revenues standard deviation, is not a significant determinant of executive pay. Overall, results using individual complexity variables lend support to hypothesis 1a that top executives receive more compensation the higher the complexity of their job.
Table 17: OLS estimation of individual complexity variables on compensation levels.
Dependent variable Independent variables Intercept Ln Revenues FETE Ln News Business Lines Ln 3 digits codes Ln StDev Revenues Industry dummies CEO dummy Adjusted R2 F-statistics Observations 0.69 74.50*** (0.00) 398 0.69 74.28*** (0.00) 398 0.64 59.55*** (0.00) 398 -0.041 (-0.90) 0.008 (0.25) Yes 0.008 (0.27) Yes 0.009 (0.31) Yes 0.035 (1.01) Yes 0.532*** (10.36) 0.68 67.09*** (0.00) 398 -0.037 (-1.00) Yes 0.539*** (9.10) 0.70 73.41*** (0.00) 398 (1) Ln TMT tot comp 11.16*** (86.96) 0.185*** (5.30) 0.004*** (4.71) 0.169*** (7.64) (2) Ln TMT tot comp 11.15*** (87.46) 0.180*** (5.20) 0.004*** (4.66) 0.169*** (7.52) -0.006 (-0.11) (3) Ln TMT cash comp 11.37*** (97.79) 0.152*** (4.38) 0.003*** (4.19) 0.136*** (6.63) -0.018 (-0.38) (4) Ln CEO cash comp 11.62*** (103.20) 0.137*** (3.63) 0.003*** (3.55) 0.120*** (8.06) -0.088 (-1.42) (5) Ln CEO tot comp 10.71*** (67.09) 0.273*** (6.51) 0.005*** (4.39) 0.167*** (6.16) -0.019 (-0.30)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses. aIn cases where the compensation of the highestpaid board member was lower than the average TMT compensation, I used the latter number. The CEO dummy variable controls for this effect.

All regressions control for industry effects. To save space, individual results for industries are not reported. Top executives in the financial and healthcare sector earn a pay premium of 55% and 40% respectively of their counterparts heading firms in the basic material sector. Executives in the technology sector receive a pay premium of 20%. Executives in the consumer goods industries, in the industrial sector and in utilities do not have compensation levels that are significantly different from those in basic material.

134

CHAPTER 8

The results further document a significant pay premium for the CEO job. Models (4) and (5) investigate compensation levels of the highest-paid executive director. In cases where the compensation of the highest-paid board member was lower than the average TMT compensation, I used the latter number. The CEO dummy variable controls for this CEO effect and indicates that CEOs get 71% (e0.539-1) higher total compensation than average top management team compensation.49

8.3.2 The influence of individual complexity variables on pay mix

Table 18 shows Tobit estimates of individual complexity variables on compensation mix. Firm size measured as the logarithm of revenues is not significantly related to the proportion of TMT or CEO equity-based compensation. This does not confirm the empirical relationship between firm size and equity-pay found in the US (e.g. Ang et al., 2002; Barron and Waddell, 2003b). Firm internationalization is significantly associated with a higher proportion of equitybased TMT and CEO compensation, corroborating the results of Sanders and Carpenter (1998). The number of newspaper articles published about a particular firm is also positively related to equity-based compensation mix. The relationship is significant at the 1%-level. This provides first evidence that compensation committees take into account the complexity arising from a politicized environment when structuring executive pay. Firm diversification and market uncertainty do not appear to be significant determinants of TMT and CEO pay structure. Diversification, measured alternatively as the logarithm of the number of 3-digits codes the firm operates in and business lines, is unrelated to pay structure. Although market uncertainty, measured as the logarithm of the revenues standard deviation, is positively related to equity-based compensation mix, the association is not significant at commonly specified levels. Market uncertainty is not robust to the inclusion of industry dummies. These two dimensions of complexity were not significant in pay level regressions of Table 17 either. There is no indication that the determinants of the structure of CEO compensation are significantly different from those of average TMT compensation structure. Overall, there is some support for hypothesis

49

Average TMT compensation includes the compensation paid to the CEO.

RESEARCH RESULTS

135

1b that greater firm complexity increases the proportion of equity-based executive compensation.
Table 18: Tobit estimates of individual complexity variables on compensation mix.
Dependent variable Independent variables Constant Ln Revenues FETE Ln News Business Lines Diversification 3 digits Ln StDev Revenues CEO dummy Industry dummies Pseudo R2 LR chi2 Observations Yes 0.55 165.26*** (0.00) 398 Yes 0.55 164.66*** (0.00) 398 0.024 (1.50) (1) TMT equity prop -0.299*** (-4.78) 0.006 (0.36) 0.001** (2.09) 0.041*** (3.66) -0.026 (-1.03) -0.016 (-0.68) 0.024 (1.52) -0.043 (-1.48) 0.025 (1.17) 0.005 (0.16) Yes 0.41 157.81*** (0.00) 398 0.024 (1.13) 0.005 (0.15) Yes 0.40 156.97*** (0.00) 398 (2) TMT equity prop -0.290*** (-4.61) 0.006 (0.35) 0.001** (2.20) 0.040*** (3.57) (3) CEO equity propa -0.451*** (-5.17) 0.026 (1.15) 0.001** (2.51) 0.039*** (2.72) (4) CEO equity propa -0.470*** (-5.41) 0.024 (1.07) 0.001** (2.40) 0.040*** (2.78) -0.038 (-1.16)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). a In cases where the compensation of the highest-paid board member was lower than the average TMT compensation, I used the latter number. The CEO dummy variable controls for this effect.

8.3.3 Survey measures of complexity and pay levels and mix

Table 19 presents summary statistics on the surveyed complexity measures. Perceived complexity has a mean (median) value of 0.87 (one). This means that in 87% of cases respondents perceived the required technical, management and human relations knowhow for their CEOs position to be high or very high. The strategic thinking environment variable has a mean (median) value of 0.38 (zero). In other words, in 38% of cases board or compensation committee chairmen perceived the CEOs strategic thinking environment to be very difficult. The span of control variable has a mean (median) value of 2.9 (3). A three corresponds to 6-9 direct reports. The corresponding standard deviation is very low.

136 Table 19: Descriptive statistics on survey complexity measures.

Variable Perceived complexity Strategic thinking environment Span of control Year 2004 2004 2004 N 47 47 47 Mean 0.87 0.38 2.91 Median 1 0 3

CHAPTER 8

Standard deviation 0.34 0.49 0.69

Table 20 shows univariate estimates of CEO compensation on surveyed complexity variables. All three variables have the hypothesized positive signs. Moreover, perceived complexity is a significant predictor of both CEO total compensation (model 1) and CEO cash compensation (model 2). The number of direct reports is significantly associated with CEO cash compensation, but not with CEO total compensation. Strategic thinking environment is significantly associated with neither total nor cash compensation. Only perceived complexity explains a relevant proportion of pay with Rsquared of 0.09 and 0.08 respectively. Overall, the results lend some support to hypothesis 1a, which states a positive relationship between complexity and compensation levels.
Table 20: Univariate estimates of surveyed complexity on CEO compensation.
Dependent variable Independent variables Intercept Perceived complexity Strategic thinking environment Direct reports Adjusted R2 F-statistics Observations 0.09 4.41** (0.04) 47 0.08 5.20** (0.03) 47 0.01 1.65 (0.205) 47 0.004 1.19 (0.28) 47 (1) Ln CEO tot comp 13.05*** (38.04) 0.771** (2.10) (2) Ln CEO cash comp 12.89*** (41.78) 0.753** (2.28) 0.333 (1.29) 0.258 (1.09) 0.194 (1.037) 0.002 1.07 (0.31) 47 0.166* (1.81) 0.05 3.26* (0.08) 47 (3) Ln CEO tot comp 13.59*** (84.65) (4) Ln CEO cash comp 13.44*** (91.93) (5) Ln CEO tot comp 13.155 (23.51)*** (6) Ln CEO cash comp 12.67*** (25.57)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses. In cases where the compensation of the highest-paid board member was lower than the average TMT compensation, I used the latter number.

Table 21 shows the result of individual Tobit regressions of survey complexity variables on the proportion of CEO equity-based compensation. Perceived complexity and

RESEARCH RESULTS

137

strategic thinking environment are not related to the extent to which CEOs are paid with options and shares. The number of direct reports is significantly associated with CEO equity compensation, but the association has the reverse sign: All other things equal, this result suggests that CEOs with more direct reports have less of their compensation tied to shareholder wealth. These results do not support the perspective that increased complexity is associated with monitoring difficulty and will therefore lead the board of directors to rely more heavily on shares and options to align the interests of managers and shareholders. Results based on survey complexity measures therefore do not lend support to hypothesis 1b that greater complexity is associated with a higher reliance on equity-based executive compensation.
Table 21: Univariate Tobit estimates of CEO equity-based compensation on survey complexity variables.
Dependent variable Independent variables Intercept Perceived complexity Strategic thinking environment Direct reports Pseudo R2 LR chi2 Observations 0.003 0.1 (0.75) 47 0.03 0.93 (0.34) 47 CEO equity proportion (1) 0.112 (1.02) -0.037 (-0.31) 0.079 (0.96) -0.086* (-1.51) 0.06 2.25 (0.13) 47 (2) 0.050 (0.93) (3) 0.335** (1.96)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed). t-values in parentheses.

8.3.4 The complexity construct and executive pay

Multicollinearity may be an issue in regressions on individual complexity variables. This problem can be alleviated by applying a principal component analysis of the complexity variables to define an overall complexity construct (Greene, 2000: 258). The correlation matrix in Table 22 confirms that the archival complexity variables are highly correlated with each other. Interesting to note that firm size is significantly correlated with all remaining measures of firm complexity, especially market uncertainty. This may make it difficult to single out the contribution of individual complexity factors and may provide an explanation for why the measures of firm diversification and market uncertainty were

138

CHAPTER 8

insignificant predictors of pay level and structure in the above regressions. Nevertheless they may provide important information about what facets of complexity influence executive pay.
Table 22: Correlation matrix of archival complexity variables.
Ln Revenues Ln revenues FETE Ln news Ln revenues Std Dev Ln 3 digits codes 1 0.496 0.679 0.907 0.420 1 0.404 0.496 0.358 1 0.660 0.318 1 0.366 1 FETE Ln News Ln Revenues Std Dev Ln Num 3 digits

Table 23 shows factor-analyzed results of archival complexity variables based on principal components extraction. Using Eigenvalues greater than 1 as a cut-off value, one principal component is retained. This suggests that the variables pick up different facets of the same complexity construct, which I term objective complexity. All complexity variables load relatively high on the principal component with revenues and Ln News exhibiting the highest scores. The high loading of revenues volatility, indicates the importance of market uncertainty, inconsistent with its insignificant effect in regressions using individual variables. Overall, Table 23 lends strong support to the complexity construct developed in chapter 5. I also estimate factor scores, which I use as inputs in subsequent analysis. Scoring coefficients are shown based on promax rotated factors. Firm size has the highest coefficient, diversification the lowest.
Table 23: Principal component analysis and scoring coefficients of archival complexity construct.
Factor loadings Variable Ln Revenues Ln Rev Stdev FETE Ln News Ln Num 3 digits 1 0.92305 0.68113 0.79863 0.90680 0.57806 Uniqueness 0.1480 0.5360 0.3622 0.1777 0.6658 Scoring coefficients 0.2968 0.2916 0.2190 0.2568 0.1859

RESEARCH RESULTS

139

Factor analyzed results of the survey complexity variables are shown in Table 24. One principal component was retained. As with objective complexity, all variables load highly on the common complexity construct. Perceived complexity scores highest, strategic thinking lowest. I term this construct subjective complexity.
Table 24: Principal component analysis and scoring coefficients of survey complexity construct.
Factor loadings Variable Perceived complexity Strategic thinking Direct reports 1 0.7477 0.5690 0.6770 Uniqueness 0.4409 0.6763 0.5416 Scoring coefficients 0.5575 0.4242 0.5048

In Table 25, the relationship between complexity and CEO compensation is synthesized. Both the archival-based objective as well as the survey-based subjective complexity constructs show a significant association with CEO compensation levels. Archival complexity explains more than 50% of the variation in CEO compensation. Archival complexity, but not survey complexity, also shows a significant association with CEO equity-based compensation mix. These results lend strong support to hypothesis 1a that greater complexity is associated with greater levels of compensation. And they offer some support for hypothesis 2 that greater complexity is associated with greater reliance on equity-based compensation.

140

CHAPTER 8

Table 25: Relationship between complexity constructs and CEO compensation.

Dependent variable Independent variables Intercept Subjective complexity Objective complexity Pseudo R2 LR chi2 Adjusted R2 F-statistics Observations 0.12 7.55*** (0.00) 47 47 0.01 0.26 (0.61) 0.52 438.5*** (0.00) 398 398 (1) Ln CEO cash compa 13.54*** (125.74) 0.299*** (2.75) (2) CEO equity comp propa 0.080* (1.91) -0.020 (-0.51) 0.722*** (20.94) 0.158*** (9.55) 0.24 92.32 (0.00) (3) CEO total compa 13.51*** (392.16) (4) CEO equity comp propa -0.020 (-1.12)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses. a In cases where the compensation of the highest-paid board member was lower than the average TMT compensation, I used the latter number.

8.4 The influence of power on executive pay

8.4.1 The influence of individual power variables on pay levels
Table 26 shows the influence of individual power variables on total TMT compensation. Most of the variables carry the hypothesized sign and are statistically significant. Power arising from stock ownership has some effects on executive compensation. Large outside shareholders are associated with significantly lower TMT and CEO compensation levels. This supports prior empirical evidence (e.g. Core et al., 1999; Ryan and Wiggins, 2001). Large executive ownership is positively correlated with TMT compensation, but the result is not significant. Non-executive ownership does not affect compensation. Structural power also matters as a determinant of executive pay. When CEOs also hold the position of the Board chairman, total TMT compensation is higher by 18% (e0.168-1) in model 1 and CEO compensation is higher by 35% (e0.300-1) in model 4. That is, CEOs receive a 35% pay-premium when they assume the position of the chairman of the board in addition. This is larger than the pay premium found in other studies. Core et al. (1999) report a 16% pay premium, Sridharan (1996) 18% for cash compensation. Inconsistent with prior studies (e.g. Gosh and Sirmans, 2005), board size is not significantly related to compensation levels, although the association is positive. Board structure, measured as

RESEARCH RESULTS

141

the proportion of non-executive directors on the board, is not significantly linked to executive compensation. Part of the low explanatory power of board structure may be due to data specifics as outlined in the section on descriptive statistics. In the majority of Swiss boards, the CEO is the only executive director. For instance in 2002, the ratio of non-executive to total directors had a standard deviation 0.18 only. Thus, after controlling for CEO duality and other structural variables, it is not surprising that nonexecutive directors do not affect compensation. In an additional analysis where I regress only firm size, non-executive directors and the CEO dummy on CEO compensation, I find that the proportion of non-executive directors is significantly and negatively associated with total CEO compensation at the 10%-level. The corresponding t-value is of the correlation coefficient is -1.82. The proportion of independent directors is negatively associated with compensation levels. Results are significant for both TMT compensation and CEO compensation levels. It suggests that independent directors non-executive directors with no material business relationship with the focal firm and who have not had any executive position over the past three years constitute a powerful counterpart to CEOs. This supports calls for more board independence. Contrary to what has been hypothesized, but consistent with empirical UK results (Conyon and Peck, 1998), the presence of a compensation committee is positively associated with executive compensation. Not only is this variable statistically highly significant, it is also economically significant. The corresponding regression coefficient suggests that boards with a compensation committee grant their executives more than 30% higher compensation than boards without such a committee, also after controlling for board size. This result is not consistent with the view of Zahra and Pearce (1989) that powerful boards split their tasks among board committees. There is little support that power arising from tenure plays a significant role in shaping top executive compensation. Neither the proportion of interdependent directors nor CEO tenure are associated with compensation levels. Similarly, I find little evidence that network power affects compensation. Neither the number of outside board memberships of executives nor the number of interlocks increases TMT compensation.50

I also include the number of outside board memberships for non-executive directors in regressions not reported in this work. The relationship is not significant and the inclusion of this variable does not affect the overall results.

50

142

CHAPTER 8 Table 26: OLS estimation of individual power variables on compensation level.
Dependent variables Ln total TMT compensation Ln total CEO compensation (3) 11.46*** (100.2) -0.004*** (-4.40) (4) 11.31*** (72..69) 0.005*** (-4.38) (5) 11.38*** (81.02) -0.005*** (-4.34)

Independent variables Intercept Largest shareholder (%) Executive ownership Non-executive ownership CEO duality Board size Non-executives directors Independent directors Compensation committee Interdependent directors Ln CEO tenure Executives on boards Interlocks Celebrity status Doctoral degree Log revenues CEO dummy Adjusted R2 F-statistics Observations

(1) 11.23*** (59.06) -0.004*** (-3.71) 1.117 (1.48) 0.076 (0.12) 0.168** (2.21) 0.008 (0.71) 0.040 (0.21)

(2) 11.36*** (86.16) -0.004*** (-3.85) 0.794 (1.09) 0.005 (0.01) 0.129* (1.79) 0.011 (0.98)

0.147** (2.16)

0.300*** (3.45) 0.019 (1.44)

0.289*** (3.52)

-0.204*** (-2.41) 0.283*** (4.73) -0.001 (-0.01) 0.286*** (4.80) 0.035 (0.41)

-0.202** (-2.47) 0.303*** (5.19)

-0.239** (-2.45) 0.367*** (4.64)

-0.227** (-2.36) 0.330*** (4.90)

-0.048 (-0.13) 0.047 (1.23) 0.045 (1.17) -0.031 (-0.62) 0.677*** (3.72) 0.047 (0.64) 0.277*** (16.48) 0.284*** (16.74) 0.286*** (18.26) 0.225*** (15.359) 0.434*** (6.29) 0.56 52.36*** (0.00) 398 0.57 48.77*** (0.00) 398 0.57 106.9*** (0.00) 398 0.62 66.97*** (0.00) 398 0.309*** (16.52) 0.423*** (6.25) 0.62 95.45*** (0.00) 398 0.695*** (3.84)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses. a In cases where the compensation of the highest-paid board member was lower than the average TMT compensation, I used the latter number. I use a dummy variable to control for this effect (CEO dummy).

Credibility power appears to influence CEO compensation. Successful CEOs, who manage to build a celebrity status around their person, get substantially higher pay. The

RESEARCH RESULTS

143

coefficient suggests that, all other things equal, CEOs of a particularly high status earn as much as 95% (e0.668-1) more than other CEOs. A doctoral title does not confer enough credibility power to significantly affect pay. Overall, results in Table 26 lend some support to hypothesis 2a that greater CEO and top management team power leads to greater levels of executive compensation. Shareholder concentration, board independence, CEO duality, compensation committee existence and CEO celebrity status all influence the level of executive compensation.

8.4.2 The influence of individual power variables on pay mix

In Table 27, I show Tobit estimates of individual power variables on executive compensation mix. There is strong evidence that ownership affects compensation mix, but not fully consistent with a power hypothesis. Higher ownership concentration is negatively associated with equity compensation mix, inconsistent with a power hypothesis, but consistent with the result found in Mehran (1995). Large shareholders do not use their influence to subject executive compensation to equity-based instruments. TMT ownership is negatively associated with TMT and CEO equity compensation mix. This is in line with a power hypothesis that more powerful executive teams are able to obtain a higher proportion of fixed pay and supports other empirical evidence (Zajac and Westphal, 1994; Sanders and Carpenter, 1998; Bryan et al., 2000a; Khan et al., 2005). Non-executive ownership is positively associated with equity compensation mix, also supporting a power hypothesis. When non-executive directors get paid in stock and options, they also tie executive pay more closely to shareholder wealth. There is also evidence that power arising from the structure of the board is related to executive pay mix. Again, results are not fully consistent with the stated hypothesis. Firms where the CEO is also the chairman of the board tend to rely more heavily on equity-based TMT compensation. Larger boards are associated with higher reliance on TMT equity compensation mix, but not CEO equity compensation mix. The presence of a compensation committee is related to higher reliance on equity-based TMT compensation. The results of all these three variables are inconsistent with my hypothesis. The other structural variables non-executives ratio and independent directors ratio are not related to equity compensation mix. Overall, board structure affects compensation mix, but not in the way argued by the standard power perspective.

144

CHAPTER 8 Table 27: Tobit estimates of individual power variables on executive compensation mix.
Dependent variable TMT equity proportion CEO equity proportiona (3) -0.160** (-2.32) -0.003*** (-4.75) -1.550*** (-2.69) 0.622* (1.75) 0.120*** (2.90) 0.030 (0.69) 0.032 (0.34) 0.014*** (2.78) 0.162*** (5.63) 0.023 (0.58) 0.014*** (2.68) 0.164*** (5.69) 0.027 (0.69) -0.003 (-0.17) 0.027 (1.21) 0.033** (1.92) 0.007* (1.66) 0.026 (1.19) 0.033** (1.92) 0.007 (1.60) 0.053** (2.04) 0.043** (2.09) 0.008 (1.47) 0.330*** (4.20) -0.059 (-1.57) 113.7*** (0.00) 0.38 398 113.3*** (0.00) 0.38 398 128.43*** (0.00) 0.33 398 -0.001 (-0.07) 0.052** (2.01) 0.043** (2.08) 0.008 (1.42) 0.332*** (4.22) -0.059 (-1.58) 127.99*** (0.00) 0.33 398 0.007 (1.12) 0.200*** (5.61) 0.025 (0.21) 0.007 (1.09) 0.202*** (5.67) (4) -1.590 (-1.42) -0.003*** (-4.80) -1.574*** (-2.67) 0.621* (1.75) 0.118*** (2.78)

Independent variables Intercept Largest shareholder (%) Executive ownership Non-executive ownership CEO duality Independent directors ratio Non-executives ratio Board size Compensation committee Interdependent directors Ln CEO tenure Interlocks Executives on boards Non-executives on boards Celebrity status Doctoral degree LR chi2 Pseudo R2 Observations

(1) -0.190*** (-3.41) -0.003*** (-5.38) -0.817* (-1.94) 0.650** (2.23) 0.111*** (3.39) 0.027 (0.74)

(2) -1.98** (-2.21) -0.003*** (-5.42) -0.826** (-1.91) 0.649** (2.22) 0.111*** (3.26)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses. aIn cases where the compensation of the highestpaid board member was lower than the average TMT compensation, I used the latter number. To test whether this procedure affects the results, I use a dummy variable for cases where I replace the highest board compensation with average TMT compensation. The dummy variable is statistically insignificant and does not alter the results.

Tenure power, measured as the ratio of interdependent directors to total directors and CEO tenure, is not related with compensation mix. Network power is a significant

RESEARCH RESULTS

145

predictor of compensation mix. However, the results are contrary to hypothesis 2b that greater executive power decreases the proportion of equity in total compensation. The number of interlocks is significantly positively related with the proportion of equitybased CEO compensation, contrary to what Fich and White (2005) find. The number of executives outside board memberships increases the proportion of equity-based CEO and TMT compensation. The number of non-executive directors outside board memberships is positively related to compensation mix, but the coefficient is significant only in model (1). Table 27 further shows that credibility power influences compensation mix. CEO celebrity significantly and positively affects the proportion of shares and options in the compensation contract. Again, this result is contrary to hypothesis 2b. Doctoral education, though, is not related to compensation mix. In short, hypothesis 2b that greater executive power leads to less reliance on equitybased compensation is not confirmed by regressions on individual power variables. Although some of the variables are significantly associated with equity-based pay mix, only executive ownership and non-executive ownership are consistent with a power hypothesis. The results are not consistent with the notion that executives prefer cash compensation over equity instruments and use their power to make their compensation package less risky (e.g. Grabke-Rundell and Gomez-Mejia, 2002).

8.4.3 Survey measures of power

In Table 28, I show descriptive statistics on survey measures of power. 42% of respondents discussed issues of executive compensation with their CEO outside official board meetings. 81% of survey participants met the CEO four or more times outside official board meetings in 2004. 15% responded that a good relationship with the CEO was a major criterion for the selection of new board members and 11% called their relationship with the CEO as friendship. 60% are very satisfied with the past performance of the CEO, 62% estimate the CEOs expertise of the industry and the firm to be very high. In 30% of responding companies, the CEO sets the agenda for board meetings. And 43% of responding firms use compensation consultants to determine executive pay.

146

CHAPTER 8

Table 28: Descriptive statistics on survey power measures.

Variable Discussion Meeting CEO Good relationship Friendship Performance satisfaction Expertise First proposal Agenda Process power Consultants Year 2004 2004 2004 2004 2004 2004 2004 2004 2004 2004 N 47 47 47 47 47 47 47 47 47 47 Mean 0.74 0.81 0.15 0.11 0.60 0.62 0.19 0.30 0.15 0.43 Median 1 1 0 0 1 1 0 0 0 0 Standard deviation 0.44 0.40 0.36 0.31 0.50 0.49 0.40 0.46 0.36 0.50

Table 29 presents univariate estimates of survey power variables on CEO compensation. There is a positive association between Chairmen and CEOs discussing issues of executive compensation outside board meetings and CEO compensation. However, the result is not statistically significant (model 1). There is a positive association between CEO influence during the pay negotiation process and CEO compensation level (model 9). This association is significant at the 10%-level. As indicated in the appendix, high scores for this variable are given when the CEO is able to negotiate his compensation exclusively with the board chairman or the compensation committee. Lower scores are given when the CEO cannot negotiate his own pay at all or only after the board has discussed the compensation committees suggestion in the absence of the CEO. None of the other power variables shows any significant association with CEO total compensation. Regressions of the different power variables on CEO cash compensation yield similar results and are therefore not reported here. The use of compensation consultants, not a direct power variable, is the other significant determinant of CEO compensation. Companies have a sort of institutional pride, when consultants compare pay across peer companies, and consciously paying a CEO below the average constitutes a blow to that institutional pride (Crystal, 1991: 221). Thus, the hiring of compensation consultants will in most cases increase the CEOs compensation.

RESEARCH RESULTS

147

Table 29: Univariate estimates of survey power variables on CEO compensation.

Dependent variable Independent variables Discussion Meeting CEO Good relationship Friendship Performance satisfaction Expertise First proposal Agenda Process power Consultants Adjusted R2 F-statistics Observations 0.01 1.67 (0.20) 47 -0.01 0.38 (0.54) 47 -0.01 0.99 (0.33) 47 -0.02 0.22 (0.64) 47 -0.02 0.01 (0.94) 47 0.01 1.40 (0.24) 47 -0.01 0.53 (0.47) 47 0.02 1.00 (0.32) 47 0.04 3.09* (0.09) 47 Ln total CEO compensation (1) 0.38 (1.29) -0.20 (-0.6) 0.36 (1.00) 0.196 (0.47) -0.02 (-0.1) -0.31 (-1.2) 0.214 (0.73) -0.28 (-1.0) 0.56* (1.76) 0.79*** (3.4) 0.19 11.5*** (0.00) 47 (2) (3) (4) (5) (6) (7) (8) (9) (10)

*Means significant at 0.10 level (two-tailed); ***Means significant at 0.01 level (two-tailed). t-values in parentheses. In cases where the compensation of the highest-paid board member was lower than the

average TMT compensation, I used the latter number. To save space, intercepts are not shown.

8.4.4 The power construct and executive pay

In order to reduce the number of variables and to improve interpretation of the executive power construct, I use principal component analysis. The cut-off point for retaining a factor is an Eigenvalue of 1 and higher. The results sort into six components (see Table 30). The four variables executive ownership, CEO duality, independent directors and non-executive directors load heavily on the first factor, which I term structural power. Interestingly, non-executive directors load especially high on this component, although it was not significant in regression on individual power variables. Interdependent directors and CEO tenure load heavily on the second component, tenure power. The largest outside shareholder and non-executive ownership load on the third component, outside

148

CHAPTER 8

ownership power. Executive outside board members and compensation committee existence load into a fourth component, committee and board membership power. Non-executive outside board members and doctoral degree load on the fifth component, which I do not define. And interlocks load on the sixth component, interlocks power. Board size and CEO celebrity status have no factor loading above the 0.5 threshold.
Table 30: Principal component analysis of archival power construct.
Variable 1 Largest shareholder Executive ownership Non-executive ownership CEO duality Independent directors Non-executive directors Board size CC existence Interdependent directors Ln CEO tenure Interlocks Executive outside board members Non-executive outside board members CEO cult Doctoral degree -0.149 0.675 0.051 0.634 -0.572 -0.760 -0.488 -0.155 0.020 0.105 -0.115 0.144 -0.223 0.148 0.035 2 0.076 0.062 0.010 0.280 0.251 -0.138 0.173 0.131 0.847 0.878 0.082 -0.011 -0.157 -0.216 0.184 Promax rotated factor loadings 3 0.744 -0.221 0.794 -0.115 -0.124 -0.009 -0.157 -0.094 -0.033 0.124 -0.260 0.254 -0.018 -0.323 0.057 4 0.135 -0.044 0.122 0.192 -0.118 -0.017 0.225 0.531 -0.005 0.021 0.199 0.867 0.332 0.292 -0.128 5 -0.285 -0.116 0.249 0.020 0.066 -0.017 -0.201 0.040 0.062 0.088 -0.012 -0.118 0.504 0.103 0.825 6 0.244 0.193 0.149 0.029 0.035 -0.067 0.150 0.133 -0.051 -0.051 -0.960 -0.334 0.022 0.192 0.022 Uniqueness 0.405 0.489 0.435 0.471 0.605 0.406 0.452 0.507 0.284 0.241 0.215 0.341 0.462 0.609 0.347

I also estimate factor scores, which I use as inputs in subsequent analysis. In Table 31 scoring coefficients are shown based on the promax rotated factors.

RESEARCH RESULTS

149

Table 31: Scoring coefficients of archival power construct.

Variable 1 Largest shareholder Executive ownership Non-executive ownership CEO duality Independent directors Non-executive directors Board size CC existence Interdependent directors Ln CEO tenure Interlocks Executive outside board members Non-executive outside board members CEO cult Doctoral degree -0.065 0.291 0.050 0.272 -0.288 -0.347 -0.268 -0.110 0.031 0.014 -0.005 0.078 -0.104 0.044 0.022 2 0.052 0.032 -0.043 0.134 0.251 -0.042 0.158 0.097 0.473 0.477 0.003 -0.034 -0.109 -0.107 0.040 Scoring coefficients 3 0.448 -0.162 0.516 -0.096 -0.120 -0.019 -0.204 -0.179 -0.053 0.058 -0.011 0.146 -0.066 -0.291 0.042 4 0.053 -0.011 0.024 0.150 -0.045 -0.003 0.233 0.450 -0.002 0.005 -0.006 0.550 0.277 0.299 -0.079 5 -0.256 -0.099 0.212 0.010 0.028 -0.014 -0.188 0.035 -0.007 0.014 -0.000 -0.074 0.451 0.113 0.694 6 0.112 0.171 -0.016 0.061 0.094 -0.015 0.268 0.270 -0.014 -0.019 -0.771 -0.196 0.091 0.278 -0.019

Factor-analyzed results of the survey power variables are shown in Table 32. Four principal components were retained. The different loadings are shown below. Performance satisfaction and expertise load most heavily on the first component, which I term credibility power. Discussion and meeting CEO load on the second component, persuasion power. Agenda and process power load on the third component, process power. And first proposal loads most heavily on the fourth component, proposal power.

150

CHAPTER 8

Table 32: Principal component analysis on survey power measures.

Variable 1 Discussion Meeting CEO Good relationship Friendship Performance satisfaction Expertise First proposal Agenda Process power 0.002 -0.019 -0.393 0.429 0.878 0.849 -0.092 0.200 -0.127 Promax rotated factor loadings 2 0.757 0.918 0.204 0.259 -0.040 0.026 -0.190 -0.189 -0.123 3 -0.094 0.151 0.453 -0.157 0.082 0.121 0.063 0.593 -0.790 4 0.075 -0.221 0.235 0.486 -0.004 -0.064 0.936 0.021 0.006 Uniqueness 0.376 0.205 0.509 0.377 0.235 0.283 0.179 0.589 0.370

Estimated factor scores used in subsequent analysis are shown in Table 33.
Table 33: Scoring coefficients of survey power construct.
Variable 1 Discussion Meeting CEO Good relationship Friendship Performance satisfaction Expertise First proposal Agenda Process power 0.012 -0.017 -0.232 0.245 0.462 0.444 -0.043 0.070 -0.021 Scoring coefficients 2 0.488 0.553 0.123 0.214 -0.025 0.010 -0.046 -0.146 -0.038 3 -0.117 0.085 0.368 -0.184 0.016 0.047 0.029 0.473 -0.622 4 0.128 -0.113 0.189 0.435 -0.000 -0.045 0.755 -0.019 0.024

8.5 The overall influence of power and complexity on executive pay

In this section I use the results of the principal component analyses of power and complexity to test the influence of these two constructs on executive pay. As Table 34 indicates, both constructs of complexity collected through archival and survey data collection are significantly and positively associated with executive pay. Objective

RESEARCH RESULTS

151

complexity, a construct of five dimensions, shows a very strong association with TMT and CEO compensation, even after controlling for all power components. Some of the power components also show a significant relationship with compensation levels and provide supportive evidence for hypothesis 2a that more powerful top management teams and CEOs are related to higher compensation. Structural power, on which CEO duality, non-executive directors, independent directors and executive ownership load especially high, is significantly associated with both top management and CEO compensation. Tenure power interdependent directors and CEO tenure shows no significant association with pay levels. Outside ownership power largest outside shareholder and non-executive ownership show a strong negative relationship with executive compensation levels. Committee and board membership power is also significantly positively associated with pay levels. Archival power component 5 doctoral degree and non-executives outside board memberships is not significantly related with compensation after controlling for objective complexity. Finally, interlocks power is not significantly associated with pay levels. Model (6) shows that complexity and power interact to influence executive pay. The significance of several interaction terms suggests that complexity moderates the relationship between managerial power and pay. The greater the level of complexity, the greater is the association between power and pay. This suggests that powerful managers have better opportunities to extract rents under conditions of high complexity.

152

CHAPTER 8

Table 34: Regressions of the archival complexity and power constructs on executive pay.
Dependent variable Ln total TMT compensation Independent variables Intercept Objective complexity Structural power Tenure power Outside ownership power Committee and board membership power Archival power 5 Interlocks power Ln revenues CEO dummy Complexity * structural power Complexity * tenure power Complexity * outside ownership power Complexity * committee and board membership power Complexity * archival power 5 Complexity * interlocks power Adjusted R2 F-statistics Observations 0.58 78.57*** (0.00) 398 0.58 78.71*** (0.00) 398 0.58 78.94*** (0.00) 398 0.61 79.11*** (0.00) 398 0.62 82.28*** (0.00) 398 0.081*** (2.88) -0.008 (-0.30) -0.112*** (-3.73) 0.141*** (4.43) 0.07** (2.57) 0.020 (0.68) 0.264*** (15.22) (1) 11.52*** (103.1) (2) 13.17*** (501.3) 0.495*** (15.24) 0.032 (1.18) -0.008 (0.28) -0.088*** (-2.89) 0.135*** (4.20) 0.23 (0.85) 0.023 (0.79) 0.12*** (3.41) 0.024 (0.75) -0.137*** (-3.88) 0.204*** (5.43) 0.06** (1.98) -0.012 (-0.35) 0.30*** (14.52) 0.447*** (6.28) 0.10*** (3.32) 0.012 (0.44) -0.075** (-2.52) 0.147*** (4.64) 0.045* (1.65) 0.008 (0.27) 0.26*** (15.00) 0.444*** (7.37) 0.385*** (5.37) 0.353*** (5.06) 0.066** (2.08) -0.012 (-0.37) -0.053 (-1.53) 0.167*** (4.62) -0.047 (-1.40) -0.065 (-1.51) 0.65 53.17*** (0.00) 398 (3) 11.37*** (85.54) Ln CEO compensation (4) cashb 11.46*** (102.1) (5) totalb 13.29*** (256.2) 0.576*** (14.75) 0.083** (2.33) 0.048 (1.46) -0.108*** (-3.02) 0.194*** (5.18) 0.013 (0.39) -0.009 (-0.26) (6) totalb 13.24*** (250.1) 0.580*** (15.02) 0.105*** (2.77) 0.068** (2.06) -0.076** (-2.14) 0.150*** (3.84) 0.011 (0.34) 0.035 (0.77)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses. aIn cases where the compensation of the highestpaid board member was lower than the average TMT compensation, I used the latter number. To test whether this procedure affects the results, I use a dummy variable for cases where I replace the highest board compensation with average TMT compensation. bCEO age is not significantly related to pay levels in these regressions and therefore not included as a control variable in reported regressions.

RESEARCH RESULTS

153

Table 35 presents Tobit regression results of the objective complexity and power components on executive pay structure. Objective complexity and the proportion of CEO compensation that is equity-based are positively and significantly related. This result is robust to the inclusion of the power components. Overall, there is strong evidence supporting the notion of hypothesis 1b that higher complexity leads to higher reliance on equity-based compensation components due to information asymmetry between the board of directors and the members of the top management team. The relationship between the power components and compensation mix is more ambiguous. I do not find a significant relationship between structural power and pay mix. There is evidence that tenure power is positively related to CEO and TMT pay structure. However, the result that power arising from a CEOs tenure decreases the proportion of cash in total compensation is inconsistent with a standard power perspective. Outside ownership power is strongly associated with lower reliance on equity-based compensation. Again, this result is not consistent with what has been hypothesized. Committee and board membership are also related to equity pay mix. There is also support that power dimension 5 primarily made up of doctoral degree and non-executives outside board memberships positively influences pay mix. There is also supportive evidence that interlocks power helps structure executive pay to become less equity-based. The control variable CEO age is significantly associated with pay structure. Older CEOs receive a smaller proportion of compensation in equity instruments supporting other empirical studies (Zajac and Westphal, 1994; Bryan et al., 2000a). There is only weak evidence that capital structure affects pay mix.

154

CHAPTER 8

Table 35: Tobit regressions of the complexity and power constructs on executive pay structure.
Dependent variable Independent variables Intercept Objective complexity Structural power Tenure power Outside ownership power Committee and board membership power Archival power 5 Interlocks power CEO age Debt-assets ratio LR chi2 Pseudo R2 Observations -0.020 (-1.14) 0.033* (1.91) -0.063*** (-3.67) 0.12*** (7.01) 0.036** (2.24) -0.041*** (-2.58) -0.009*** (-4.29) -0.000 (-0.56) 107.97*** (0.00) 0.28 398 (1) CEO equity pay structure 0.496*** (4.17) (2) CEO equity pay structure 0.602*** (5.34) 0.12*** (6.41) -0.007 (-0.42) 0.018*** (2.91) -0.028* (-1.67) 0.076*** (4.48) 0.013 (0.83) -0.026* (-1.75) -0.010*** (-5.04) -0.001* (-1.85) 149.07*** (0.00) 0.38 398 -0.000 (-0.54) 91.89*** (0.00) 0.31 398 -0.001 (-1.63) 121.75*** (0.00) 0.41 398 -0.016 (-0.78) 0.056 (1.19) -0.052*** (-3.77) 0.087*** (6.30) 0.035*** (2.71) -0.017 (-1.27) (3) TMT equity pay structure 0.026 (0.68) (4) TMT equity pay structure 0.064* (1.77) 0.084*** (5.47) -0.004 (-0.33) 0.024* (1.91) -0.029** (-2.14) 0.055*** (3.90) 0.018 (1.45) -0.006* (-1.63)

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses.

Table 36 presents regressions of survey complexity and power constructs on CEO compensation level and mix. Factor-analyzed subjective complexity is positively associated with CEO cash and total compensation. However, there is no significant association between complexity and pay structure. None of the power categories is significantly associated with neither compensation levels nor mix. This additional analysis based on survey results provides further support for hypothesis 1a, but no support for hypotheses 1b, 2a and 2b.

RESEARCH RESULTS

155

Table 36: Regressions of the survey complexity and power constructs on CEO compensation.
Dependent variable Ln CEO compensation Independent variables Intercept Subjective complexity Persuasion power Credibility power Process power Proposal power Adjusted R2 / Pseudo R2 F-statistics / LR chi2 Observations (1) cash 11.52*** (103.1) 0.260** (2.31) -0.110 (-0.99) -0.004 (-0.03) -0.153 (-1.38) 0.090 (0.77) 0.11 2.12* (0.08) 47 (2) total 13.71*** (111.36) 0.246* (1.93) -0.104 (-0.83) 0.017 (0.13) -0.149 (-1.19) 0.127 (0.97) 0.07 1.70 (0.16) 47 CEO equity proportion (3) 0.08* (1.94) -0.026 (-0.65) -0.006 (-0.14) 0.045 (0.81) 0.020 (0.49) 0.031 (0.73) 0.06 2.19 (0.82) 47

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses.

8.6 Sensitivity analysis

8.6.1 Does the composition of the compensation committee matter?
Earlier results show that the presence of a compensation committee increases, rather than decreases executive pay, inconsistent with the notion that the specialization of boards increases relative board power over top management (Zahra and Pearce, 1991). To test whether the result is driven by the composition of the compensation committee, I conduct an additional analysis on the sub-sample of companies that have a compensation committee. Table 37 shows the effect of compensation committee composition on TMT compensation. There is no evidence that executives present on the compensation committee lead to pay arrangements that disproportionally benefit executives even before controlling for other determinants. This may be partially explained by the fact that many executives, although they do not formally serve on the compensation committee, actually attend their minutes, sometimes in the function as committee secretary.

156

CHAPTER 8 Table 37: OLS estimation of the composition of the compensation committee on TMT compensation.
Dependent variable Ln total TMT compensation Ln total CEO compensation (2) 13.473 (234.065)*** 0.002 (0.01) (2) 11.30*** (55.69) 0.158 (0.54) 0.724*** (6.11) -0.005*** (-2.99) 0.372*** (14.50) 0.00 0.00 (0.99) 240 0.56 77.34*** (0.000) 240

Independent variables Intercept Compensation committee composition CEO duality Largest shareholder (%) Log revenues Adjusted R2 F-statistics Observations

(1) 13.473*** (234.07) -0.517 (-1.62)

(1) 11.316*** (70.051) -0.2143 (-0.90) 0.356*** (3.73) -0.0032 (-2.13) 0.2945*** (14.15)

0.01 2.62 (0.107) 240

0.52 65.48*** (0.000) 240

*Means significant at 0.10 level (two-tailed); **Means significant at 0.05 level (two-tailed), ***Means significant at 0.01 level (two-tailed). t-values in parentheses.

However, while committee composition does not appear to have a direct influence on compensation levels, executives present on the compensation committee may influence more subtle elements of compensation such as the likelihood of option repricing (e.g. Chidambaran and Prabhala, 2004) or the timing of option grants (Yermack, 1997).

8.6.2 Alternative measures for complexity variables

Diversification measured as the number of three digits SIC codes or as a dummy variable for the existence of multi-divisional firms, does not appear to be reflected in compensation levels and structure. One conclusion may be that boards do not compensate their executives for the complexity arising of the diversification of the firm. Another conclusion may be that the proxy used to measure diversification may be inappropriate in the context of compensation. I therefore rerun some of the regression including the size of the top management team as an indication of how diversified a firm is. The regression coefficient of TMT size is significant at the 10%-level in regressions of total CEO compensation on all individual complexity variables. However, factor loadings and scoring coefficients in factor analysis are lower than using the number of three digits SIC codes.

RESEARCH RESULTS

157

In my analysis, I use the logarithm of revenues as the proxy for firm size. Alternatively, I use a factor-analyzed composite variable of firm size including revenues, assets and number of employees. Results are unchanged. I also use the proportion of foreign employees to total employees (FETE) in my regressions as a proxy for the degree of internationalization. I rerun regressions using a factor-analyzed composite variable of internationalization including the proportion of foreign sales out of total sales in addition to FETE. This composite measure captures two of the three attributes of a firms degree of internationalization suggested by Sullivan (1994). Again, this alternative measure does not significantly affect the results.

8.6.3 Alternative measures for power variables

I measured board ownership with two individual variables, executive ownership and non-executive ownership. Following other studies (Dunn, 2004; Beiner et al., 2005), I build a relative measure of ownership power, the number of voting rights by executives as a percentage of voting rights by non-executive directors. This variable is not related with compensation in a number of test regressions. As an alternative measure of CEO duality, I measure whether the CEO also sits on the board of directors. Descriptive statistics have shown that this is a more frequent event than CEO duality. In roughly half of the sample firms, the CEO is a member of the board. CEO board member is measured as a dummy variable 1, if the CEO sits on the board of directors and 0 otherwise. Results are essentially unchanged to the ones reported.

8.6.4 Alternative measure of firm performance

The relatively large ownership concentration of Swiss firms suggests only medium market liquidity. I therefore also measure accounting performance in terms of return on sales. The inclusion of an accounts-based measure, in addition to a direct measure of shareholder returns, reflects the view that incentive-based compensation systems should include variables that are not only observable but also controllable by the actions of the chief executive (Holmstrm, 1979). In this context it can be argued that accounts-based measures are more stable and less subject to speculative and exogenous shocks than stock market-based measures. Also, it has been suggested that both accounting and market return performance measures should be included to avoid underestimating the overall impact of financial performance on executive pay (Joskow and Rose, 1994). A

158

CHAPTER 8

further argument for accounts-based measures of performance is the fact that Swiss companies heavily rely on accounting-based metrics such as EBIT or Net Profit to evaluate the performance of their executives (Stern and Peck, 2003: 23). My measure of return on sales is the average of the current-year return plus the returns of the two previous years. Rerunning regressions on the basic determinants of executive compensation suggests a negative association between accounting performance and pay levels. In other words, firms which had poor accounting returns paid their executives higher compensation than firms that had strong accounting returns. This result is inconsistent with a pay-for-performance perspective. However, it is consistent with the idea that executives of particularly troubled firms get compensated for the additional riskiness of their job.

DISCUSSION

159

Discussion

The basic discussion of results including a comparison with previous research was conducted in the results chapter, which also includes a sensitivity analysis. This chapter discusses additional topics in a more detailed manner.

9.1 Complexity and executive pay

This work has attempted to provide a formal test of the complexity-pay hypothesis allowing for a more definitive statement about the effects of complexity on executive compensation. Factor analyzed results suggest that the objective complexity construct helps explain more than 50% of the variation in total CEO compensation and about a fourth of CEO equity-pay mix. The results further show that there are multiple facets of complexity with firm size just being one such facet. In addition to size, internationalization and diversification, this paper introduces two additional dimensions of complexity. These are market uncertainty and a politicized environment. Despite strong theoretical ground I find no evidence that market uncertainty, measured as the volatility of sales revenues, helps explain compensation levels, when regressing executive pay on individual complexity variables. This may be due to high collinearity of complexity variables. Indeed, as the correlation matrix indicated, volatility of sales revenues is highly correlated with the other measures of complexity. Using factor analysis results, the market uncertainty variable loads very high on the one principal component retained to build the complexity construct. The other new dimension of complexity is politicized environment. The measure used, the number of newspaper articles published about the focal firm, is a highly significant determinant of executive compensation level and structure. This suggests that compensation committees take into account the extent to which a firms environment is politicized when making compensation decisions. The results confirm that multidimensional constructs also need to be tested on a multidimensional basis and that the application of individual variables will not suffice. Indeed, all five variables of complexity score relatively high on the principal component retained for complexity.

160

CHAPTER 9

9.2 The incremental contribution of firm size

Firm size has been established as the major determinant of executive pay levels, accounting for about 50% of the variation in pay levels (Tosi et al., 2000). However, as argued in this paper, the contribution of firm size to explain executive pay may be overestimated and suffer from an omitted variable problem. Other measures beyond firm size, which have often been omitted in prior studies, contribute to the complexity of the CEOs job and are therefore a relevant predictor of pay levels. Some of these other dimensions of complexity help explain executive pay even after controlling for firm size. Due to the high correlation between firm size and the other complexity measures, it is difficult to estimate the true contribution of size to explain compensation levels. However, as the factor scores indicate internationalization and a politicized environment almost equally contribute to the complexity construct. And market uncertainty and diversification do not score much lower either. Two investigations help to qualify the unique contribution of firm size in explaining executive pay. First, this study reports an elasticity of CEO cash and total pay to firm size of 0.17 and 0.27 respectively after including other complexity variables. Before the inclusion of the other complexity variables the corresponding elasticities are 0.32 and 0.39 respectively. Second, I run additional regressions with the complexity variables. In the first regression, I exclude the firm size variable, in the second regression, I include all complexity variables. By including firm size, the adjusted R-squared increases from 0.486 to 0.531. In other words, the inclusion of size in explaining pay levels increases the explained variance by 4.5 percentage points.

9.3 Technological complexity and executive pay

In addition to the dimensions studied in this paper, firm complexity may also arise from high reliance on research and technology (Henderson and Fredrickson, 1996; Balkin et al., 2000). Increased R&D activity is typically associated with higher levels of technological uncertainty, more sophisticated and differentiated products, more specialized personnel and higher levels of intra-firm technology pooling. This argument suggests that if executives are paid for the technological complexity of the firm they run, executive pay will be positively related to R&D activity. And as high complexity associated with R&D activity is also decreases monitoring-ability, this may lead to an increased reliance on equity-based compensation.

DISCUSSION

161

Although not explicitly evoking a complexity explanation, empirical research in accounting and finance mainly corroborates this view. R&D expenditure has been studied frequently in prior executive compensation research, but used as a proxy for the growth opportunities of firms. Evidence has been mostly supportive. Ryan and Wiggins (2001) find that the absolute use of stock options is positively related to R&D intensity. However, R&D intensity exerts a negative influence on the absolute use of restricted stock in their study. Brick et al. (2002) and Buck et al. (2003) find a positive association between R&D-intensity and CEO total compensation. Kole (1997) shows that the use of restricted stock plans is positively related to R&D-intensity and Barron and Waddell (2003b) document that the use of equity-based compensation increases with the R&D-tobook value of assets ratio. Bryan et al. (2000a) find a positive relationship between R&D intensity and the proportion of CEO stock options out of total pay. Milkovich et al. (1991) find that higher technology businesses preclude all but outcome-based compensation designs, since evaluating appropriate behavior is inefficient in complex situations. I do not include technological complexity into my study, as I use companies from a wide variety of industries including banks, insurance companies and real estate firms. By controlling for industry effects, I find some evidence that technological complexity is related to executive compensation: Executives in the technology industry earn substantially higher total compensation than their counterparts in the basic material industry. The difference is only larger for executives in the financial and healthcare industries.

9.4 The components of complexity

The results of correlation and principal component factor analysis showed that the complexity variables internationalization, size, diversification, market uncertainty and politicized environment are highly interrelated. This suggests that the variables pick up different nuances of the same complexity construct. For instance, as firms grow, other dimensions of complexity also appear to increase. Larger firms seem to be more likely to have international operations. Acquisition-driven growth is likely to increase the number of businesses the firm operates in and thus to increase diversification. Larger firms may also be more likely to attract public attention and therefore face a more politicized environment. Furthermore, the high correlation between market uncertainty and a

162

CHAPTER 9

politicized environment suggests that firms which are exposed to stronger market shocks also appear to receive more public attention.

9.5 The components of managerial power

The fifteen measures of power sort into six factors. Largest shareholder and nonexecutive director ownership sort into an outside ownership power category. Executive ownership, CEO duality, non-executive directors and independent directors sort into a structural power category. These results are generally consistent with hypothesized dimensions and are strikingly similar to those obtained by Boyd (1994). His five measures of board control sort into two factors with CEO duality, ratio of insiders and level of director compensation sorting into a structural power category, and board stock and institutional ownership sorting into an ownership power category. The two tenure variables interdependent directors and CEO tenure form the dimension tenure power. And interlocks are an own dimension interlocking power, as hypothesized. However, the remaining two components are inconsistent with the hypothesized conceptual framework. Executive outside board members and compensation committee existence form a common dimension of power. The former variable was hypothesized to be a network power variable, the latter to be part of structural power. Non-executive outside board members and doctoral degree form a final category. The former variable was hypothesized to be part of network power, the latter to be part of expertise power. It suggests that executive power has more aspects than originally thought. The results confirm the importance of ownership structure as a determinant of executive compensation. The outside ownership power dimension is significantly negatively related with TMT and CEO compensation levels. This suggests that companies with a large outside shareholder are more effectively constraining executive compensation. And providing equity-based incentives to non-executive directors may also be an effective way of eliminating rents for CEOs.

9.6 Interdependent directors and executive pay

In this study, the proportion of interdependent directors out of the total number of board directors did not affect compensation level or structure. There is only a weak positive relationship between tenure power, which includes the interdependent directors variable, and TMT equity compensation mix. This non-existent or weak effect of

DISCUSSION

163

interdependent directors on executive pay may partially be due to a conflicting hypothesis. It has been argued that escalation behavior by the board in favor of the CEO (e.g. higher pay and decoupling pay and performance) is more likely to occur when the current board hired the CEO because board members feel personally responsible for observed performance outcomes (Staw, 1976; Gomez-Mejia, 1994). This suggests a negative association between interdependent directors and CEO compensation. On the contrary, the managerial power hypothesis suggests that directors elected after the CEO has taken office, will feel thankful for their election and be less critical with respect to the CEOs performance.

9.7 Compensation committee existence and executive pay

The strong positive association between the presence of a compensation committee and executive pay level is consistent with prior studies in the UK (Conyon and Peck, 1998; Main and Johnston, 1993) and raises questions about the theoretical prediction that specialized committees are a characteristic of powerful boards (Zahra and Pearce, 1991). The effect is also economically significant. CEOs of firms with a compensation committee earn almost 40% higher total compensation. The difference is still more than 30% for the average TMT member. Codes of Best Practice in Corporate Governance have taken up the idea that the professionalization of boards requires specialized committees. Implicit in this assumption is the idea that effective governance through a specialized compensation committee will lead to more optimal or efficient compensation levels. However, the presence of a compensation committee in fact increases rather than decreases compensation levels. This result is clearly inconsistent with the idea that a compensation committee decreases the relative power of executives over their board. The result is not influenced by the composition of the compensation. There is no support for the hypothesis that the presence of executives on the compensation committee influences compensation levels. Thus, compensation committee existence appears to be a very poor measure of board power in the context of executive pay. However, the result is consistent with other interpretations of a compensation committee existence. Another explanation for the positive association between compensation committees and compensation is that poor governance companies were quick to establish such a committee to show their formal compliance, while actually not making much

164

CHAPTER 9

use of it. Such a symbolic view of governance practices is portrayed by Westphal and Zajac (1994) who show that many companies adopted long-term performance plans without actually using them. Influential CEOs drove the adoption of such plans but not their use. In a similar vein, influential CEOs may be quick to accept compensation committees and to use their power to get above-average pay. Also, the result is consistent with a bidding-up hypothesis (Ezzamel and Watson, 1998). This argues that boards respond to deviations from the going market rate with subsequent pay adjustment and that this adjustment is especially observable for executives that are paid below market. A further possible explanation is that better quality managers may simply join companies with a compensation committee. In addition to the issue of conflicting hypotheses, the measure of disclosed compensation committee structure may be very noisy. Frequently, the CEO participates in the meetings of the compensation committee without formal voting power. However, this presence allows him to engage in persuasion behavior (Westphal, 1998). This presence not considered in the committee structure variable may obscure the effects of a truly independent compensation committee.

9.8 Managerial power and pay structure

I hypothesized that CEOs and TMT members will use their power to negotiate a pay structure according to their preference, which is more stable and less dependent on performance. This is the view commonly advocated by other researchers (e.g. Hill and Phan, 1991; Grabke-Rundell and Gomez-Mejia, 2002). I expected power to be inversely related to the proportion of compensation that is equity-based. However, not only can I find very little support for this hypothesis, but some dimensions of executive power also appear to positively influence the proportion of stock and option pay in managerial compensation contracts. It is either possible that my measures are insufficient proxies of the underlying construct. Or, and more likely, the assumption that executives prefer cash over equity compensation may be flawed. The existence of political constraints is consistent with empirical results. More powerful CEOs are able to extract rents in their compensation contracts, but they are limited by internal and external political constraints. Public opinion tends to perceive high pay as unconscionable (Jensen and Murphy, 1990a). Arguably, these outrage constraints are higher for cash compensation than for equity-based compensation (Bebchuk and

DISCUSSION

165

Grinstein, 2005b). There is anecdotal evidence that equity compensation is more legitimate as it provides a direct link between the fortunes of the CEO and those of shareholders (Dial and Murphy, 1995). Also, stock options are a less visible means of increasing executive pay (Hall and Liebman, 1998). Option valuation is inherently imprecise and provides companies with considerable discretion in the valuation of options. Companies may only disclose the number of options granted, making it very costly for the uninformed reader to gauge the value of these options. Some Swiss companies disclose the tax value of stock options. However, this tax value may significantly deviate from the Black-Scholes option value, the most accepted valuation method for stock options. More particularly, the tax value decreases with the blocking or vesting period of an option as a bonus for a more long-term orientation but the Black-Scholes value increases with the time to maturity (which is generally linked to the vesting period).51 In light of political constraints, a positive relationship between power and equity compensation mix is therefore plausible. A positive relationship between power and the proportion of equity-compensation should also be expected for US data, where law prohibits the tax-deduction of CEO cash compensation above 1 MUSD. Beyond a pure tax effect of making incentive pay more attractive to firms, political constraints may be in place and more powerful CEOs may be compensated with higher equity compensation. This view is consistent with the empirical finding that stock options have become the majority component of US and UK executive pay and that option compensation usually has been granted in addition to and not as a replacement for cash compensation (Murphy, 1999; Main et al., 1996). In the US, many large companies now pay their top executives so called gross-ups, benefits that are carefully hidden in the annual reports footnotes in order to not provoke public displeasure (Niederberger, 2005; Bebchuck and Fried, 2004). This also provides an explanation for the ambiguous results found in prior studies linking managerial power to pay structure. Despite their potential importance, these political constraints have not received much attention. An explanation may be that they are difficult to document because they operate in informal and indirect ways (Jensen and Murphy, 1990a: 227). This discussion suggests a new perspective of how to interpret corporate governance structures, managerial power and pay-performance relationships. For example, the
51

For each year that the options are blocked, the tax value is reduced by roughly 6%.

166

CHAPTER 9

positive relationship between CEO duality and equity-based compensation mix could be interpreted as firms substituting the lack of structural balance with strong payperformance incentives. However, the power perspective advanced in this work argues that more powerful CEOs influence the board to get more compensation in equity instruments, which are politically more acceptable than cash payments. The only acceptable way to get higher compensation may be through stock options and restricted stock awards. For instance, additional regressions show that the CEO duality pay premium is significantly larger for total compensation than for cash compensation: While CEOs who are also the chairman of the board earn, on average, 38% more in total compensation, they earn 27% more in cash compensation. Also, the proportion of equity-based compensation and total compensation are significantly correlated. This suggests that more powerful CEOs get disproportionally large increases in equity-based compensation. The assumption that CEOs prefer more stable pay may be generally correct, but is very misleading in the context of political constraints. From a power perspective, all executives want is more compensation, and if it is difficult to get higher cash compensation, options and stock may do it as well. This perspective of managerial power challenges the view that optimal contracting requires higher pay to compensate managers for their increased risk-taking as a result of equity pay (Core et al. (2004). In short, the managerial power hypothesis may still be valid, but its assumption on pay structure may have to be refined.

9.9 CEO duality as a self-serving opportunity?

The combination of the CEO and board chairman positions has been met with criticism by corporate governance activists for a long time. The results of this work suggest that CEOs who also occupy the position of the chairman of the board earn a pay premium of about 35-40%. At least two reasons provide legitimate reason for a positive association between CEO duality and pay levels. It has already been outlined that CEOs who also occupy the position as a chairman of the Board of Directors should be rewarded with higher pay for their dual responsibility (Guay, 1999). This explains part of the pay differential. Furthermore, it has been argued that the duality structure may be advantageous in situations requiring strong leadership (Finkelstein and DAveni, 1994). This is the case, for instance, in a turnaround situation. Hence, the association between CEO duality and higher compensation may be due to special circumstances leading to

DISCUSSION

167

higher pay. Nevertheless, this large premium suggests that there is at least some managerial self-serving in place. Core et al. (1999) find support for this contention. In their study, CEO excess compensation resulting from a weak governance structure such as CEO duality is negatively associated with subsequent accounting performance. This provides support for a managerial self-serving view.

9.10 Board size and executive pay

I argued that larger boards are easier to control by the CEO and that executives prefer cash over contingent equity compensation. As a result, board size should be negatively associated with the proportion of equity-based compensation. Contrary to this hypothesis board size is positively related to TMT equity compensation mix. Further, results do not indicate that board size influences compensation levels. Neither does board size load significantly on any of the power categories. This does not support prior empirical results from the UK (Conyon and Peck, 1998) and the US (Core et al., 1999). Part of the explanation may be that board size has been optimized over time to reflect best practice in Corporate Governance. Indeed, Swiss boards typically have between six and nine members and there is little variability, which makes it difficult to find a statistical effect. Another possible explanation is that smaller boards may in some circumstances actually be easier to control. This is the case when CEOs are heavily involved in choosing new board members, for instance by being a member of the nomination committee. In these cases, the board will more quickly fill up with loyal board members (Grinstein and Hribar, 2004).

9.11 The influence of Anglo-Americans

Although only invoked as a control variable the influence of Anglo-American board members on executive compensation deserves closer attention. Empirical results clearly indicate that Swiss companies that have a larger proportion of Anglo-Americans on their board or compensation committee are associated with significantly higher executive compensation. Anglo-American board members seem to be a major driving force behind the postulated increase in executive compensation in Switzerland. The same is true for Anglo-American TMT members. The inclusion of these variables makes many other variables insignificant. However, the interpretation of these results is not straightforward. Does it mean that, beyond firm size, the nationality of board and TMT

168

CHAPTER 9

members explains compensation levels and structure? Several explanations are consistent with this result. First, a social comparison perspective (OReilly et al., 1988) suggests that Anglo-Americans take their US or UK executives as comparison when deciding on pay. And this work has shown that US and UK executives tend to have higher total compensation and higher reliance on restricted stock and stock options. Second, this is closely related to an institutional perspective. Equity-based compensation has become institutionalized in most Anglo-American countries. It is therefore only normal to postulate it in Switzerland as well. Third, the relationship may also be spurious, picking up an underlying effect that has not been included such as risk. The Anglo-American TMT members in my sample may be especially powerful. Due to these theoretical and methodological reservations, I do not include the ratios of AngloAmericans on boards and TMTs, fully aware that regressions may suffer from an omitted variable problem. However, other studies have not included this variable either. To sum up, it is unclear what influences this important relationship and its reasons need to be further investigated. A first contribution in getting a clearer understanding on the influence of the Anglo-American system on European executive pay is the study by Oxelheim and Randy (2005). They show that higher exposure to Anglo-American financial influence increases CEO compensation in Norway and Sweden.

CONCLUSIONS

169

10 Conclusions
10.1 Contributions to the literature
This paper has broadened the usual empirical approach to the study of executive compensation. In the first place, a broad concept of complexity and power has been developed and their influence on executive compensation has been tested. Prior research has never tested how the concept of complexity influences compensation contracts despite strong theoretical ground and compensation consultants inclusion of complexity in job evaluation work. Suggested more than 15 years ago as a potentially important determinant of executive pay (Finkelstein and Hambrick, 1989), the complexity-pay relation has received very little attention ever since and operationalization of complexity has by and large been limited to single sources of complexity such as firm diversification (Finkelstein and Hambrick, 1989) and internationalization (Sanders and Carpenter, 1998). By building and testing a multidimensional framework of complexity, this work shows the large explanatory power of complexity in explaining top management pay levels and structure. Two of the analyzed dimensions of complexity market uncertainty and politicized environment have not received any attention in prior research. Although power has attracted much more attention from compensation researchers, empirical results are often inconsistent and ambiguous. Therefore, Finkelstein and Hambrick (1996) suggest that a formal test of the power hypothesis is needed, one that incorporates multiple aspects of executive power and allows for a more definitive statement about the effects of power on executive compensation. In this paper I have developed and tested a framework of executive power that incorporates multiple dimensions. In short, this study offers some insight into methodological issues surrounding the study of complexity and managerial power. In particular, it provides a strong argument for future researchers to consider using multiple indicators when examining complexity and power. A second major innovation of this work is the identification of more subtle power and complexity variables through means of a questionnaire survey. My results provide firsttime evidence that the compensation setting process provides different degrees of power for CEO to influence their compensation. I also find evidence that subjective judgments of complexity affect top management compensation. In general, the results of this work

170

CHAPTER 10

contribute to our understanding of the determinants of compensation levels and structure. Power and complexity have been established as multi-faceted concepts, which also need to be met with corresponding methodological sophistication. To reflect a multi-faceted construct, a large number of variables may need to be included. And to capture the nuances of these potentially highly correlated variables, factor analysis has been suggested as an important method to achieve a more definite statement on the complexity-pay relation and the power-pay relation. In general, the right way forward appears to be to try to capture potentially multi-faceted constructs such as power or complexity with adequate methods such as principal component analysis to obtain a more complete picture of power. Third, this work has provided an in-depth and updated literature review on the determinants and consequences of executive pay. Good general reviews of the literature were written in the 1990s (e.g. Murphy, 1999; Gomez-Mejia and Wiseman, 1997; Finkelstein and Hambrick, 1996). Since then, a large volume of theoretical and empirical research has been published. Chapters 3 through 6 have provided an update of the accumulated knowledge about executive compensation.

10.2 Summary of results

10.2.1 Overview of executive and director compensation
In 2004, the average (median) CEO in Swiss stock-listed companies received total compensation of 2.5 (1.0) MCHF. There is large variation in compensation levels. The minimum total CEO compensation in my sample in 2004 was 180,000 CHF and the highest was 23.3 MCHF. CEOs further received average (median) cash compensation of 1.6 (0.9) MCHF in the same year. On average, CEOs received company stock options with a Black-Scholes value of 450,000 CHF. However, the median CEO did not receive any company stock options. On top of that, the average CEO further received shares with a market value of 550,000 CHF. Again, the median CEO was not awarded company shares. This suggests that equity-based pay is used for only a relatively small group of highly paid CEOs. The average proportion of equity-based pay out of total pay was 13.9%. The compensation of members of the top management team is substantially lower. Average and median total compensation values are 980,000 CHF and 560,000 CHF respectively. Average (median) cash compensation amounted to 720,000 CHF (500,000

CONCLUSIONS

171

CHF). Total awards made in equity-based elements were, on average, 11.3% of total pay. Thus, average TMT compensation is somewhat less incentivized than CEO compensation. The median value for option and share awards individually is zero. Those executives receiving higher total compensation also tend to receive a larger part of their remuneration in equity instruments. Over the years 2002 to 2004, boards partially replaced option-based compensation with direct shares awards and higher cash compensation. In this way, the increase in share compensation more than offset the decrease in option-based compensation. The average non-executive director received total compensation of 135,000 CHF in 2004. Total median non-executive director compensation was 83,000 CHF. Mean (median) cash compensation for non-executive directors was 104,000 CHF (72,000 CHF) in 2004 respectively. The majority of boards did not grant stock options to its nonexecutive members. Also, the median board did not grant any shares to its nonexecutives. On average, non-executive directors received company shares with a value of 23,000 CHF. As is the case with executive pay, non-executive board members replaced much of the stock option grants with restricted stock awards in their own compensation contracts. For non-executives, the additional value of shares granted also more that offset the reduction in the value of stock option pay. Non-executive directors received approximately 14% of their total compensation in equity-based pay.

10.2.2 Basic determinants of executive pay

Consistent with prior literature (Tosi et al., 2000), the main determinant of compensation levels is firm size. Current-year total shareholder return is not a significant predictor of TMT and CEO compensation levels. Interestingly, when regressions are run for the years 2002 and 2003 separately, shareholder return is positively related to total TMT compensation in 2003 at the 5% significance level and negatively, but not significantly, in 2002. As average stock market returns were positive in 2003 and negative in 2002, this suggests that managers are rewarded for higher shareholder returns but not punished for weak performance. This result is consistent with prior US research (Hambrick and Finkelstein, 1995). There is also some evidence that internationalization is positively related to total CEO compensation. Shareholder concentration, measured as the ownership stake of the largest outside shareholder, is negatively associated with TMT compensation level, consistent with prior studies (e.g. Werner and Tosi, 1995; Khan et al., 2005). I also include CEO duality as a firmly established determinant of CEO

172

CHAPTER 10

compensation (e.g. Conyon, 1997). There is a wage differential of approximately 40% (e0.341-1) between CEOs who are also chairman of the board and those who are not. Overall, the results confirm the common determinants of CEO compensation levels established in the executive pay literature. These are shareholder concentration, firm size, internationalization and CEO duality. They also confirm the weak and sometimes non-existent relationship between current shareholder returns and executive pay. My results of the determinants of executive pay structure are also broadly consistent with earlier studies. Larger firms tend to award a larger proportion of equity-based compensation to the TMT and the CEOs. Firm internationalization is a significant positive determinant of CEO equity-compensation mix. Firms with more concentrated ownership tend to have lower reliance on equity-based compensation for both the average TMT member and the CEO. CEO age is negatively related to CEO equity compensation mix. Executive ownership is negatively, but not significantly, related to executive equity-pay mix. Capital structure is not significantly related to compensation structure.

10.2.3 Complexity and executive pay

I built a framework which divides complexity into the five different dimensions: size, internationalization, diversification, politicized environment and market uncertainty. If I regress individual complexity variables on executive pay, relationships are significant for size, internationalization and a politicized environment, but not for diversification and market uncertainty. However, correlation and factor analysis indicate that the individual complexity variables are highly correlated with each other, and that they all sort into one principal component. Supporting the notion that managers whose job is more complex should be compensated with higher pay, I document a positive relation between various measures of executive pay levels and the factor-analyzed complexity construct. It helps explain more than 50% of the variation in executive pay levels. I also find some support for the notion that CEOs and top management teams who are more difficult to monitor due to more complex jobs should have their compensation more closely tied to shareholder value creation. The factor-analyzed complexity construct is significantly associated with CEO and TMT equity-compensation mix. On an individual variables basis, internationalization and politicized environment are

CONCLUSIONS

173

significant predictors of compensation mix, while size, diversification and market uncertainty are not. Compensation decisions are made by a small group of board members. As they consider complexity as a justifiable way to determine compensation levels and structure, their personal judgment about the complexity of the CEOs job may add additional information. I therefore ran a survey with board or compensation committee chairmen. The results of the survey support the notion that perceived complexity is an important determinant of executive pay. Also, factor-analyzed complexity based on the variables surveyed (subjective complexity) is positively related to total CEO compensation. Board members personal judgment about the complexity of the CEOs job appears to influence compensation levels. However, the subjective complexity construct is not significantly related to CEO equity-pay mix.

10.2.4 Power and executive pay

Similarly to the complexity construct, I also built a broad conceptual framework of power in the context of executive compensation. Regressing individual power variables on compensation levels returns several significant relationships. CEO and TMT compensation levels are lower, the higher the ownership stake of the largest outside shareholder and the larger the proportion of independent directors on the board. CEO and TMT compensation levels are substantially higher when the CEO also occupies the position of the chairman of the board, when the CEO has a particularly high status, and when the board has a compensation committee. With the exception of the presence of a compensation committee, all signs of these relationships are consistent with hypothesis 2a that greater executive power is associated with higher compensation levels. Tobit regression results document significant relationships between managerial power variables and executive compensation mix. The likelihood that a given Swiss franc of TMT and CEO compensation is equity-based is higher when non-executive ownership is higher, CEOs also occupy the board chairman position, the board has a compensation committee, or TMT members occupy board positions on other stock-listed companies. The likelihood that a given Swiss franc of TMT and CEO compensation is equity-based is lower the higher the ownership stake of the largest shareholder and the higher executive stock ownership. Further, CEO equity-pay mix is positively associated with interlocks and CEO celebrity status, and TMT equity-pay mix is positively associated

174

CHAPTER 10

with board size. However, only the sign of the regression coefficients for executive and non-executive ownership as well as compensation committee existence are consistent with hypothesis 1b that greater managerial power is associated with a lower proportion of shares and options in total compensation. To detect structure in the power variables I apply factor analysis. The fifteen archivalbased power variables sort into six different principal components. The three structural variables, CEO duality, independent directors and non-executive directors as well as executive ownership, load heavily on the first factor, which I termed structural power. The two tenure variables, interdependent directors and CEO tenure, load heavily on the second component tenure power. The two ownership variables, largest outside shareholder and non-executive ownership, load on the third component, outside ownership power. Executive outside board members and compensation committee existence load into a fourth power component. Non-executive outside board members and doctoral degree load into the fifth power component. And interlocks loads on the sixth component interlocking power. Some of these power components are significantly related to pay levels. Structural power as well as board membership and committee power are positively associated with TMT and CEO compensation levels. Outside ownership power is negatively associated with compensation levels. Power component five including doctoral degree and non-executives outside board memberships is positively associated with compensation levels, but only before controlling for complexity. Overall, these results provide support for hypothesis 2a that more powerful CEOs and TMTs are associated with higher cash and total compensation levels. The power components also influence compensation structure. Outside ownership power and interlocks power are negatively associated with equity compensation mix. Power components four and five as well as tenure power are positively associated with equity-pay mix. Overall, these ambiguous results yield mixed support to the contention that more powerful executives are associated with more invariable pay. I argued that CEOs prefer cash compensation over equity compensation, as the former provides more stable income. Thus, CEOs gaining the preponderance of power would be able to influence the compensation contract to get a larger proportion of non-contingent cash pay.

CONCLUSIONS

175

This study also provides first evidence on more subtle elements of power. The nine power variables from the survey sort into four components. Performance satisfaction and expertise load most heavily on the first component, credibility power. Discussion and meeting CEO load on the second component, which I term persuasion power. Agenda and process power load on the third component process power. And first proposal loads most heavily on the fourth component proposal power. In a subsequent analysis, I found no evidence that these dimensions of power influence compensation levels and structure.

10.3 Implications for policy makers and practitioners

This study has several implications for policy makers and practitioners. First of all, structural power including the variables CEO duality, non-executive directors, independent directors and executive ownership is significantly related to pay levels. This suggests that the preference and recommendations of the Swiss Code of Best Practice in Corporate Governance (Economiesuisse, 2002) to split the positions of the CEO and the board chairman and to fill the board with non-executive and independent directors helps preventing self-serving managerial behavior. Second, the results of this study indicated that some aspects of managerial power increase rather than decrease contingent pay. This surprising result was explained by a relatively lower political constraint of awarding stock options and shares as opposed to straight cash compensation. This perspective implies that comments in the media may actually help constrain pay levels by telling CEOs what is considered excessive and what is not. Media and institutional investors voicing their concern with pay levels may play an important role in providing guidance to compensation consultants and compensation committee members. It further suggests that the voting on executive compensation packages at the annual general meeting may function as a suitable ex-ante political constraint on pay levels, even though very few packages may actually be voted down in the end. Third, this studys result suggests that boards take complexity into consideration in the design of compensation contracts. A firms degree of complexity arising from strategy, organization and environment affects board monitoring costs and the inherent risks top managers face. Boards need to recognize these costs and risks, and ensure that top executive compensation contracts are consistent with their degree of complexity. In

176

CHAPTER 10

particular, firms with a higher degree of complexity have more difficulty in monitoring managers actions. Compensation contracts in such firms should therefore have a higher proportion of equity-based compensation. In addition, to ensure that high quality top executives are retained, the board should compensate executives for the added complexity. Fourth, survey measures of power have revealed a positive relationship between CEO involvement in the compensation negotiation process and total CEO compensation. For companies to improve their internal corporate governance arrangements, this suggests that boards need to make an informed decision about CEO pay without the involvement of the CEO himself. At the least, CEO compensation should not be a matter of negotiation between the chairman and the CEO only.

10.4 Limitations
My findings, their interpretation and my subsequent discussion must be considered in the context of the studys limitations. A primary set of limitations is related to the measurement of variables. First, in regression of CEO compensation, I include average TMT compensation values for a large minority of companies that do not disclose CEO compensation. To mitigate this limitation, I control for the inclusion of average TMT values with a CEO dummy variable in all regressions. For future empirical research conducted in Switzerland, CEO compensation data should soon be universally available, as a new law has been proposed to require the disclosure of compensation data for all board members and the CEO on an individual basis (Wittwer, 2005c: 25; EJPD, 2003). Second, network power was assessed by executives memberships on other boards of firms stock-listed at the Swiss Exchange SWX and related interlocking directorates resulting from these board memberships. While these board memberships certainly are an important means of building a network, top managers have further opportunities to develop their network power. Such opportunities include the Swiss-American Chamber of Commerce or the local alumni association of Harvard Business School, which organize regular meetings. It is possible that by extending the measurement of network power, one might get a better picture of how the managerial network provides power. Third, I measured the degree of internationalization as FSTS and FETE. This may not capture adequately the cultural diversity associated with international diversification. There is evidence that the extent of cultural relatedness between firm operations and

CONCLUSIONS

177

markets is related to the complexity of the firm (Gomez-Mejia and Palich, 1997). A more refined measure of internationalization may consider cultural relatedness. Nevertheless, I was able to find significant effects with my variables. Another potential limitation concerns the use of the questionnaire survey. While the survey responses allowed a first idea of how more subtle and subjective elements of managerial power and complexity influence compensation, results are based on 47 responses. This relatively small sample size suggests that results can only be viewed as preliminary. To keep an acceptable number of degrees of freedom, I was restricted to running univariate regressions. Nevertheless, the surveys results provide some first estimates. Finally, it has to be noted that this study investigates corporate governance structure in terms of CEO and TMT compensation. A board that is optimized for making compensation decisions may destroy value by making bad decisions on other potentially more crucial items (Core et al., 2004). Optimal board structure and corporate governance arrangements need to maximize the value of the firm and not just the costs of managerial compensation.

10.5 Future research

While of explanatory nature, my study indicates the need for more theoretical and empirical investigation of the effects of complexity and power on executive compensation. This work suggests a multifaceted framework of complexity to explain compensation level and structure capturing a construct with corresponding methodological sophistication. To corroborate the results of this study, it would be warranted to duplicate this study with data in other countries, namely the US, where much of the existing literature on executive pay has been produced. One obvious direction for future research would be to apply such a wide construct of power on design features of pay other than level and mix. This work incorporates multiple aspects of managerial power and measures its contribution to explaining executive pay level and structure. Little is known on how managerial power in a broad sense affects stock option and restricted stock plan design. For instance, researchers may investigate whether executive power is related to the holding and vesting periods of stock options and restricted stock or the time to maturity of stock options. Further, it may

178

CHAPTER 10

be fruitful to investigate whether a broad power construct helps explain the timing of option awards and news announcements (Aboody and Kasznik, 2000; Yermack, 1997). In the US, researchers may attempt to measure managerial power as a determinant of the amount of wealth transferred to CEOs by stealth compensation52 (Bebchuk and Fried, 2004: 101). Researchers may also study whether managerial power affects the relative proportion of option and stock in total compensation. Were firms of more powerful CEOs quicker to shift option pay to restricted shares awards during the bear market in 2001 and 2002? Another question relates to the dynamic link between complexity and executive compensation. This work provides strong evidence that higher complexity is associated with higher compensation levels and equity-pay mix. Future research may analyze how the link between complexity and executive compensation evolves over time. Prior studies investigated the dynamic response of executive pay to changes in firm size (Bebchuk and Grinstein, 2005a) and firm diversification (Kerr, 1985). Little is known about the dynamic effects of the other dimensions of complexity on pay. How do firms adapt their compensation strategies, when they expand internationally? How do changes in market uncertainty or the politicized environment affect executive pay? Finally, researchers may want to investigate differences in pay within the same top management team. More than 10 years ago, Finkelstein (1992) developed a model for the distribution of power within a top management team. I use parts of his model as inputs in my framework, which focuses on the power of the CEO or the top management team in total relative to non-executive directors and outside shareholders. It would be a welcome contribution to test the Finkelstein model of within-TMT power on withinTMT pay differentials. Are more powerful TMT members able to negotiate rents as opposed to their less powerful colleagues on the top management team?

Stealth compensation refers to elements of pay that are not directly observable such as the promise of beneficial pension promises, the life-long use of a company apartment, office or secretary services, or the promise of after-retirement compensation. These elements of pay are becoming more and more popular.

52

APPENDIX - QUESTIONNAIRE

179

Appendix - questionnaire
Questions are coded into measures of power and complexity. Questions 3 and 11-14 are related to the concept of complexity. They are coded into a limited number range of 1 through 4, the latter indicating a highest amount of complexity. Questions 2, 4-10 and 15-16 are questions related to managerial power. Different degrees of power are assigned different numbers as indicated below.
Question 1: What characteristics do you consider to be important for new board members? Please select the three most important criteria. Good industry knowledge Good (geographical) market knowledge Good relationship with other board members 0 0 Good relationship with the CEO Good network to strategically important people and organizations Other: ___________________ 1 0

Question 2: Which of the following activity / activities apply to the CEO with respect to the selection of new members of the board of directors? Several answers are possible. Makes suggestions for potential new board members Interviews all candidates Interviews only candidates short-listed by the board, its compensation committee and / or an executive search firm 1 1 2 Is the primary contact to the executive search firm Makes the final decision alone Makes the final decision together with the board or the nomination committee None of the above 0 1 2 2

Question 3: How complex would you estimate the operative / strategic thinking environment (thinking challenge, problem solving challenge) of your CEOs job to be? not complex somewhat complex 1 2 complex very complex 3 4

Question 4: How often have you met with the CEO outside official board meetings in 2004? Never 1-2 times 1 2 3-4 times More than 4 times 3 4

180

APPENDIX - QUESTIONNAIRE

Question 5: Which of the following consulting firms are used within your company for setting the compensation of the CEO? And who hires the respective consultancy firms / who is the primary contact? Please tick the appropriate box / boxes. Hired by / primary contact Consulting firm used Board of directors / compensation committee Corporate HR The consulting firm has other business relations with our company

CEO

Watson Wyatt New Bridge Street Mercer Spencer Stuart Obermatt Partners Towers Perrin Hewitt Associates McLagan other: ______________ Question 6: Which of the following descriptions best characterizes your relationship with the CEO? friendship acquaintance 4 3 a fair, but competitive business relation difficult 2 1

Question 7: How satisfied are you with the performance of your CEO since 2000 (or the time of appointment of the incumbent CEO)? not satisfied somewhat satisfied 1 2 satisfied very satisfied 3 4

Question 8: How high would you rate the CEOs current expertise of the firm, its industry and customers? low average 1 2 high very high 3 4

Question 9: Who writes / develops the first proposal for a compensation package for the CEO? Please tick the most appropriate box.

APPENDIX - QUESTIONNAIRE

181

the Compensation committee the entire board of directors the chief executive officer

0 0 1

the head of HR the compensation consulting firm other: ___________________

1 1 0

Question 10: Who sets the agenda for the board meetings? the CEO the chairman of the board 1 0 the CEO / chairman in case it is the same person other: __________________ 1 0

Question 11: How many executives are directly reporting to the CEO? 12 35 1 2 69 10 and more 3 4

Question 12: How high would you estimate the technical know-how required for your CEOs job? not very high Somewhat high 1 2 high very high 3 4

Question 13: How high would you estimate the management know-how required for your CEOs job? Not very high Somewhat high 1 2 high very high 3 4

Question 14: How high would you estimate the human relation skills necessary for your CEOs job? Not very high Somewhat high 1 2 high very high 3 4

182

APPENDIX - QUESTIONNAIRE

Question 15: Please tick the most appropriate description: The compensation package of the CEO is determined by negotiations between the CEO and the entire board of directors upon proposal from the compensation committee negotiations between the CEO and the entire board of directors after the board has discussed with the compensation committee (without the participation of the CEO) negotiations between the CEO and the compensation committee and subsequent discussion with the entire board 2 proposal by the board or its compensation committee and mandatory acceptance by the CEO without further negotiations 0 0 negotiations between the chairman and the CEO without negotiations with the entire board 4 1 negotiations between the CEO and the compensation committee without discussion with the entire board 3

Question 16: Have you discussed issues of executive compensation with your CEO outside official meetings of the board or the compensation committee? Yes 1 No 0

REFERENCES

183

References
ABERNETHY, M. / VAGNONI, E. (2004). Power, organization design and managerial behavior. Accounting, Organizations and Society 29, 207-225. ABOODY, D. / KASZNIK, R. (2000). CEO stock option awards and the timing of corporate voluntary disclosures. Journal of Accounting and Economics 29, 73-100. ABOWD, J. (1990). Does performance-based managerial compensation affect corporate performance? Industrial Labor Relations Review 43, 52-73. ABOWD, J. / BOGNANNO, M. (1995). International differences in executive and managerial compensation. In: FREEMAN, R. / KATZ, L. (ed.) Differences and changes in wage structures. Chicago: The University of Chicago Press, 67-103. ABOWD, J. / KAPLAN, D. (1999). Executive compensation: six questions that need answering. Journal of Economic Perspectives 13, 4, 145-168. ACHARYA, V. / JOHN, K. / SUNDARAM, R. (2000). On the optimality of resetting executive stock options. Journal of Financial Economics 57, 65-101. ADAMS, R. (2003). What do boards do? Evidence from board committee and director compensation data. Working Paper Federal Reserve Bank of New York. ADAMS, G. (2004). Power plays: A longitudinal examination of CEO/board power circulation and its impact on organizational performance. Doctoral dissertation Florida State University. ADAMS, R. / ALMEIDA, H. / FERREIRA, D. (2004). Powerful CEOs and their impact on corporate performance. Review of Financial Studies 18, 1403-1432. AGARWAL, N. (1981). Determinants of executive compensation. Industrial Relations 20, 1, 36-45. AGARWAL, J. / MANDELKER, G. (1987). Managerial incentives and corporate investment and financing decisions. Journal of Finance 42, 4, 823-837. AGARWAL, J. / WALKING, R. (1994). Executive careers and compensation surrounding takeover bids. Journal of Finance 49, 3, 985-1014. AGGRAWAL, R. / SAMWICK, A. (1999). The other side of the trade-off: the impact of risk on executive compensation. Journal of Political Economy 107, 1, 65-105. AGRAWAL, A. / KNOEBER, C. (1996). Firm performance and mechanisms to control agency problems between managers and shareholders. Journal of Financial and Quantitative Analysis 31, 377-397. AGRAWAL, A. / KNOEBER, C. (1998). Managerial compensation and the threat of takeover. Journal of Financial Economics 47, 219-239.

184

REFERENCES ALLEN, M. (1981a). Power and privilege in the large corporation: Corporate control and managerial compensation. American Journal of Sociology 86, 1112-1123. ALLEN, M. (1981b). Managerial power and tenure in the large corporation. Social Forces 60, 2, 482-494. ALMAZAN, A. / HARTZELL, J. / STARKS, L. (2005). Active institutional shareholders and costs of monitoring: Evidence from executive compensation. University of Texas Working Paper. ANDERSON, R. / BATES, T., BIZJAK, J. / LEMMON, M. (2000). Corporate governance and firm diversification. Financial Management 29, 5-22. ANDERSON, R. / BIZJAK, J. (2003). An empirical examination of the role of the CEO and the compensation committee in structuring executive pay. Journal of Banking & Finance 27, 7, 13231348. ANDJELKOVIC, A. / BOYLE, G. / MCNOE, W. (2002). Public disclosure of executive compensation: Do shareholders need to know? Pacific-Basin Finance Journal 10, 97-117. ANG, J. / LAUTERBACH, B. / SCHREIBER, B. (2002). Pay at the executive suite: How do US banks compensate their top management teams? Journal of Banking & Finance 26, 1143-1163. ANGEL, P. / FUMAS, V. (1997). The compensation of Spanish executives: A test of a managerial talent allocation model. International Journal of Industrial Organization 15, 511-532. ANTLE, R. / SMITH, A. (1986). An empirical investigation of the relative performance evaluation of corporate executives. Journal of Accounting Research 24, 1-39. ARCHARYA, V. / KOSE, J. / SUNDARAM, R. (2000). On the optimality of resetting executive stock options. Journal of Financial Economics 57, 65-101. ARTHUR, N. (2001). Board composition as the outcome of an internal bargaining process: empirical evidence. Journal of Corporate Finance 7, 307-340. BABER, W. / JANAKIRAMAN, S. / KANG, S. (1996). Investment opportunities and the structure of executive compensation. Journal of Accounting and Economics 21, 3, 297-318. BAKER, G. / HALL, K. (2004). CEO incentives and firm size. Journal of Labor Economics 22, 767798. BAKER, G. / JENSEN, M. / MURPHY, K. (1988). Compensation and incentives: Practice versus theory. Journal of Finance 43, 593-616. BALKIN, D. / MARKMAN, G. / GOMEZ-MEJIA, L. (2000). Is CEO pay in high-technology firms related to innovation? Academy of Management Journal 43, 6, 1118-1129. BALSAM, S. (2002). An introduction to executive compensation. Academic Press: San Diego. BARKEMA, H. / GOMEZ-MEJIA, L. (1998). Managerial compensation and firm performance: A general research framework. Academy of Management Journal 41, 2, 135-145.

REFERENCES

185

BARKEMA, H. / PENNINGS, J. (1998). Top management pay: Impact of overt and covert power. Organization Studies 19, 6, 975-1003. BARRON, J. / WADDELL, G. (2003a). Hard work versus good intentions: stock options as compensation. University of Oregon Working Paper. BARRON, J. / WADDELL, G. (2003b). Executive rank, pay and project selection. Journal of Financial Economics 67, 305-349. BARTLETT, C. / GHOSHAL, S. (2002). Managing across borders the transnational solution. Boston: Harvard Business School Press. BARTOV, E. / MOHANRAM, P. (2003). Private information, earnings manipulations, and executive stock option exercises. Columbia University Working Paper. BAUMOL, W. (1959). Business behavior, value and growth. New York: Macmillan. BEATTY, R. / ZAJAC, E. (1994). Managerial incentives, monitoring and risk bearing: A study of executive compensation, ownership, and board structure in initial public offerings. Administrative Science Quarterly 39, 313-335. BEBCHUCK, L. / FRIED, J. / WALKER, D. (2001). Executive compensation in America: Optimal contracting or extraction of rents? NBER Working Paper 8661. BEBCHUCK, L. / FRIED, J. (2003). Executive compensation as an agency problem. Journal of Economic Perspectives 17, 3, 71-92. BEBCHUCK, L. / FRIED, J. (2004). Pay without performance. Cambridge, Massachusetts, and London, England: Harvard University Press. BEBCHUCK, L. / GRINSTEIN, Y. (2005a). Firm expansion and CEO pay. NBER Working Paper 11886. BEBCHUCK, L. / GRINSTEIN, Y. (2005b). The growth of U.S. executive pay. Oxford Review of Economic Policy 21, 2, 283-303. BECHER, D. / CAMPBELL, T. / FRYE, M. (2005). Incentive compensation for bank directors: The impact of deregulation. Journal of Business 78, 5, 1753-1778. BECKER, G. (1975). Human capital: A theoretical and empirical analysis, with special reference to education. Chicago: University of Chicago Press. BEINER, S. / SCHMID, M. / WANZENRIED, G. (2005). Product market competition, managerial incentives, and firm valuation. University of Basel Working Paper. BELLIVEAU, M. / OREILLY, C. / WADE, J. (1996). Social capital at the top: the effects of social similarity and status on CEO compensation. Academy of Management Journal 39, 6, 1568-1593. BENTSON, G. (1985). The self-serving management hypothesis. Journal of Accounting and Economics 7, 67-84.

186

REFERENCES BENZ, M. / KUCHER, M. / STUTZER, A. (2001). Are stock options the managers blessing? Stock option compensation and institutional controls. University of Zurich Working Paper. BERGER, P. / OFEK, E. / YERMACK, D. (1997). Managerial entrenchment and capital structure decisions. Journal of Finance 52, 4, 1411-1435. BERGSTRESSER, D. / PHILIPPON, T. (2003). CEO incentives and earnings management: Evidence from the 1990s. Harvard Business School Working Paper. BERLE, A. / MEANS, G. (1932). The Modern Corporation and Public Property. New York: Macmillan. BERTRAND, A. / MULLAINATHAN, G. (2000). Agents without principals. American Economic Review 90, 2, 203-208. BERTRAND, A. / MULLAINATHAN, G. (2001). Are CEOs rewarded for luck? The ones without principals are. Quarterly Journal of Economics 116, 901-932. BETTIS, C. / BIZJAK, J. / LEMMON, M. (2005). Exercise behavior, valuation, and the incentive effects of employee stock options. Journal of Financial Economics 76, 445-470. BLACK, F. / SCHOLES, M. (1973). The pricing of options and corporate liabilities. Journal of Political Economy 81, 637-654. BLANCHARD, O. / LOPEZ-DE-SILANES, F. / SHLEIFER, A. (1994). What do firms do with cash windfalls? Journal of Financial Economics 36, 337-360. BLAZEKOVIC, M. (2004). Incentive analysis of executive compensation a case study. University of St. Gallen Master Thesis. BLISS, R. / ROSEN, R. (2001). CEO compensation and bank mergers. Journal of Financial Economics 61, 1, 107-138. BLOOM, M. (1999). The performance effects of pay dispersion on individuals and organizations. Academy of Management Journal 42, 1, 25-40. BLOOM, M. / MILKOVICH, G (1998). Relationships among risk, incentive pay, and organizational performance. Academy of Management Journal 41, 3, 283-297. BOEKER, W. (1992). Power and managerial dismissal: scapegoating at the top. Administrative Science Quarterly 37, 400-421. BOLLIGER, G. / KAST, M. (2004). Executive compensation and analyst guidance: The link between CEO pay and expectations management. National Centre of Competence in Research Working Paper No. 167. BOLTEN, P. / SCHEINKMAN, J. / XIONG, W. (2002). Executive compensation and short-termist behavior in speculative markets. NBER Working Paper 9722. BOSCHEN, M. / SMITH, R. (1995). You can pay me now and you can pay me later: The dynamic response of executive compensation to firm performance. Journal of Business 68, 4, 577-608.

REFERENCES

187

BOYD, B. (1994). Board control and CEO compensation. Strategic Management Journal 15, 335344. BRATTON, W. (2005). The academic tournament over executive compensation. California Law Review 93, 5. BRENNER, M. / SUNDARAM, R. / YERMACK, D. (2000). Altering the terms of executive stock options. Journal of Financial Economics 57, 103-128. BRICK, I. / PALMON, O. / WALD, J. (2002). CEO compensation, director compensation, and firm performance: Evidence of cronyism. Rutgers Business School Working Paper. BRICKLEY, J. / BHAGAT, S. / LEASE, R. (1985). The impact of long-range managerial compensation plans on shareholder wealth. Journal of Accounting and Economics 7, 115-129. BRICKLEY, J. / COLES, J. / JARRELL, G. (1997). Leadership structure: separating the CEO and chairman of the board. Journal of Corporate Finance 3, 189-220. BRUNELLO, G. / GRAZIANO, C. / PARIGI, B. (2001). Executive Compensation and Firm Performance in Italy. International Journal of Industrial Organization, 19, 133-161. BRYAN, S. / HWANG, L. / LILIEN, S. (2000a). CEO stock-based compensation: An empirical analysis of incentive-intensity, relative mix, and economic determinants. Journal of Business 73, 4, 661-693. BRYAN, S. / HWANG, L. / LILIEN, S. (2000b). Outside director compensation: An empirical analysis of economic determinants. City University of New York Working Paper. BRYAN, S. / HWANG, L. / LILIEN, S. (2005). CEO compensation after deregulation: The case of electric utilities. Journal of Business 78, 5, 1709-1752. BUCK, T. / BRUCE, A. / MAIN, B. / UDUENI, H. (2003). Long term incentive plans, executive pay and UK company performance. Journal of Management Studies 40, 7, 1709-1727. BUSHMAN, R. / INDJEJIKIAN, R. / SMITH, A. (1996). CEO compensation: The role of individual performance evaluation. Journal of Accounting and Economics 21, 161-193. BUSHMAN, R. / CHEN, Q. / ENGEL, E. / SMITH, A. (2004). Financial accounting information, organizational complexity and corporate governance systems. Journal of Accounting and Economics 37, 167-201. CALLAGHAN, S. / SALY, J. / SUBRAMANIAM, C. (2004). The timing of option repricing. Journal of Finance 59, 4, 1651-1677. CANELLA, A. / MONROE, M. (1997). Contrasting perspectives on strategic leaders: toward a more realistic view of top managers. Journal of Management 23, 3, 213-237. CARPENTER, J. (1998). The exercise and valuation of executive stock options. Journal of Financial Economics 48, 127-158. CARPENTER, J. (2000). Does option compensation increase managerial risk appetite? Journal of Finance 55, 5, 2311-2331.

188

REFERENCES CARPENTER, J. / REMMERS, B. (2001). Executive stock option exercises and inside information. Journal of Business 74, 4, 513-534. CARPENTER, M. / SANDERS, G. (2002). Top management team compensation: The missing link between CEO pay and firm performance? Strategic Management Journal 23, 367-375. CARPENTER, M. / SANDERS, G. (2004). The effects of top management team pay and firm internationalization on MNC performance. Journal of Management 30, 4, 509-528. CARPENTER, M. / SANDERS, G. / GREGERSEN, H. (2001). Bundling human capital with organizational context: the impact of international assignment experience on multinational firm performance and CEO pay. Academy of Management Journal 44, 493-511. CARTER, M. / LYNCH, L. (2001). An examination of executive stock option repricing. Journal of Financial Economics 61, 207-225. CATRINA, W. (2003). ABB Die verratene Vision. Zurich: Orell Fssli Verlag AG. CHANCE, D. / KUMAR, R. / TODD, R. (2000). The repricing of executive stock options. Journal of Financial Economics 57, 129-154. CHARKHAM, J. (1995). Keeping good company: A study of corporate governance in five countries. Oxford: Oxford University Press. CHAUVIN, K. / SHENOY, C. (2001). Stock price decreases prior to executive stock option grants. Journal of Corporate Finance 7, 53-76. CHEFFINS, B. (2003). Will executive pay globalize along American lines? Corporate Governance 11, 1, 8-24. CHEN, M. (2004). Executive option repricing, incentives, and retention. Journal of Finance 59, 3, 1167-1199. CHENG, S. / FIRTH, M. (2005). Ownership, corporate governance and top management pay in Hong Kong. Corporate Governance: An International Review 13, 2, 291-302. CHIDAMBARAN, N. / PRABHALA, N. (2003). Executive stock option repricing, internal governance mechanisms and management turnover. Journal of Financial Economics 69, 153-189. CHOE, C. (2001). Maturity and exercise price of executive stock options. Review of Financial Economics 10, 227-250. CISCEL, D. (1974). Determinants of executive compensation. Southern Economic Journal 40, 4, 613-617. CISCEL, D. / CARROLL, (1980). The determinants of executive salaries: An econometric survey. Review of Economics and Statistics 62, 7-13. COLES, J. / NAVEEN, D. / NAVEEN, L. (2002). Executive compensation and managerial risk-taking. Arizona State University Working Paper.

REFERENCES

189

COMBS, J. / SKILL, M. (2003). Managerialist and human capital explanations for key executives pay premiums: a contingency perspective. Academy of Management Journal 46, 1, 63-73. CONYON, M. (1997). Corporate governance and executive compensation. International Journal of Industrial Organization 15, 493-509. CONYON, M. / SADLER, G. (2001). CEO compensation, option incentives, and information disclosure. Review of Financial Economics 10, 251-277. CONYON, M. / MURPHY, K. (2000). The prince and the pauper? CEO pay in the US and the UK. Economic Journal 110, 640-671. CONYON, M. / PECK, S. / READ, L. / SADLER, G. (2000). The structure of executive compensation contracts: UK evidence. Long Range Planning 33, 478-503. CONYON, M. / PECK, S. (1998). Board control, remuneration committees, and top management compensation. Academy of Management Journal 41, 2, 146-157. CONYON, M. / PECK, S. (2004). Compensation consultants and executive pay: UK evidence. University of Pennsylvania Working Paper. CONYON, M. / PECK, S. / SADLER, G. (2001). Corporate tournaments and executive compensation: Evidence from the UK. Strategic Management Journal 22, 805-815. CONYON, M. / SCHWALBACH, J. (1999). Corporate governance, executive pay and performance in Europe. In: CARPENTER, J. / YERMACK, D, Executive Compensation and Shareholder Value Theory and Evidence. Dordrecht / Boston / London: Kluwer Academic Publishers, 13-33. CONYON, M. / SCHWALBACH, J. (2000). Executive Compensation: Evidence from the UK and Germany. Long Range Planning 33, 504-526. COOMBS, J. / GILLEY, M. (2005). Stakeholder management as a predictor of CEO compensation: Main effects and interactions with financial performance. Strategic Management Journal 26, 827840. COPPETTI, T. (2002). Die wundersame Geldvermehrung. Handelszeitung 10.07.2002: 29. CORE, J. / GUAY, W. (2001). Stock option plans for non-executive employees. Journal of Financial Economics 61, 253-287. CORE, J. / GUAY, W. / THOMAS, R. (2004). Is US CEO pay inefficient pay without performance? University of Pennsylvania Working Paper. CORE, J. / HOLTHAUSEN, R. / LARCKER, D. (1999). Corporate governance, chief executive officer compensation, and firm performance. Journal of Financial Economics 51, 371-406. COUGHLAN, A. / SCHMIDT, R. (1985). Executive compensation, management turnover, and firm performance: An empirical investigation. Journal of Accounting and Economics 7, 3-9. COWHERD, D. / LEVINE, D. (1992). Product quality and pay equity between lower-level employees and top management: An investigation of distributive justice theory. Administrative Science Quarterly 37, 302-320.

190

REFERENCES CRAWFORD, A. / EZZELL, J. / MILES, J. (1995). Bank CEO pay-performance relations and the effects of deregulation. Journal of Business 68, 2, 231-255. CRYSTAL, G. (1991). In search of excess the overcompensation of American executives. New York: Norton & Company. CUNAT, V. / GUADALUPE, M. (2005). How does product market competition shape incentive contracts? Journal of the European Economic Association 3, 5, 1058-1082. CUNY, C. / JORION, P. (1995). Valuing executive stock options with an endogenous departure decision. Journal of Accounting and Economics 20, 193-205. DAILY, C. / JOHNSON, J. (1997). Sources of CEO power and firm financial performance: A longitudinal assessment. Journal of Management 23, 97-117. DAILY, C. / JOHNSON, J. / ELLSTRAND, A. / DALTON, D. (1998). Compensation committee composition as a determinant of CEO compensation. Academy of Management Journal 41, 2, 209-220. DATTA, S. / ISKANDAR-DATTA, M. / RAMAN, K. (2001). Executive compensation and corporate acquisition decisions. Journal of Finance 56, 2299-2336. DAVENI, R. (1990). Top managerial prestige and organizational bankruptcy. Organization Science 1, 123-142. DAVID, P. / KOCHHAR, R. / LEVITAS, E. (1998). The effect of institutional investors on the level and mix of CEO compensation. Academy of Management Journal 41, 2, 200-208. DAVILA, A. / VENKATACHALAM, M. (2004). The relevance of non-financial performance measures for CEO compensation: Evidence from the airline industry. Academy of Management Journal 9, 4, 443-464. DAVIS, B. / SHELOR, R. (1995). Executive Compensation and Financial Performance in the Real Estate Industry. Journal of Real Estate Research 10, 2, 141-151. DECKOP, J. (1988). Determinants of chief executive officer compensation. Industrial and Labor Relations Review 41, 2, 215-226. DEFUSCO, R. / JOHNSON, R. / ZORN, T. (1990). The effect of executive stock option plans on stockholders and bondholders. Journal of Finance 45, 617-627. DEMIGUEL, A. / PINDADO, J. / DE LA TORRE, C. (2004). Ownership structure and firm value: New evidence from Spain. Strategic Management Journal 25, 1199-1207. DEMSETZ, H. / LEHN, K. (1985). The structure of corporate ownership: Causes and consequences. Journal of Political Economy 93, 1155-1177. DIAL, J. / MURPHY, K. (1995). Incentives, downsizing, and value creation at General Dynamics. Journal of Financial Economics 37, 261-314. DUNN, P. (2004). The impact of insider power on fraudulent financial reporting. Journal of Management 30, 3, 397-412.

REFERENCES

191

DYL, E. (1988). Corporate control and management compensation: evidence on the agency problem. Managerial and Decision Economics 91, 21-25. EATON, J. / ROSEN, H. (1983) Agency, delayed compensation and the structure of executive remuneration. Journal of Finance 38, 1489-1505. ECONOMIESUISSE (2002). Swiss Code of Best Practice for Corporate Governance. Economiesuisse. ECONOMIST (1999) Share and shares unalike. The Economist, August 7: 18-20. ECONOMIST (2005a) Executive pay: Too many turkeys. The Economist, November 24. ECONOMIST (2005b) What shareholder democracy? The Economist, March 23. ECONOMIST (2005c) The corporate savings glut. The Economist, July 7. EGELHOFF, W. (1982) Strategy and structure in multinational corporations: An informationprocessing approach. Administrative Science Quarterly 27, 435-458. EISENHARDT, K. (1988) Agency- and institutional theory explanations: The case of retail sales compensation. Academy of Management Journal 31, 3, 488-511. EISENHARDT, K. (1989) Agency Theory: An Assessment and Review. Academy of Management Review 14, 1, 57-74. EJPD (2003) Bericht und Vorentwurf fr eine nderung des Obligationenrechts. Transparenz betreffend Vergtungen an Mitglieder des Verwaltungsrats und der Geschftsleitung. ELSON, J. (2003) Whats wrong with executive compensation? Harvard Business Review January, 68-77. ELSTON, J. / GOLDBERG, L. (2003) Executive compensation and agency costs in Germany. Journal of Banking & Finance 27, 1391-1410. EMIRBAYER, J. / GOODWIN, L. (1994) Network analysis, culture, and the problem of agency. American Journal of Sociology 99, 1411-1554. ERIKSSON, T. (2000). What determines managers pay? Long Range Planning 33, 544-559. EZZAMEL, M. / WATSON, R. (1998). Market comparison earnings and the bidding-up of executive cash compensation: evidence from the United Kingdom. Academy of Management Journal 441, 221-231. FAMA, E. (1980). Agency Problems and the Theory of the Firm. Journal of Political Economy 88, 2, 288-307. FAMA, E. / JENSEN, M. (1983). Separation of ownership and control. Journal of Law and Economics 26, 2, 301-325. FATEMI, A. / DESAI, A. / KATZ, J. (2003). Wealth creation and managerial pay: MVA and EVA as determinants of executive compensation. Global Finance Journal 14, 159-179.

192

REFERENCES FEE, E. / HADLOCK, C. (2003). Raids, rewards, and reputations in the market for managerial talent. Review of Financial Studies 16, 4, 1315-1357. FEHR, K. / BISWAS, C. (2005). Die Boni sprudeln wieder. NZZ am Sonntag 30.01.2005: 43. FELTHAM, G. / WU, M. (2001). Incentive efficiency of stock versus options. Review of Accounting Studies 6, 1, 7-28. FELTON, R. (2004). A new era in corporate governance. McKinsey Quarterly 2, 28-41. FENN, G. / LIANG, N. (2001). Corporate payout policy and managerial stock incentives. Journal of Financial Economics 60, 45-72. FERRARINI, G. / MOLONEY, N. / VESPRO, C. (2003). Executive remuneration in the EU: Comparative law and practice. ECGI Working Paper Series in Law. FICH, E. / WHITE, L. (2005). Why do CEOs reciprocally sit on each others boards? Journal of Corporate Finance 11, 175-195. FINKELSTEIN, S. (1992). Power in top management teams: Dimensions, measurement, and validation. Academy of Management Journal 35, 3, 505-538. FINKELSTEIN, S. / BOYD, B. (1998). How much does the CEO matter? The role of managerial discretion in the setting of CEO compensation. Academy of Management Journal 41, 2, 179-199. FINKELSTEIN, S. / DAVENI, R. (1994). CEO duality as a double-edged sword: How boards of directors balance entrenchment avoidance and unity of command. Academy of Management Journal 37, 5, 1079-1108. FINKELSTEIN, S. / HAMBRICK, D. (1988). Chief executive compensation: a synthesis and reconciliation. Strategic Management Journal 9, 543-558. FINKELSTEIN, S. / HAMBRICK, D. (1989). Chief executive compensation: a study of the intersection of markets and political processes. Strategic Management Journal 10, 121-134. FINKELSTEIN, S. / HAMBRICK, D. (1996). Strategic Leadership: Top executives and their effects on organizations. St. Paul: West Publishing. FISHER, J. / GOVINDARAJAN, V. (1992). Profit center manager compensation: An examination of market, political and human capital factors. Strategic Management Journal 13, 205-217. FRENCH, J. / RAVEN, B. (1968). The bases of social power. In: CARTWRIGHT, D. / ZANDER, A. (eds.) Group Dynamics. New York: Harper and Row. FREY, B. (1997). On the relationship between intrinsic and extrinsic work motivation. International Journal of Industrial Organization 15, 427-439. FREY, B. / OBERHOLZER-GEE, F. (1997). On the relationship between intrinsic and extrinsic work motivation. American Economic Review 87, 4, 746-755.

REFERENCES

193

FREY, B. (2004). Etwas vage ist besser als ganz genau Die Forderung nach Lohntransparenz und Leistungslohn klingt vernnftig, ist aber oft kontraproduktiv. NZZ am Sonntag 18.01.2004: 22. FRYE, M. (2004). Equity-based compensation for employees: Firm performance and determinants. Journal of Financial Research 27, 1, 31-54. GAREN, J. (1994). Executive compensation and principal-agent theory. Journal of Political Economy 102, 6, 1175-1199. GAVER, J. / GAVER, K. (1993). Additional evidence on the association between the investment opportunity set and corporate financing, dividend, and compensation policies. Journal of Accounting and Economics 16, 125-160. GAVER, J. / GAVER, K. (1995). Compensation policy and the investment opportunity set. Financial Management 24, 1, 19-32. GELETKANYCZ, M. / BOYD, B. / FINKELSTEIN, S. (2001). The strategic value of CEO external directorate networks: Implications for CEO compensation. Strategic Management Journal 22, 889-898. GERHARDT, B. / MILKOVICH, G. (1990). Organizational differences in managerial compensation and financial performance. Academy of Management Journal 33, 4, 663-691. GERNY, D. (2004). Spitzenmanager sollen Lhne offen legen Bundesrat verabschiedet Transparenzbestimmungen. Neue Zrcher Zeitung 24.06.2004: 13. GIBBONS, R. (1998). Incentives in organizations. Journal of Economic Perspectives 12, 4, 115132. GIBBONS, R. / MURPHY, K. (1990). Relative performance evaluation for chief executive officers. Industrial and Labor Relations Review 43 (special issue), 30s-51s. GILSON, S. / VETSUYPENS, M. (1993). CEO compensation in financially distressed firms: An empirical analysis. Journal of Finance 48, 425-458. GOLDEN, B. / ZAJAC, J. (2001). When will boards influence strategy? Inclination x power = strategic change. Strategic Management Journal 22, 1087-1111. GOMEZ, P. (2004). On the discretionary power of top executives. International Studies of Management and Organization 34, 2, 37-62. GOMEZ-MEJIA, L. (1992). Structure and process of diversification, compensation strategy, and firm performance. Strategic Management Journal 13, 5, 381-397. GOMEZ-MEJIA, L. (1994). Executive compensation: A reassessment and a future research agenda. Research in Personnel and Human Resources Management 12, 161-222. GOMEZ-MEJIA, L. / BALKIN, D. (1989). Effectiveness of individual and aggregate compensation strategies. Industrial Relations 28, 431-445.

194

REFERENCES GOMEZ-MEJIA, L. / PALICH, L. (1997). Cultural diversity and the performance of multinational firms. Journal of International Business Studies 28, 2, 309-335. GOMEZ-MEJIA, L. / TOSI, H. / HINKIN, T. (1987). Managerial control, performance, and executive compensation. Academy of Management Journal 30, 1, 51-70. GOMEZ-MEJIA, L. / WISEMAN, R. (1997). Reframing Executive Compensation: An Assessment and Outlook. Journal of Management 23, 3, 291-474. GOMEZ-MEJIA, L. / LARRAZA-KINTANA, M. / MAKRI, M. (2003). The determinants of executive compensation in family-controlled public corporations. Academy of Management Journal 46, 2, 226-237. GOSH, C. / SIRMANS, C. (2005). On REIT CEO compensation: Does board structure matter? Journal of Real Estate Finance and Economics 30, 4, 397-428. GOYAL, V. / PARK, C. (2002). Board leadership structure and CEO turnover. Journal of Corporate Finance 8, 49-66. GRABKE-RUNDELL, A. / GOMEZ-MEJIA, L. (2002). Power as a determinant of CEO Compensation. Human Resource Management Review 12, 3-23. GREENE, W. (2000). Econometric analysis Fourth edition. New York: Prentice Hall International. GRAY, S. / CANNELLA, A. (1997). The role of risk in executive compensation. Journal of Management 23, 4, 517-540. GRINSTEIN, Y. / HRIBAR, P. (2004). CEO compensation and incentives: Evidence from M&A bonuses. Journal of Financial Economics 73, 119-143. GROSSMANN, S. / HART, O. (1983). An analysis of the principal-agent problem. Econometrica 51, 7-45. GUAY, W. (1999). The sensitivity of CEO wealth to equity risk: An analysis of the magnitude and determinants. Journal of Financial Economics 53, 43-71. HADLOCK, C. / LUMER, G. (1997). Compensation, turnover, and top management incentives: Historical evidence. Journal of Business 70, 2, 153-187. HALEBLIAN, J. / FINKELSTEIN, S. (1993). Top management team size, CEO dominance, and firm performance. Academy of Management Journal 36, 4, 844-859. HALL, B. (1998). The pay to performance incentives of executive stock options. NBER Working Paper. HALL, B. / LIEBMAN, J. (1998). Are CEOs really paid like bureaucrats? Quarterly Journal of Economics, 123, 3, 654-692. HALL, B. / MURPHY, K. (2000). Optimal exercise prices for executive stock options. American Economic Review 90, 2, 209-214.

REFERENCES

195

HALL, B. / MURPHY, K. (2002). Stock options for undiversified executives. Journal of Accounting and Economics 33, 3-42. HALL, B. / MURPHY, K. (2003). The trouble with stock options. Journal of Economic Perspectives 17, 3, 49-70. HALL, J. (2005). Financial performance, CEO compensation, and large-scale information technology outsourcing decisions. Journal of Management Information System 22, 1, 193-221. HALLOCK, K. (1997). Reciprocally interlocking boards of directors and executive compensation. Journal of Financial and Quantitative Analysis 32, 331-344. HALLOCK, K. (1999). Dual agency: Corporate boards with reciprocally interlocking relationships. In: CARPENTER, J. / YERMACK, D, Executive compensation and shareholder value Theory and evidence. Dordrecht / Boston / London: Kluwer Academic Publishers, 55-75. HAMBRICK, D. (1981). Environment, strategy, and power within top management teams. Administrative Science Quarterly 26, 253-276. HAMBRICK, D. / ABRAHAMSON, E. (1995). Assessing the amount of managerial discretion in different industries: A multimethod approach. Academy of Management Journal 38, 1427-1441. HAMBRICK, D. / FINKELSTEIN, S. (1995). The effects of ownership structure on conditions at the top: The case of CEO pay raises. Strategic Management Journal 16, 175-193. HAMBRICK, D. / FINKELSTEIN, S. (2005). Executive job demands: New insights for explaining strategic decisions and leader behaviors. Academy of Management Review 30, 3, 472-491. HAMBRICK, D. / FUKUTOMI, G. (1991). The seasons of a CEOs tenure. Academy of Management Review 16, 719-742. HANDELSZEITUNG. (2003). Top 2003 Die grssten Unternehmen in der Schweiz. Zurich: Handelszeitung und Finanzrundschau AG. HANDELSZEITUNG. (2004). Top 2004 Die grssten Unternehmen in der Schweiz. Zurich: Handelszeitung und Finanzrundschau AG. HANDELSZEITUNG. (2005). Top 2005 Die grssten Unternehmen in der Schweiz. Zurich: Handelszeitung und Finanzrundschau AG. HANLON, M. / RAJGOPAL, S. / SHEVLIN, T. (2003). Are executive stock options associated with future earnings? Journal of Accounting and Economics 36, 3-43. HARDY, C. / CLEGG, S. (1999). Some dare call it power. In: HARDY, C. / CLEGG, S, Studying organization: Theory and method. London: SAGE Publications, 368-388. HARFORD, J. / LI, K. (2005). Decoupling CEO wealth and firm performance. University of British Columbia Working Paper. HARRIS, D. / HELFAT, C. (1997). Specificity of CEO human capital and compensation. Strategic Management Journal 18, 895-920.

196

REFERENCES HARTZELL, J. / STARKS, L. (2003). Institutional investors and executive compensation. Journal of Finance 58, 6, 2351-2374. HAUGEN, R. / SENBET, L. (1981). Resolving the agency problems of external capital through options. Journal of Finance 36, 3, 629-647. HAYES, R. / SCHAEFER, S. (2000). Implicit contracts and the explanatory power of top executive compensation for future performance. Rand Journal of Economics 31, 273-293. HAYGROUP (2003). International comparison of executive compensation. HEALY, P. (1985). The effect of bonus schemes on accounting decisions. Journal of Accounting and Economics 7, 85-107. HECKMAN, J. (1979). Sample selection bias as a specification error. Econometrica 47, 153-161. HENDERSON, A. / FREDRICKSON, J. (1996). Information-processing demands as a determinant of CEO compensation. Academy of Management Journal 39, 3, 575-606. HENGARTNER, L. (2002). Executive Compensation in Switzerland. University of St. Gallen Master thesis. HENGARTNER, L. (2003a). Managerial ownership and firm performance: Evidence from China. University of St. Gallen Doctoral seminar paper. HENGARTNER, L. (2003b). Firm internationalization and executive compensation: Evidence from Swiss multinationals. University of St. Gallen Doctoral seminar paper. HENGARTNER, L. (2004). Audit committees, managerial compensation and audit quality: Evidence form Swiss companies. University of St. Gallen Doctoral seminar paper. HENGARTNER, L. / RICK, K. (2003). Best Practice in Swiss Corporate Governance Novartis and Swiss Re. University of St. Gallen Doctoral seminar paper. HERMALIN, B. (1994). Heterogeneity in organizational form: Why otherwise identical firms choose different incentives for their managers. Rand Journal of Economics 25, 4, 518-537. HERMALIN, B. / WEISBACH, M. (1988). The determinants of board composition. Rand Journal of Economics 19, 4, 589-606. HERMALIN, B. / WEISBACH, M. (1998). Endogenously chosen boards of directors and their monitoring of the CEO. American Economic Review 88, 1, 96-118. HERMALIN, B. / WEISBACH, M. (2001). Boards of directors as an endogenously determined institution: a survey of the economic literature. NBER Working Paper 8161. HICKSON, D. / HININGS, C. / LEE, C. SCHNECK, R. / PENNINGS, J. (1971). A strategic contingencies theory of intraorganizational power. Administrative Science Quarterly 16, 2, 216-229. HILB, M. (2005). Integrierte Corporate Governance: Ein neues Konzept der

Unternehmensfhrung und Erfolgskontrolle. Zrich: SpringerVerlag.

REFERENCES

197

HILL, C. / PHAN, P. (1991). CEO tenure as a determinant of CEO pay. Academy of Management Journal 34, 3, 707-717. HIMMELBERG, C. / HUBBARD, G. / PALIA. D. (1999). Understanding the determinants of managerial ownership and the link between ownership and performance. Journal of Financial Economics 53, 353-384. HIMMELBERG, C. / HUBBARD, G. (2000). Incentive pay and the market for CEOs: An analysis of pay-for-performance sensitivity. Columbia University Working Paper. HOFSTEDE, B. (1980). Cultures consequences: International differences in work-related values. Beverly Hills CA: Sage Publications. HOLMSTRM, B. (1979). Moral hazard and observability. Bell Journal of Economics 10, 74-91. HOLMSTRM, B. (1982). Moral hazard in teams. Bell Journal of Economics 13, 324-340. HOLMSTRM, B. / KAPLAN, S. (2003). The state of U.S. corporate governance: Whats wrong and whats right? ECGI Working Paper Series in Finance. HOLMSTRM, B. / MILGROM, P. (1987). Aggregation and linearity in the provision of intertemporal incentives. Econometrica 55, 303-328. HOLTHAUSEN, R. / LARCKER, D. / SLOAN, R. (1995a). Business unit innovation and the structure of executive compensation. Journal of Accounting and Economics 19, 279-313. HOLTHAUSEN, R. / LARCKER, D. / SLOAN, R. (1995b). Annual bonus schemes and the manipulation of earnings. Journal of Accounting and Economics 19, 29-74. HOSKISSON, R. / HITT, M. (1988). Strategic control systems and relative R&D intensity in large multiproduct firms. Strategic Management Journal 9, 605-622. HUBBARD, G. / PALIA, D. (1995). Executive pay and performance: Evidence from the U.S. banking industry. Journal of Financial Economics 39, 105-130. ITTNER, C. / LAMBERT, R. / LARCKER, D. (2003). The structure and performance consequences of equity grants to employees of new economy firms. Journal of Accounting and Economics 34, 89127. JANAKIRAMAN, S. / LAMBERT, R. / LARCKER, D. (1992). An empirical investigation of the relative performance evaluation hypothesis. Journal of Accounting Research 30, 53-69. JENSEN, M. (1986). Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review 76, 323-329. JENSEN, M. (1993). The modern industrial revolution, exit, and the failure of internal control systems. Journal of Finance 48, 3, 831-880. JENSEN, M. / MECKLING, W. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305-360.

198

REFERENCES JENSEN, M. / MURPHY, K. (1990a). Performance Pay and Top-Management Incentives. Journal of Political Economy 98, 2, 225-264. JENSEN, M. / MURPHY, K. (1990b). Its not how much you pay, but how. Harvard Business Review May-June, 138-153. JENSEN, M. / WARNER, J. (1988). Management compensation and the managerial labor market. Journal of Accounting and Economics 7, 3-9. JENSEN, M. / ZIMMERMAN, J. (1985). The distribution of power among corporate managers, shareholders, and directors. Journal of Financial Economics 20, 1-24. JENTER, D. (2002). Executive compensation, incentives, and risk. MIT Sloan School of Management Working Paper. JIN, L. (2002). CEO compensation, diversification and incentives. Journal of Financial Economics 66, 24-63. JOHN, T. / JOHN, K. (1993). Top-Management compensation and capital structure. Journal of Finance 48, 3, 949-974. JOHNSON, M. / PORTER, S. / SHACKELL, M. (1997). Stakeholder pressure and the structure of executive compensation. University of Michigan Working Paper. JOHNSON, S. / TIAN, Y. (2000a). Indexed executive stock options. Journal of Financial Economics 57, 35-64. JOHNSON, S. / TIAN, Y. (2000b). The value and incentive effects of nontraditional stock options. Journal of Financial Economics 57, 3-34. JOHNSTON, J. (2002). Tenure, promotion and executive compensation. Applied Economics 34, 993-997. JOSKOW, P. / ROSE, N. (1994). CEO pay and firm performance: dynamics, asymmetries, and alternative performance measures. NBER Working Paper 4976. JOSKOW, P. / ROSE, N. / WOLFRAM, C. (1996). Political constraints on executive compensation: Evidence from the electric utility industry. Rand Journal of Economics 27, 1, 165-182. KADAN, O. / SWINKELS. J. (2005). Stocks or options? Moral hazard, firm viability and the design of compensation contracts. Washington University Working Paper. KATO, T. (1997). Chief executive compensation and corporate groups in Japan: New evidence from micro data. International Journal of Industrial Organization 15, 455-467. KE, B. / PETRONI, K. / SAFIEDDINE, A. (1999). Ownership concentration and sensitivity of executive pay to accounting performance measures: Evidence from publicly and privately-held insurance companies. Journal of Accounting and Economics 28, 185-209. KELLER, H. (2003). The determinants and effects of interlocking directorships and board composition: an empirical analysis of corporate governance in Switzerland. University of St. Gallen Dissertation.

REFERENCES

199

KERR, S. (1975). On the folly of rewarding A, while hoping for B. Academy of Management Journal 18, 769-783. KERR, J. (1985). Diversification strategies and managerial rewards: An empirical study. Academy of Management Journal 28, 1, 155-179. KERR, J. / KREN, L. (1992). The effect of relative decision monitoring on chief executive compensation. Academy of Management Journal 35, 2, 370-397. KHAN, R. / DHARWADKAR, R. / BRANDES, P. (2005). Institutional ownership and CEO compensation: a longitudinal examination. Journal of Business Research 58, 1078-1088. KIDDER, D. / BUCHHOLTZ, A. (2002). Can excess bring success? CEO compensation and the psychological contract. Human Resource Management Review 12, 599-617. KOLE, S. (1997). The complexity of compensation contracts. Journal of Financial Economics 43, 79-104. KORNHAUSER, L. / NAIR, V. / DAINES, R. (2005). The good, the bad and the lucky: CEO pay and skill. New York University School of Law Working Paper. KOSTIUK, P. (1990). Firm size and executive compensation. Journal of Human Resources, 90-105. KRAUER, A. (2004). Es braucht den Sinn fr das Mass. Neue Zrcher Zeitung 03.03.2004: 23. KRAMARSCH, M. (2000). Aktienbasierte Managementvergtung. Stuttgart: Schffer-Poeschel Verlag,. KROLL, M. / SIMMONS, S. / WRIGHT, P. (1990). Determinants of chief executive officer compensation following major acquisitions. Journal of Business Research 20, 349-366. LAMBERT, R. / LARCKER, D. (1987). An analysis of the use of accounting and market measures of performance in executive compensation contracts. Journal of Accounting Research 25, 85-125. LAMBERT, R. / LANEN, W. / LARCKER, D. (1989). Executive stock option plans and corporate dividend policy. Journal of Financial and Quantitative Analysis 24, 409-425. LAMBERT, R. / LARCKER, D. / WEIGELT, K. (1993). The structure of organizational incentives. Administrative Science Quarterly 38, 438-461. LAMBERT, R. / LARCKER, D. / WEIGELT, K. (1991). How sensitive is executive compensation to organizational size? Strategic Management Journal 12, 395-402. LAMBERT, R. / LARCKER, D. / VERRECHIA, R. (1991). Portfolio considerations in valuing executive compensation. Journal of Accounting Research 29, 129-149. LARCKER, D. (1983). The association between performance plan adoption and corporate capital investment. Journal of Accounting and Economics 5, 3, 3-30. LAZEAR, E. / ROSEN, S. (1981). Rand-order tournaments as optimum labor contracts. Journal of Political Economy 89, 841-864.

200

REFERENCES LEONARD, J. (1990). Executive pay and firm performance. Industrial and Labor Relations Review 43, 13-29. LEWELLEN, W. / HUNTSMAN, B. (1970). Managerial pay and corporate performance. American Economic Review 60, 710-720. LEWELLEN, W. / LODERER, C. / MARTIN, K. (1987). Executive compensation and executive incentive problems. Journal of Accounting and Economics 9, 287-310. LIE, E. (2005). On the timing of CEO stock option awards. Management Science 51, 5, 802-812. LIPPERT, R. / MORE, W. (1994). Compensation contracts of chief executive officers: Determinants of pay-performance sensitivity. Journal of Financial Research 57, 3, 321-332. MAIN, B. (1991). Top executive pay and performance. Managerial and Decision Economics 12, 3, 219-229. MAIN, B. (1999). The rise and fall of executive share options in Britain. In: CARPENTER, J. / YERMACK, D, Executive compensation and shareholder value Theory and evidence. Dordrecht / Boston / London: Kluwer Academic Publishers, 83-112. MAIN, B. / BRUCE, A. / BUCK, T. (1996). Total board remuneration and company performance. The Economic Journal 11, 606-628. MAIN, B. / OREILLY, C. / WADE, J. (1993). Top executive pay: Tournament or Teamwork? Journal of Labor Economics 106, 1627-1644. MAIN, B. / OREILLY, C. / WADE, J. (1995). The CEO, the board of directors and executive compensation: Economic and psychological perspectives. Industrial and Corporate Change 4, 134. MASSON, R. (1971). Executive motivations, earnings, and consequent equity performance. Journal of Political Economy 79, 1278-1292. MAYERS, D. / SMITH, C. (1992). Executive Compensation in the life insurance industry. Journal of Business 65, 1, 51-74. MCCONNELL, J. / SERVAES, H. (1990). Additional evidence on equity ownership and corporate value. Journal of Financial Economics 27, 595-618. MCEACHERN, W. (1975). Managerial control and performance. Lexington, MA: D.C. Heath & Co. MCGUIRE, J. / CHIU, J. / ELBING, A. (1962). Executive income, sales, and profits. American Economic Review 52, 753-761. MCGUIRE, J. / DOW, S. / ARGHEYD, K. (2003). CEO incentives and corporate social performance. Journal of Business Ethics 45, 341-359. MEHRAN, H. (1995). Executive compensation structure, ownership and firm performance. Journal of Financial Economics 38, 2, 163-184.

REFERENCES

201

MEHRAN, H. / NOGLER, G. / SCHWARTZ, K.(1998). CEO incentive plans and corporate liquidation policy. Journal of Financial Economics 50, 319-350. MERTON, R. (1973). Theory of rational option pricing. Bell Journal of Economics 4, 141-183. MEULBROEK, L. (2001). The efficiency of equity-linked compensation: understanding the full cost of rewarding executive stock options. Financial Management 30, 2, 5-44. MILBOURN, T. (2003). CEO reputation and stock-based compensation. Journal of Financial Economics 68, 233-262. MILGROM, P. (1988). Employment contracts, influence activities and efficient organization design. Journal of Political Economy 96, 42-60. MILGROM, P. / ROBERTS, J. (1992). Economics, Organization, and Management. Englewoods Cliffs, NJ: Prentice-Hall Inc. MILKOVICH, G. / GERHARDT, B. / HANNON, J. (1991). The effects of research and development intensity on managerial compensation in large organizations. Journal of High Technology Management Research 2, 1, 133-150. MILLER, D. (1995). CEO salary increases may be rational after all: Referents and contracts in CEO pay. Academy of Management Journal 38, 5, 1361-1385. MILLER, D. / WISEMAN, R. / GOMEZ-MEJIA, L. (2002). The fit between CEO compensation design and firm risk. Academy of Management Journal 45, 4, 745-756. MINTZBERG, H. (1973). The nature of managerial work. New York: Harper & Row. MISHRA, C. / MCCONAUGHY, D. / GOBELI, D. (2000). Effectiveness of CEO pay-for-performance. Review of Financial Economics 9, 1-13. MORCK, R. / SHLEIFER, A. / VISHNY, R. (1988). Management ownership and market valuation An empirical analysis. Journal of Financial Economics 20, 293-315. MUELLER, D. / YUN, S. (1997). Managerial discretion and managerial compensation. Journal of Industrial Organization 15, 441-454. MURPHY, K. (1985). Corporate performance and managerial remuneration: An empirical analysis. Journal of Accounting and Economics 7, 11-42. MURPHY, K. (1997). Executive compensation and the Modern Industrial Revolution. International Journal of Industrial Organization 15, 417-425. MURPHY, K. (1999). Executive compensation. In: ASHFELDER, O. / CARD, D. (eds.), Handbook of Labor Economics, Vol. 3B. Amsterdam: North-Holland, 2485-2563. MURPHY, K. (2003). Stock-based pay in new economy firms. Journal of Accounting and Economics 34, 129-147. MURPHY, K. / ZABOJNIK, J. (2003). Managerial capital and the market for CEOs. University of Southern California Working Paper.

202

REFERENCES MURPHY, K. / ZABOJNIK, J. (2004). CEO pay and appointments: A market-based explanation for recent trends. American Economic Review 94, 2, 192-194. NAGAR, V. / NANDA, D. / WYSOCKI, P. (2003). Discretionary disclosure and stock-based incentives. Journal of Accounting and Economics 34, 283-309. NEWMAN, H. / MOZES, H. (1999). Does the composition of the compensation committee influence compensation practices? Financial Management 28, 3, 41-53. NIEDERBERGER, W. (2005). Stark belagerte Gewerkschaften. Tagesanzeiger 24.12.2005: 24. NORINGTON, B. (2004). Breaking the bank. The Sydney Morning Herald 03.04.2004: 36. OCASIO, W. (1994). Political dynamics and the circulation of power: CEO succession in U.S. industrial corporations, 1960-1990. Administrative Science Quarterly 39, 285-312. OREILLY, C. / MAIN, B. / CRYSTAL, G. (1988). CEO Compensation as tournament and social comparison: A tale of two theories. Administrative Science Quarterly 33, 257-274. OXELHEIM, L. / RANDY, T. (2005). The Anglo-American financial influence on CEO compensation in non-Anglo-American firms. Journal of International Business Studies 36, 470483. PENG, A. / RELL, T. (2005). Executive pay, earnings manipulation and shareholder litigation. Princeton University Working Paper. PETTIGREW, A. / MCNULTY, T. (1995). Power and influences in and around the boardroom. Human Relations 48, 845-873. PETTIGREW, A. / MCNULTY, T. (1998). Sources and uses of power in the boardroom. European Journal of Work and Organizational Psychology 7, 2, 197-214. PFEFFER, J. (1980). Power in organizations. Cambridge: Ballinger Publishing. PFIFFNER, F. (2005). Silber fr Schweizer Verwaltungsrte. NZZ am Sonntag 17.04.2005: 59. PFYFFER, H. /RUUD, F. / BODENMANN, J. (2001). Corporate Governance in der Schweiz Verwaltungsrat zwischen Verantwortung und Haftung. KPMG Schweiz and University of St. Gallen. PEARCE, J / ZAHRA, S. (1991). The relative power of CEOs and boards of directors. Strategic Management Journal 12, 135-153. PETTIGREW, A. (1973). The politics of organizational decision making. London: Tavistock. PETTIGREW, A. (1992). On studying managerial elites. Strategic Management Journal 13, 163182. POLLOCK, T / FISCHER, H. / WADE, J. (2002). The role of power and politics in the repricing of executive options. Academy of Management Journal 45, 6, 1172-1182. PORAC, J / WADE, J. / POLLOCK, T. (1999). Industry categories and the politics of the comparable firm in CEO compensation. Administrative Science Quarterly 44, 1, 112-144.

REFERENCES

203

PRENDERGAST, C. (2002). The tenuous trade-off between risk and incentives. Journal of Political Economy 110, 5, 1071-1102. PROWSE, S. (1992). Corporate governance in an international perspective: A survey of corporate control mechanisms among large firms in the U.S., U.K, Japan and Germany. Journal of Finance 47, 1-61. RAJAGOPALAN, S. / PRESCOTT, T. (1990). Determinants of top management compensation: explaining the impact of economic, behavioural, and strategic constructs and the moderating effects of industry. Journal of Management 16, 3, 515-538. RAJGOPLAN, S. / SHEVLIN, T. (2002). Empirical evidence on the relation between stock option compensation and risk taking. Journal of Accounting and Economics 33, 145-171. REDDING, G. / FRIES, A. (1995). ABB The battle for the Pacific. Long Range Planning 28, 1, 9294. REDIKER, K. / SETH, A. (1995). Board of directors and substitution effects of alternative governance mechanisms. Strategic Management Journal 16, 85-99. RICHARD, J. (2000). Global executive compensation: A look at the future. Compensation and Benefits Review 32, 3, 35-38. ROBERTS, D. (1959). A general theory of executive compensation based on statistically tested propositions. Quarterly Journal of Economics 70, 270-294. ROSEN, S. (1982). Authority, control and the distribution of earnings. Bell Journal of Economics 13, 311-323. ROSE, N. / SHEPARD, A. (1997). Firm diversification and CEO compensation: managerial ability or executive entrenchment? Rand Journal of Economics 28, 3, 489-514. ROTH, K. (1995). Managing international interdependence: CEO characteristics in a resourcebased framework. Academy of Management Journal 38, 200-231. ROTH, K. / ODONNELL, S. (1996). Foreign subsidiary compensation strategy: An agency theory perspective. Academy of Management Journal 39, 3, 678-703. ROULSTONE, D. (2003). The Relation between insider-trading restrictions and executive compensation. Journal of Accounting Research 41, 3, 525-551. RYAN, H. / WIGGINS, R. (2001). The influence of firm- and manager-specific characteristics on the structure of executive compensation. Journal of Corporate Finance 7, 101-123. RYAN, H. / WIGGINS, R. (2004). Who is in whose pocket? Director compensation, board independence, and barriers to effective monitoring. Journal of Financial Economics 73, 497-524. SALANCIK, G. / PFEFFER, R. (1980). Effects of ownership and performance on executive tenure in US corporations. Academy of Management Journal 23, 653-664. SALY, P. (1994). Repricing of executive stock options in a down market. Journal of Accounting and Economics 18, 325-356.

204

REFERENCES SANDERS, G. (2001a). Behavioral responses of CEOs to stock ownership and stock option pay. Academy of Management Journal 44, 3, 477-492. SANDERS, G. (2001b). Incentive alignment, CEO pay level, and firm performance: A case of heads I win, tails you lose? Human Resource Management 40, 2, 159-170. SANDERS, G. / CARPENTER, M. (1998). Internationalization and firm governance: The roles of CEO compensation, top team composition and board structure. Academy of Management Journal 41, 2, 158-178. SCHAEFER, S. (1998). The dependence of pay-performance sensitivity on the size of the firm. Review of Economics and Statistics 80, 3, 436-443. SCHILDKNECHT, K. (2004). Corporate Governance Das subtile Spiel um Geld und Macht. Zrich: Verlag Neue Zrcher Zeitung. SCHMID, F. (1997). Vorstandsbezge, Aufsichtsratsvergtung und Aktionrsstruktur. Zeitschrift fr Betriebswirtschaft 2, 203-217. SCHMIDT, D. / FOWLER, K. (1990). Post-acquisition financial performance and executive compensation. Strategic Management Journal 11, 7, 559-570. SCHTZ, D. (2005). Gierige Chefs Warum kein Manager zwanzig Millionen wert ist. Zurich: Orell-Fssli-Verlag. SCHWALBACH, J. / GRASSHOFF, U. (1997). Managervergtung und Unternehmenserfolg. Zeitschrift fr Betriebswirtschaft 2, 203-217. SHEN, W. (2003). The dynamics of the CEO-board relationship: An evolutionary perspective. Academy of Management Review 28, 3, 466-476. SHEN, W. (2005). Improve board effectiveness: The need for incentives. British Journal of Management 16, s81-s89. SHEN, W. / CANNELLA, A. (2002). Power dynamics within top management and their impacts on CEO dismissal followed by inside succession. Academy of Management Journal 45, 6, 11951205. SHIVDASANI, A. / YERMACK, D. (1999). CEO involvement in the selection of new board members: an empirical analysis. Journal of Finance 54, 5, 1829-1853. SHILLER, R. (2003). From efficient markets theory to behavioral finance. Journal of Economic Perspectives 17, 1, 83-104. SHLEIFER, A. / VISHNY, R. (1986). Large shareholders and corporate control. Journal of Political Economy 94, 653-664. SHLEIFER, A. / VISHNY, R. (1989). Management entrenchment the case of manager-specific investments. Journal of Financial Economics 25, 123-139.

REFERENCES

205

SIEGEL, P. / BROCKNER, J. (2005). Individual and organizational consequences of CEO claimed handicapping: Whats good for the CEO may not be good for the firm. Organizational Behavior and Human Decision Processes 96, 1-22. SIEGEL, P. / HAMBRICK, D. (2005). Pay disparities within top management groups: Evidence of harmful effect on performance of high-technology firms. Organization Science 16, 3, 259-274. SIMON, H. (1967). The compensation of executives. Sociometry, 32-35. SINGH, H. / HARIANTO, F. (1989). Top management tenure, corporate ownership structure, and the magnitude of golden parachutes. Strategic Management Journal 10, 143-156. SLOAN, R. (1993). Accounting earnings and top executive compensation. Journal of Accounting and Economics 16, 55-100. SMITH, C. / WATTS, R. (1992). The investment opportunity set and corporate financing, dividend, and compensation policies. Journal of Financial Economics 32, 263-292. SPECK, K. (2003). Der Erfolg zhlt weniger. Handelszeitung, 17.09.2003: 12-13. SPECK, K. (2005). 25 im Millionen-Club. Handelszeitung, 18.05.2005: 8. SRIDHARAN, U. (1996). CEO influence and executive compensation. Financial Review 31, 1, 5166. STAW, B. (1973). Knee-deep in the big muddy: A study of escalating commitment to a chosen course of action. Organizational Behavior and Human Performance 16, 27-44. STEIN, J. (1997). Internal capital markets and the competition for corporate resources. Journal of Finance 52, 111-133. STERN, H. / PECK, S. (2003). Executive Compensation Switzerland Trends in Vergtungsstrukturen fr Fhrungskrfte. Zurich: Obermatt Partners. STYLES, P. / TAYLOR, B. (2003). Boards at work How directors view their roles and responsiblilities. Oxford: Oxford University Press. SULLIVAN, D. (1994). Measuring the degree of internationalization of a firm. Journal of International Business Studies 25, 2, 325-342. SWX SWISS STOCK EXCHANGE. (2002). Richtlinie betreffend Informationen zur Corporate Governance. Zurich: SWX Swiss Stock Exchange. SWX SWISS STOCK EXCHANGE. (2003). Studie zur praktischen Umsetzung der Corporate Governance Richtlinien. Zurich: SWX Swiss Stock Exchange. TEHRANIAN, H. / TRAVLOS, N. / WAEGELEIN, J.(1987). The effect of long-term performance plans on corporate sell-off induced abnormal returns. Journal of Finance 42, 933-942. TIAN, Y. (2004). Too much of a good incentive? The case of executive stock options. Journal of Banking & Finance 28, 1225-1245.

206

REFERENCES TOSI, H. / GOMEZ-MEJIA, L. (1989). The decoupling of CEO pay and performance: an agency theory perspective. Administrative Science Quarterly 34, 169-189. TOSI, H. / GRECKHAMER, T. (2004). Culture and CEO compensation. Organization Science 15, 6, 657-670. TOSI, H. / MISANGYI, V. / FANELLI, A. / WALDMANN, D. / YAMMARINO, F. (2004). CEO charisma, compensation, and firm performance. Leadership Quarterly 15, 405-420. TOSI, H. / WERNER, S. / KATZ, J. / GOMEZ-MEJIA, L. (2000). A meta-analysis of executive compensation studies. Journal of Management 26, 301-339. TOYNE, M. / MILLAR, J. / DIXON, B. (2000). The relation between CEO control and the risk of CEO compensation. Journal of Corporate Finance 6, 291-306. TUSHMAN, M. / NADLER, D. (1978). Information processing as an integrating concept in organizational design. Academy of Management Review 3, 3, 613-624. UNGSON, G. / STEERS, R. (1984). Motivation and politics in executive compensation. Academy of Management Review 9, 2, 313-323. USEEM, M. (1979). The social organization of the American business elite and participation of corporate directors in the governance of American institutes. American Sociological Review 44, 553-572. USEEM, J. (2003). Have they no shame? Fortune 147, 57. VERBEEK, M. (2004). A guide to modern econometrics 2nd edition. Chichester: John Wiley & Sons Ltd. WADE, J. / OREILLY, C., CHANDRATAT, I. (1990). Golden parachutes: CEOs and the exercise of social influence. Administrative Science Quarterly 35, 587-603. WERNER, S. / TOSI, H. (1995). Other peoples money: The effect of ownership on compensation strategy and managerial pay. Academy of Management Journal 38, 6, 1672-1691. WESTPHAL, J. (1998). Board games: How CEOs adapt to increases in structural board independence from management. Administrative Science Quarterly 43, 511-537. WESTPHAL, J. (1999). Collaboration in the boardroom: Behavioral and performance consequences of CEO-board social ties. Academy of Management Journal 42, 1, 7-24. WESTPHAL, J. / ZAJAC, E. (1994). Substance and symbolism in CEOs long-term incentive plans. Administrative Science Quarterly 39, 367-390. WHITE, H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica 48, 817-838. WITTWER, J. (2005a). Die Lohnschere ffnet sich weiter. Tagesanzeiger 21.06.2005: 27. WITTWER, J. (2005b). Das Netz ber den Konzernen. Tagesanzeiger 14.05.2005: 27. WITTWER, J. (2005c). Topmanager sollen ihre Lhne offen legen. Tagesanzeiger 02.03.2005: 25.

REFERENCES WUFFLI, P. (2004). Dauerbrenner Managerlhne. Neue Zrcher Zeitung 03.03.2004: 23.

207

YERMACK, D. (1995). Do corporations award CEO stock options effectively? Journal of Financial Economics 39, 237-269. YERMACK, D. (1996). Higher market valuation of companies with a small board of directors. Journal of Financial Economics 40, 185-211. YERMACK, D. (1997). Good timing: CEO stock option awards and company news announcements. Journal of Finance 52, 449-476. YERMACK, D. (2004). Remuneration, retention, and reputation incentives for outside directors. Journal of Finance 59, 2281-2308. ZAJAC, E. (1988). Interlocking directorates as an interorganizational strategy. Academy of Management Journal 31, 428-438. ZAJAC, E. (1990). CEO Selection, compensation, and firm performance. Strategic Management Journal 11, 217-230. ZAJAC, E. / WESTPHAL, J. (1994). The costs and benefits of managerial incentives and monitoring in large U.S. corporations: When is more not better? Strategic Management Journal 15, 121-142. ZAJAC, E. / WESTPHAL, J. (1995). Accounting for the explanations of CEO compensation: Substance and symbolism. Administrative Science Quarterly 40, 283-308. ZAJAC, E. / WESTPHAL, J. (1996a). Who shall succeed? How CEO/Board preferences and power affect the choice of new CEOs. Academy of Management J0urnal 39, 1, 64-90. ZAJAC, E. / WESTPHAL, J. (1996b). Director reputation, CEO-board power, and the dynamics of board interlocks. Administrative Science Quarterly 41, 3, 507-529.

Curriculum Vitae
Education
2002 2006 University of St. Gallen Doctoral studies in business administration 1997 2002 University of St. Gallen Master of Science in International Management HSG (MIM-HSG) 2000 2002 Norwegian School of Economics and Business Administration Master of International Business (MIB-NHH) 2001 Stockholm School of Economics Exchange semester

Professional Experience
2004 today ABB Switzerland Assistant to the Executive Committee 2003 2004 ABB Group Global Graduate Program in Finance & Controlling 2000 HILTI France Internship in Marketing