1st proof

LABOR MARKETS

Why employment performance difers

It is not only labor markets but also product market regulations that chiey determine a nations employment performance

William W. Lewis, Ren Limacher, and Michael D. Longman

as the gravest economic problem for most industrial countries. In Europe, it has climbed steadily for the past 25 years to reach a postwar high (Exhibit 1). In the United States, it has been cyclical, with an average below that of Europe for the past decade. Japan has so far experienced low unemployment, although the early stages of economic reform and the recent rise in the numbers of the jobless suggest there is a risk that it will grow in the future.
NEMPLOYMENT IS WIDELY VIEWED

Much research has attempted to determine the causes of unemployment. Most of it has focused on macro analysis of the labor market. Though its ndings are based on sound economic principles, the evidence available to test them is not conclusive. Despite references to the possible impact on unemployment of the markets for goods and services (product market) and for capital, no work to date has been able fully to determine this impact. To test conclusions about the labor markets efect on unemployment, and to investigate the role played by product and capital markets, the McKinsey Global Institute studied employment performance in France, Germany, Italy, Spain, Japan, and the United States from 1980 to the early 1990s.* We conducted analysis both at the level of the national economy and within seven industries: automotive, computers, furniture, banking, general merchandise retailing, lm /TV/video, and construction. (See pp. 915 for case summaries.) Our principal ndings are: Japan and the United States have lower unemployment than Europe because they have created jobs in the market part of their economy, whereas European countries have lost jobs in theirs (Exhibits 2 and 3).
Authors note: We would like to recognize the valuable contributions of Michele Appendino, Martin Baily, Toms Calleja, Heino Fassbender, Thomas Gerstner, Kathryn Huang, Mineki Toyama, and Eric Zitzewitz to the results of this report. Employment Performance, McKinsey Global Institute, Washington, DC, November 1994.

THE McKINSEY QUARTERLY 1994 NUMBER 4

1st proof

WHY EMPLOYMENT PERFORMANCE DIFFERS

Exhibit 1

Standardized unemployment rates, 197093

Percent of total labor force
12

European Community
10

United States

Product market restrictions were as important, if not more so, than workforce rigidity in explaining why job creation in Europe was below that in the United States, especially in highgrowth service industries. The creation of large numbers of service sector jobs in the United States has harmed neither job quality nor wages.

Japan
2 0 1970 Source: OECD

1975

1980

1985

1990

1993

Exhibit 2

Employment performance, 198090

at n Ja
10 0 10 20 30

Ita l G y er Sp m ai an n y

ce

Fr an

Market economy Nonmarket economy Total

50 40 30 20 40 50 60 * Adjusted for growth in the working age population Including government, private and public education, private and public health care, etc Source: OECD; National household surveys; McKinsey analysis

ni te

pa

St

es
Exhibit 3

Net jobs created per thousand working age population*

Market economy, 198090

at pa n Ja
10 0 10 20 30

Ita l G y e Sp rma ai ny n

ce

Fr an

Market economy Agriculture Manufacturing Construction Market services

50 40 30 20 40 50 60 * Adjusted for growth in the working age population Including mining and utilities Source: OECD; National household surveys; McKinsey analysis

THE McKINSEY QUARTERLY 1994 NUMBER 4

ni te

St

es

Net jobs created per thousand working age population*

1st proof

WHY EMPLOYMENT PERFORMANCE DIFFERS

Exhibit 4

Employment growth by occupation, 198090*

er Ja ma p n Fr an y an ce
Net jobs created per thousand working age population Professional, technical, administrative, and managerial Clerical, sales, and service Production, transportation, and laborers
40 30 20 10 0 10 20 30 40 50 * France 198291; Germany 198091; Japan 197990; United States 197990; all countries extrapolated to 198090. Figures do not add up to total employment performance (Exhibit 2) since agricultural workers are excluded Adjusted for growth in the working age population Source: OECD; McKinsey analysis

The United States has created more high-skill jobs than Germany and France (Exhibit 4). Those countries have improved their job skill mix primarily by destroying low-skill jobs. The wage distribution of service jobs in the US is almost identical to that of manufacturing (Exhibit 5). Our ndings for each country are summarized in the insert on pp. 67.
Exhibit 5

US wage distribution, 1992

Total industry workforce
4.0

Percent in each $25 wage interval

3.5 3.0 2.5 2.0

The causes

Manufacturing

1.5 Diferences in employment perServices 1.0 formance ultimately stem from diferences in the rates at which 0.5 industrial economies evolve. All 0.0 200 400 600 800 1,000 1,200 economies are driven by natural Weekly wage ($) evolutionary forces. As agriculture Source: Cleveland Federal Reserve Bank; US Department of Labor; US Bureau of Labor Statistics developed beyond the point of providing for self-suciency in food, productivity increases released labor for employment in manufacturing and services. As a result, jobs in these sectors rose as a fraction of total employment in all industrial countries during the rst half of the twentieth century. In the second half, manufacturing employment peaked in industrial countries. Service employment, however, continued to rise.

U n St i t e d at es
1,400

Today, productivity gains from innovative new products and processes continue to drive economic evolution and turnover in the labor market. Contrary to popular wisdom, productivity gains do not generally bring mounting unemployment and a declining economy. Rather, they provide more income for workers in old jobs, and free other workers to ll new jobs created by entrepreneurs. These entrepreneurs, in turn, produce

THE McKINSEY QUARTERLY 1994 NUMBER 4

1st proof

PERFORMANCE BY COUNTRY
France had weak job growth in both manufacturing and services during the 1980s. Restructuring occurred in manufacturing and in many traditional services because of slow growth in demand and the disadvantaged productivity/wage position of many companies. Numerous product market restrictions inhibited new job creation in services and construction. High levels of benets and payroll taxes further reduced employment opportunities in low-wage industries such as retailing. Italy experienced strong employment growth in the service sector, but not enough to compensate for heavy losses in manufacturing and agriculture. Employment declined in all branches of manufacturing as high productivity growth was unmatched by output growth. The pace of restructuring was more rapid here than elsewhere because of the initial low productivity of many Italian manufacturers. Companies that had historically faced little competition in Italys oligopolistic markets began to feel the pressure to become more cost-competitive. Service employment grew, but largely in traditional products rather than in such emerging areas as mortgages, securities, new retailing formats, and cable networks. Spain began the 1980s at an earlier stage of economic evolution than most other countries in Europe. Large numbers of people were still occupied in agriculture and traditional manufacturing industries. During the decade, farm jobs were lost as productivity increased. Manufacturing employment fell too, despite the initially low penetration of investment goods. High real interest rates dampened demand growth in investment goods, while low productivity levels, rising wages, and a real appreciation of the peseta meant that the increase in demand that did occur was met by imports. Spain did improve service sector and construction employment, but numerous product market barriers prevented this growth from being sufcient to absorb the outow from manufacturing and agriculture. Germany began the 1980s with a high level of manufacturing productivity relative to the rest of Europe. This helped its manufacturers gain market share in Europe and postpone major restructuring during the decade. Consequently, its small loss of manufacturing

innovative new products and services, which workers with increased incomes want to buy. The supply/demand balance in the economy is not automatic. An imbalance in one direction can generate ination; in the other, it can frustrate growth and usher in recession. Evidence shows that slow productivity growth does not relieve recession, any more than rapid productivity growth deepens it. We would expect to see a pattern of primarily productivity improvement in the older industries (agriculture and manufacturing), and both productivity improvement and output growth in the newer industries (services). Our results broadly bear this out. Governments are the chief reason why the process of economic evolution is working better in some countries than in others. Their actions to manage and control the evolution of their economies are the most important determinants of diferences in employment performance. Most of these actions regulate individual service industries through restrictions on output

THE McKINSEY QUARTERLY 1994 NUMBER 4

1st proof

employment allowed Germany to achieve a better job creation record than the rest of Europe, despite having the slowest growth in service sector employment. Product market restrictions in services and construction and high total labor costs have held back job creation in many of the emerging sectors of the economy. That Germanys performance could not be sustained became obvious at the beginning of the 1990s, when heavy layoffs in manufacturing could not be absorbed by the service industries, leading to an increase in unemployment. Japan experienced rising employment in both manufacturing and services during the 1980s, though the sustainability of many current jobs remains in question since the country avoided much of the restructuring that took place in the United States and Europe. Such highproductivity industries as auto and machine tools were able to increase employment while extending their world lead in productivity thanks to the extraordinary growth in demand for investment goods during the bubble economy of the late 1980s. Lower-productivity industries, such as retailing and food processing, were often protected from foreign

competition and prevented from evolving rapidly, and thus remained uncompetitive. Employment dislocations look set to occur in both types of industry in the future. Demand for consumer durables and investment goods will slow now that product penetration has approached Western levels and industries are experiencing overcapacity. Consumer pressure for lower prices will force productivity to rise in the lagging industries or bring in greater numbers of imports. The United States lost a few manufacturing jobs as a direct result of its mounting trade decit, but it lost even more to the corporate restructuring that took place in response to intensied Japanese competition. Its overall employment performance did not suffer, however, thanks to the creation of vast numbers of service sector and construction jobs in an environment of relatively few restrictions in product markets and some helpful regulation. About half of the jobs created were high-skill jobs, but low unemployment benets and low payroll taxes also made it possible to expand employment in low-skill, low-wage industries such as retailing.

in the product market.* These restrictions hurt productivity, employment, or both together. If wages are free to fall, then employment can be maintained, but the loss in productivity keeps real wages low. If wages are maintained, then restrictions show up in unemployment. Since product market restrictions are specic to individual industries, aggregate analysis is inadequate for identifying them and determining their impact. We have had to analyze them at industry level. This case work constitutes the main efort of our study, and diferentiates it from most other research on the unemployment problem.

Manufacturing declines
No industrial country should look to manufacturing for net job creation. This is clearly evident in the computer industry: job creation from computer innovation now stems from the service dimension of applying the innovation (sotware and distribution), rather than from the manufacturing
The manufacturing sector is less amenable to government inuence than are service industries because strong inuence on manufacturers is exerted by market pressure from trade and foreign direct investment.

THE McKINSEY QUARTERLY 1994 NUMBER 4

1st proof

WHY EMPLOYMENT PERFORMANCE DIFFERS

of the computer itself. The impact of hardware manufacturing on employment performance is almost negligible. Even in a high-tech industry, then, virtually all net job growth derives from services. Our analysis identies two important reasons for Japans superior employment performance in manufacturing. The rst was its efort to catch up with the United States in plant and equipment per employee. The second was its overinvestment in capacity in heavy manufacturing and electronics during the No industrial country bubble economy of the late 1980s. Almost half of Japans growth in GDP came from investment, rather than consumption. Since the production of capital goods like machine tools is concentrated in the manufacturing sector, it is not surprising that Japans manufacturing output rose so strongly. Of the growth in manufacturing value-added, 71 percent came from machinery, equipment, and electronics, where capital goods tend to be produced. Employment in these industries is now beginning to fall because of overcapacity. Japan had the greatest growth in manufacturing productivity throughout the 1980s. France and Italy also had high levels of productivity improvement. However, the European countries did not enjoy strong growth in domestic demand, and these improvements led to widespread job destruction in manufacturing. Exhibit 6 shows the employment performance of the six countries for the three manufacturing cases studied automotive, computers, and furniture along with the total for the three taken together. The auto case illustrates
Exhibit 6

should look to manufacturing for net job creation

Employment performance in manufacturing

d St at G er es m Ja a pa n n y
0 1 2

Net jobs created per thousand working age population*

ce

Ita ly Sp ai n

Fr an

Automotive
198092

Computer hardware
198191

Furniture
198091

Manufacturing case
Total

Computer software
198191 7 6 5 4 3 2 1 3 4 5 * Adjusted for growth in the working age population Including semiconductor manufacturing Computer software figures unavailable for Germany, Italy, and Spain Source: National household and establishment surveys; McKinsey analysis

THE McKINSEY QUARTERLY 1994 NUMBER 4

ni te

1st proof

AUTOMOTIVE

ntense competition in the 1980s increased cost pressure and triggered restructuring in the automotive industry worldwide. Employment fared best in Japan, where demand growth in the local market was combined with cost-driven improvements in trade performance.

wide. The difference for consumption was 64 percent; for trade, 61 percent. Each translated into an extra 4 percent a year in employment growth for the better-performing country. Wage levels relative to productivity were the critical determinant of trade performance. Countries with high productivity and low wages, like Japan at the beginning of the 1980s, fared well in the global market. Japanese companies were able to exploit their superior cost position to build market share rapidly, mostly at the expense of US manufacturers. Germany was able to postpone productivity improvements because of its high initial level of productivity, but has recently suffered large job losses. Concerns about market saturation in mature industries like this one are valid, but often overstated. Although real consumption of autos rose by more than 2 percent a year in all the countries studied, sustaining such growth will become harder in the future. High levels of penetration in traditional markets will limit further expansion in sales. Innovation must therefore continue if value-added per car is to increase. Emerging markets offer potential, but may not stimulate many jobs in developed countries, since import restrictions and cheap labor are prompting more and more foreign direct investment by traditional manufacturers.

Employment performance
Net jobs created per thousand working age population

France Germany Italy Japan Spain United States

5.3 0.5 3.0 0.7 2.3 0.6

Employment performance varied widely in the 1980s. Japan, Germany, and the United States were all able to keep jobs stable, while France, Italy, and Spain suffered heavy losses. These differences reected the countries varying ability to match annual productivity gains with at least equal growth in output. High output growth was the result of increased domestic consumption or improved trade performance both of which were related to the cost position and innovativeness of the rms in the industry. Local consumption and trade affected job creation too. Output differences between the best- and worst-performing countries were

the relatively strong employment performance of Japan and the relatively weak showing of France and Italy. Japan had stronger growth in domestic demand for automotive products because of the low penetration of its domestic market in 1980. But demand was also partly driven by the bubble economy. As a result, Japan overinvested in automotive capacity. Its performance was also boosted by an improving trade balance in the auto industry, which shows how trade can afect employment at the industry level. All the same, our aggregate analysis revealed that the employment benets produced by an improving trade performance in autos were entirely ofset by a declining trade performance in other manufacturing areas. In the automotive industry, the decline in manufacturing employment performance in France and Italy was partly caused by competitive pressure

THE McKINSEY QUARTERLY 1994 NUMBER 4

1st proof

COMPUTERS

f the United States has the worlds most innovative computer industry, why did Japan create more jobs? The answer: differences in competitive intensity.

and software/ services. These effects account for the weaker employment performance of the United States. While Japan was not insulated from these developments, structural change in its industry was slowed by low competitive intensity. This, in turn, was the result of a closed distribution structure (hardware vendors largely controlled retail prices), language barriers (the interpretation of kanji characters was difcult with early PC hardware and software), and a lack of venture capital (which helped speed vertical disintegration and led to the emergence of specialized players in the United States). This is why Japan had more employment growth and less productivity growth in the 1980s. The computer industry stands out from the other cases studied in that natural barriers to economic evolution played a more important role than articial (regulatory) barriers. As these barriers are overcome by the forces of globalization, we expect Japan to undergo the restructuring that the United States and Europe have already experienced.

Employment performance
Net jobs created per thousand working age population

France Germany Italy Japan Spain United States

NA NA NA

4.0

6.0 3.8

The computer industry illustrates the impact of fast technological change on job creation in an innovative business. In the United States and, to a lesser extent in Europe, technological breakthroughs linked to the emergence of PCs changed industry structure and prompted vertical disintegration. Specialized companies at each stage of the value chain translated economies of scale into sizable productivity gains, while large integrated suppliers such as IBM and DEC suffered dramatic restructurings. Vertical disintegration also forced US-based companies to source components on an increasingly worldwide basis. At the same time, these technological breakthroughs caused a shift in value creation and job growth from hardware to semiconductors

from the German industry through trade. At the beginning of the 1980s, Germany had a cost advantage over France and Italy, mainly because of its higher productivity. The consequent pressures led to massive restructuring of the French and Italian industries, resulting in a steep fall in automotive employment. As this example shows, trade can also have an indirect impact on employment performance through competitive intensity.

Constraints afecting services

Services constitute the only sector in industrial economies that is growing. In services, we chose particular industries in order to understand why the United States had an employment performance superior to that of any of the other countries studied. Exhibit 7 shows employment performance for the three cases selected banking, general merchandise retailing, and lm/TV/video. We also chose construction because not only does it reect the superior employment performance of the US, but it is also a major sector of the economy.

10

THE McKINSEY QUARTERLY 1994 NUMBER 4

1st proof

FURNITURE

s in most mature manufacturing sectors, the long-term employment trend in the furniture industry has been and will continue to be downward. Innovation helped to slow this trend in the United States, Italy, and Germany.

Employment performance
Net jobs created per thousand working age population

France Germany Italy Japan Spain United States 3.2

1.4 0.4 0.7 1.2 0.2

Demand was at in the French and Spanish industries chiey because new products and styles emerged more slowly here than elsewhere. Consumers tended to stick with traditional designs, and producers were hesitant to invest in new projects. In Italy, however, exible networks of small rms ourished by rapidly bringing new ideas to market. Large German retailers and manufacturers competed by spending heavily on design, equipment, and promotion. US companies focused on creating and expanding specialty niches. Furniture manufacturers everywhere face constant nancial exposure due to highly cyclical demand, limited nancial reserves, and few barriers to entry or exit. In order to survive, they have found ways to hire people when they are needed and lay them off during down cycles. As a result, employment is not as sticky as it is in some other sectors. Nearly 30 percent of furniture workers in Spain, France, and Japan, for example, lost their jobs in the global recession of the late 1970s and early 1980s. Employment rebounded somewhat in the second half of the 1980s as real output expanded. Traditional furniture manufacturers can expect additional employment pressure from lowwage countries in the near future, although it is highly unlikely that the bulk of the industry will relocate its production base. Although labor-intensive componentry will be increasingly outsourced, design and assembly will remain close to the consumer.

Trends in furniture employment may reect a broader pattern playing out in many lowprole, mature manufacturing industries. Some countries, such as France, lost jobs mainly because output stagnated. Others, like Japan, suffered falling employment because rapid productivity rises outstripped gains in output. Still others, such as Spain, experienced both a sharp increase in productivity and a decline in value-added. The result for all three countries was poor employment performance. By contrast, the net decline in jobs was relatively mild in the United States, Germany, and Italy. The reason can be found in the structure of the furniture manufacturing and retailing industries in each country, rather than in capital, labor, or product market differences.

Exhibit 7

Employment performance in services and construction

Sp a Fr in an Ita ce l Ja y pa G n e U rma n St ite ny at d es
1 0 1 2 3 4 5

Net jobs created per thousand working age population*

Banking
198292

Retail
198090

Film/TV/video
198092

Service case
Total

Construction
198090 11 10 9 8 7 6 5 4 3 2 6 7 8 9 * Adjusted for growth in the working age population Source: National household and establishment surveys; McKinsey analysis

THE McKINSEY QUARTERLY 1994 NUMBER 4

11

1st proof

BANKING

anking illustrates how product, labor, and capital market barriers can affect employment through changes in competitive intensity. The United States, France, and Spain all experienced falling employment in traditional banking products as competitive intensity increased during the 1980s. In the United States, however, the right combination of product, labor, and capital market factors led to rising employment in mortgages and securities.

but it would not have been nearly so big without a radical change in the mortgage business system. Specialized mortgage banks used securitization and IT-driven underwriting to increase efciency and reduce interest margins and origination fees, thus stimulating demand. This innovation depended on a favorable regulatory environment, securitization, and exible labor. The absence of these conditions in Europe and Japan prevented similar developments. In the US securities industry, regulations such as strong antitrust enforcement, Chinese walls in underwriting, and transparent accounting standards encouraged high retail penetration and boosted employment. The same regulations also helped the United States develop competitive advantage in securities origination, structuring, and trading. Japan achieved high securities employment without having many of these regulations, although a substantial part of its employment is maintained by high regulated commissions in equity trading. The productivity of banking is becoming a concern for national policymakers in terms of both competitive advantage and the efciency of intermediation. The importance of the industry goes well beyond the 1 to 2 percent of the working age population it employs. Employment policy in this industry allows for employment growth consistent with high productivity and efcient intermediation services to the rest of the economy.

Employment performance
Net jobs created per thousand working age population

France Germany Italy Japan Spain United States

0.5 1.9 0.9 1.0 0.7 1.9

Deregulation and privatization brought increasing competition in the United States, France, Spain, and among Japanese city banks during the 1980s. In mature traditional banking products, additional cost pressures and newly available automation technology prompted banks to improve efciency and reduce employment. Competition remained low in Germany, Italy, and Japan (outside the city banks), allowing employment to continue to grow. Tougher competition in the United States generated a boom in residential mortgages that created more than enough jobs to compensate for the decline in traditional products. Falling interest rates, demographics, and tax law changes also fuelled the boom,

Exhibit 8 illustrates the relative importance of various factors in explaining the diferences in employment performance between the benchmark country with the highest net job creation (the US in every case) and the other countries. Employment growth rates are directly determined by output and productivity growth rates. These are in turn determined by industry dynamics, in particular, competitive intensity, innovation, and trade performance. Finally, industry dynamics are themselves determined by conditions in the markets for capital, labor, and products.

12

THE McKINSEY QUARTERLY 1994 NUMBER 4

1st proof

RETAILING

etailing is undergoing a structural shift: the destruction of old formats and the creation of innovative new ones that offer more value to the customer and operate with greater efciency. Rigidities in the product and labor market retarded the emergence of high-value formats in France, Germany, and Japan, leading to lower overall employment.

high-value formats tends to create jobs; a shift toward high efciency tends to destroy them. Policymakers are in a position to inuence the kinds of formats that retailing entrepreneurs develop. By modifying zoning regulations, anti-competitive practices, and opening hour limits, they can foster job creation in high-value formats. Product market barriers restricted the development of high-value formats in favor of high-efciency ones in France, Germany, and Japan, and ultimately harmed employment. Weaker constraints in the United States resulted in a shift beyond high-efciency to high-value formats, and created jobs. Italy and Spain increased employment by protecting their low-productivity stores. Retailing is one of the few industries we studied where labor market factors appear to have a major effect on employment. High labor costs and a lack of exibility in using part-time labor and in hiring and ring prevented entrepreneurs from exploiting the advantages inherent in new formats.

Employment performance
Net jobs created per thousand working age population

France Germany Italy Japan Spain United States

3.6 2.1 4.1 2.3 3.5 4.9

The transformation of retailing involves innovation in the form of new store formats focusing either on increasing efciency through scale economies, or on increasing value to the customer through specialization in a particular product group. A shift toward

Exhibit 8

Reasons for superior employment performance

de o Se r av vic er e c ag as e e ng vi uc Co

Benchmark country: United States

in

V/

ili

nk

ta

Ba

Re

Capital market More pressure from owners Less government ownership/support Readily available capital Labor market Low labor cost High availability/low benefits More flexibility Product market Fewer restrictions on output and competition More new business facilitation Rapid demand growth Industry dynamics/competitive intensity Better trade/FDI performance More price competition/restructuring More innovation/new products Higher output growth Higher productivity growth Lower productivity growth
NA

Fi

lm

ns

Important

Secondary

Undifferentiating

/T

tr

tio

Service sector cases

THE McKINSEY QUARTERLY 1994 NUMBER 4

13

1st proof

FILM/TV/VIDEO

n this high-prole industry, innovation is the driving force in both the creation and the destruction of jobs. Slower employment growth in Europe can be explained by regulations that inhibited the emergence of new labor-intensive segments.

television, which blossomed after European state monopolies were ended, and in new outlets such as videocassette and cable/satellite, where few barriers blocked emergence. Employment-reducing blockages hampered French video retailing and private TV. These segments grew slowly, in part because of regulations intended to preserve cinemas. Similar negative effects can be seen in Spain and Italy, where government delays in deploying cable and in dening rules for the cable/satellite market slowed output and employment growth. Lacking many of these impediments, the US industry had rising employment. It also beneted from a mounting trade surplus in lm/TV production. Most other countries were net importers. Restrictions on advertising limited the ability of European TV channels to pay for expensive domestic production. Governments that tried actively to manage this industrys evolution may have retained more employment in existing forms of entertainment, but only at the expense of better net job performance.

Employment performance
Net jobs created per thousand working age population

France Germany Italy Japan Spain United States

1.1 0.9 0.5 1.3 0.3 1.5

Employment growth in the industry that creates and distributes television programs and feature lms was high in the United States, Japan, France, and Germany, but low in Spain and Italy. Any similarity in aggregate performance belies signicant differences in the expansion of various segments. Innovation had a radical effect on how people watch this industrys products. Latent demand for TV and lmed entertainment manifested itself wherever allowed: in private broadcast

The service cases, along with construction, show that the United States had higher output growth across the board. We found that this distinction derived from important factors both within and beyond each industry. Important factors within the industry were more innovation and new product development, and fewer product market restrictions. The innovations that made big diferences were securitization and derivatives in banking; specialty formats in general merchandise retailing; and new lms, cable television channels, and video rental formats in the lm/TV/ video industry. Industry-specic restrictions that constrained employment performance in Europe and Japan included product prohibitions and lack of transparency in banking; zoning laws and controls on the forms of competition in retailing; and a series of constraints, such as maximum numbers of homes per cable licence in Japan, in the various segments of lm/TV/video. Construction followed the same pattern of product market restrictions. Here zoning laws limit land use, thereby raising prices and suppressing demand for housing. These restrictions spill over into banking and

14

THE McKINSEY QUARTERLY 1994 NUMBER 4

1st proof

CONSTRUCTION

he myriad rules and regulations governing the construction industry in Europe and Japan provide an excellent example of how product market restrictions stie output. Faster output growth together with slower gains in productivity secured a better employment performance for the United States.

Employment performance
Net jobs created per thousand working age population

France 10.7 Germany Italy Japan Spain United States 8.1 6.3 1.3 2.8 3.0

Spain experienced growth in real construction volume in line with overall growth in GDP. Germany, France, and Italy, by contrast, saw construction growth lag GDP growth. Strict zoning laws are a prime cause of this lag in that they restrict the supply of residential and commercial land. This leads to high land prices and high costs of construction, both of which constrain output growth. Building codes and norms also have a negative impact on volume by increasing material and labor costs. European and Japanese authorities further distort housing markets through rent controls and tax incentives. Japan and Spain raised output and employment in spite of such product market restrictions; extensive upgrading of housing and public investments in infrastructure acted as drivers of growth. Productivity levels and growth rates must also be taken into account. The strong employment performance of the United States is partly due to its low productivity growth rate. Labor market factors such as benet levels and wage-bargaining bodies played only a secondary role in accounting for growth in output, productivity, and employment. For Japanese and European policymakers, the implication is that relaxing the regulatory environment is likely to increase construction output and employment. This raises the possibility of tradeoffs with broader social objectives. Policymakers should question whether restrictions serve the public good or merely protect vested interests.

The construction industry is subject to numerous regulations regarding land use, building quality, safety, rent, and taxes. However desirable such regulations may be in their own right, they adversely affect construction volume and, therefore, employment. The effect on output of different degrees of regulation rather than competitive intensity or industry structure directly explains differences in employment. France, Germany, and Italy lost jobs; Japan more or less maintained employment; and Spain and the United States were able to create jobs. Construction employment depends largely on output growth. The United States, Japan, and

retailing through lower demand for mortgages and a reduced land area available for shopping centers. Factors outside the industries drove up demand for banking and retailing services to a much greater extent in the United States than elsewhere. These factors were productivity increases across the economy and new business formation, both of which gave rise to higher per capita incomes that increased demand for these services. In services and construction, we found that labor market factors seemed to have less impact on employment performance than product market factors. The one signicant exception is that relatively high labor costs in retailing

THE McKINSEY QUARTERLY 1994 NUMBER 4

15

1st proof

WHY EMPLOYMENT PERFORMANCE DIFFERS

in Europe kept prices high, suppressed demand, and made some new types of retailing less economic. These high labor costs are the product of union bargaining power, high minimum wages, and high unemployment benets that leave people understandably unwilling to work at low wages a problem that is particularly acute in Europe at present. The result in France, for example, is that the average retailing wage is higher than the average manufacturing wage.
Exhibit 9

Gross job creation and destruction

Average annual rates of total employment by operating site
Percent Job creation Job destruction Openings Closures Expansions Contractions

France
198489

7.3 6.9 2.5 1.9 4.1 3.6 8.9 7.2 2.9 7.0 5.6 6.5 7.5 8.6

6.6 5.9 9.0

13.9 12.8

Germany
198390

Italy
198489

12.7 10.6 4.3 10.1 13.2

United States
198488 Source: OECD

Contrary to conventional wisdom, the availability of skilled workers from Germanys apprenticeship program has not prevented an employment problem there; almost half (43 percent) of the countrys unemployed have followed such a program. Moreover, data on gross job creation and destruction indicate that business operations in Europe (with the exception of Germany) create and destroy jobs about as much as in the United States (Exhibit 9). Unions, it seems, do not limit workforce exibility to the extent commonly believed.

Product market restrictions in the service sector are, in our view, probably the key factor behind diferences in employment performance. For lowwage industries such as retailing and construction, high minimum wages and high unemployment benets are also important.

Policy implications
The main implication of our ndings for policy is that if industrial countries want to improve their employment performance, they must remove product market restrictions in services and let the economy evolve naturally. Product market restrictions are chiey designed to protect existing interests, including existing jobs. Our research strongly indicates that such eforts invariably fail. In fact, in most cases they have exactly the opposite of their intended efect. Preserving existing jobs in any industry slows productivity growth and weakens the competitiveness of that industry. Sooner or later competition from best practice comes along, and when it does, restructuring is severe. Aterwards, there will be fewer jobs in the industry than there would have been if it had practised continuous productivity improvement.

16

THE McKINSEY QUARTERLY 1994 NUMBER 4

1st proof

WHY EMPLOYMENT PERFORMANCE DIFFERS

In low-productivity industries such as retailing and construction, we have found that reducing the minimum wage and cutting unemployment benets will also help create employment. It is better, we believe, for low-skilled workers to be employed in these industries than to be unemployed. The economy will perform better, and income distribution objectives can be achieved by adjusting atertax income, rather than by intervening directly in the workings of the market. Liting product market restrictions will also afect some noneconomic dimensions of society. Some areas that might be afected are the agricultural base, green space, the structure of urban development, popular culture, condence in the banking system, fairness for each individual, and stability in individuals lives. We have not investigated the tradeofs that would be involved against these other areas. However, the unemployment situation, especially in Europe, is so severe that liting product market restrictions and making some of these tradeofs seem necessary. If European countries had matched the job creation performance of Japan and the United States over the 1980s without sufering any productivity penalty, their GDP would be 5 to 15 percent higher and their unemployment problems would have disappeared.

What rms can do

Corporations clearly have a stake in the health of the societies in which they operate. Since unemployment now threatens the cohesion of some industrial countries, it is a natural concern for corporations. Indeed, some have been reluctant to restructure to improve their performance because they fear that job losses would exacerbate the social problem. Our ndings show that such a reaction does not help. It simply postpones the unemployment problem and makes it worse when the inevitable restructuring nally occurs. Corporations can best serve the societies in which they operate by boosting their performance through continuous productivity improvements and innovations that lead to successful new business formation. The main barriers in their path lie in current restrictions in the product market that reduce competitive pressure on corporations in the short term and allow them to delay restructuring. These regulations seem to be in the interests of both shareholders and employees, but they actually help neither in the long term. Corporations, especially those with potentially superior performance, should therefore urge governments to lit these restrictions. Bill Lewis is Director of the McKinsey Global Institute. Ren Limacher and Michael Longman are consultants in the McKinsey Global Institute. Copyright 1994 McKinsey & Company. All rights reserved.

THE McKINSEY QUARTERLY 1994 NUMBER 4

17