Journal of Purchasing & Supply Management 16 (2010) 221229

Contents lists available at ScienceDirect

Journal of Purchasing & Supply Management

journal homepage: www.elsevier.com/locate/pursup

Dispute resolution and litigation in the construction industry. Evidence on

conicts and conict resolution in The Netherlands and Germany
Frits Tazelaar a,n, Chris Snijders b
a
b

Department of Sociology, Utrecht University, PO Box 80.140, 3508 TC, Utrecht, The Netherlands
School of Innovation Sciences, Eindhoven University of Technology, PO Box 513, 5600 MB, Eindhoven, The Netherlands

a r t i c l e in fo

Keywords:
Construction industry
Conict resolution
Litigation

abstract
The construction industry is regarded to be a tough and competitive business characterized by shortterm and opportunistic relations rather than being based on cooperative partnerships. In particular,
conicts and litigation have been claimed to proliferate in the construction industry. Upon closer
inspection of the literature, it seems that the empirical basis of these claims is largely circumstantial.
Using data on contractorsubcontractor relations in the construction industry in The Netherlands, we
consider the extent to which litigation in construction is common. Then we compare the results to
similar data sets on IT-purchasing both in The Netherlands and Germany, and to a data set with more
general business-to-business transactions of larger Dutch and German rms. We nd some evidence
that the construction industry has higher percentages of transactions leading to either arbitration,
suspension of the relation, or legal steps (1.6% versus 1.2, 0.4 and 0.6). The differences are however not
as extreme as one might conclude based on supercial reading of the popular and scientic literature,
and certainly not bigger than the differences between the other data sets.
& 2010 Elsevier Ltd. All rights reserved.

1. Introduction
The construction industry has often been characterized as a
harsh business. Even a brief literature review suggests that
behaviour in the construction industry is characterized to be
antagonistic and confrontational (Cox and Thompson, 1997; Saad
et al., 2002) and relying on detailed contract specication and
close performance monitoring (Kadefors, 2004). An important and
typical consequence of the tough and competitive world that the
construction industry apparently is, is that there seems to be a
high and increasing level of conict and disputes (Lavers, 1992;
Brooker and Lavers, 1997). In addition, researchers have reported
a dominant blame-culture and a strong tendency towards the use
of litigation to resolve disputes.1
A closer look at the scientic literature reveals that there is
surprisingly little representative empirical evidence to back up
the harshness of the construction industry. Many publications on
disputes and conict resolution in the construction industry are
largely conceptual, or make claims about the state of affairs
without reference to empirical data based on specic experiences,

Corresponding author. Tel.: + 31 30 6910047; fax: + 31 30 6910046.

E-mail addresses: f.tazelaar@uu.nl
(F. Tazelaar), c.c.p.snijders@tue.nl (C. Snijders).
1
See, for instance, Khalfan et al. (2007), Davis et al. (1989), Fenn (1991),
Fellows (1992), Latham (1994), Brooker and Lavers (1997), Saad et al. (2002),
Harmon (2003), Colledge (2005), and Gibbons (2007).
1478-4092/$ - see front matter & 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.pursup.2010.08.003

impressions, perceptions and opinions of practitioners (cf. Bryde,

2008, and several contributions to the proceedings of the First
International Construction Management Conference at UMIST/UK
in 1992). In still other publications empirical generalizations are
formulated based on case study research with only a limited
number of interesting but not necessarily representative cases.
Even when researchers themselves have been cautious about and
warned against generalizations in the original study (e.g. Khalfan
et al., 2007), subsequent researchers who present literature
overviews on the subject sometimes loose the cautiousness of
the original.
Let us illustrate the issue with an example. The conclusions in
the inuential Latham Report (1994) with respect to the widespread use of litigation in the UK construction industry have been
quoted numerous times, and the report is generally used as a
reference that is showing that the construction industry is a tough
and troublesome business (e.g. Bresnen and Marshall, 2000; Saad
et al., 2002). However, these conclusions are largely based on
analyses following several poorly performing projects, so that
representativeness is problematic. The same holds for the work
of the Dispute Avoidance and Resolution Task Force of the
American Arbitration Association (as mentioned in Stipanowich,
1997; Colledge, 2005). Reports such as these and the general
scientic literature have contributed to an image of the
construction industry characterized by inefciency and toughness, but it is not clear whether this image is correct. This image
might be completely correct, but nobody bothered to really count

222

F. Tazelaar, C. Snijders / Journal of Purchasing & Supply Management 16 (2010) 221229

in a representative way. We do. The aim of our article is to add to

the empirical evidence with respect to conict resolution and
litigation in the construction industry. Is it really as harsh as
people tend to think?
In our analysis we focus on conict resolution and litigation as
our main indicators of harshness (as is customary in the
literature). First we outline the literature on conict resolution
and litigation and introduce the concept of a dispute pyramid as
introduced by Sarat (1984) as a useful tool to analyze dispute
resolution. We then introduce the Dutch construction industry,
which is also generally seen as the tough and competitive world
that construction seems to be in general. Our data collection
allows us to then present estimates on conict and dispute
resolution. In the subsequent section, we compare these data to
similar data in the IT-sector in the Netherlands, so that we get a
feel for the magnitude of a within-country sector difference. Then,
we compare these IT-data with an identical data collection of ITtransactions in Germany, so that we can estimate the difference
between countries, within a sector. Fourth, we compare these
data to a more general set of purchasing transactions of Dutch
rms. A conclusion and discussion section concludes the paper.

2. Conict resolution and litigation in a broader business

context
In the 1960s and 70s there were opposite views on litigation in
business relations. At one end of the spectrum, many researchers in
the United States had the impression that there was too much
litigation. This was often described as the law explosion or the
legal explosion, signifying the growing intrusion of law in every
aspect of American Society (Barton, 1974, p. 567). Similar buzzwords that were used were hyperlexis (Manning, 1976, p. 767) or
the litigation explosion (for an overview, see Sarat, 1984, p. 319).
At the other end of the spectrum there was a line of research
sparked by a publication by Macaulay (1963). Macaulay investigated the social functioning of contract law in the business world.
His research consisted of interviews with 68 businessmen and
lawyers from 43 companies and 6 law rms. An important
conclusion drawn from the ndings of the research was that in
practice, disputes were frequently settled without reference to
either the contract or potential or actual legal sanctions: Even
where the parties have a detailed and carefully planned agreement
[y] often they will never refer to the agreement but will negotiate
a solution when the problem arises, apparently as if there had
never been any original contract. (Macaulay, 1963, p. 61). Later,
Macaulays study was labeled a classic not just for the fact that it is
(now almost 50 years) old, but also because it is nowadays widely
recognized as one of the most inuential contributions to the eld
of conict resolution in the world of business. Even though most
people expected that many rms run to the courts as soon as they
can, closer inspection suggested they did not.
Of course there were and are potential counter-arguments to
the validity of Macaulays claim. An obvious one is that the picture
painted by Macaulay is now outdated. While in the 50s and 60s of
the previous century the companies, their mutual relationships and
their products and prices were stable for a relatively long time, our
perception of todays business arena is rather different: more and
more business players are everywhere, and there is more general
turbulence. These factors may in the meantime have undermined
Macaulays notion of long-standing relationships, as recognized by
Macaulay (e.g. Galanter and Rogers, 1991; Macaulay, 1985;
Galanter et al., 1991). The most important differences would
appear to lie in the increased internationalization, the increased
competition in business, and the increased litigiousness in the
world of business. Initially, it seems obvious and consistent with

what we see happening around us, that the parties in a dispute

confront each other in court more often than in the past.
It is indeed a fact that over the past decades there has been a
substantial increase in the number of lawsuits in the area of
business contracts that are brought before the various courts,
both in and outside the United States. But appearances can be
deceptive. According to Galanter the litigation explosion
hypothesis has not been based on the growth in lings in federal
courts and the growth in size of the legal profession only, but also
on specic accounts of monster cases (such as the AT & T and IBM
antitrust cases), the vast amount of resources consumed in such
litigation, and on war stories (Galanter, 1983). He concludes that
we need a more contextual reading of the facts. The increase in
lawsuits obviously needs to be viewed in the perspective of a
tremendous growth in economic interactions and business
activities (Sarat, 1984). What we need to consider is what is
happening with the percentage of transactions that ultimately
ends up in a lawsuit. Sarat noticed that what looks like a ood of
litigation as viewed from the courthouse steps, might appear
rather modest when compared against the vast magnitude of
lawsuits that could have materialized but did not (Sarat, 1984).
Dunworth and Rogers (1996) likewise consider conicts in the
light of the number of potential conicts. They took a more
precise look at the growth in the number of lawsuits between
companies in the 70s and 80s of the previous century, and
concluded that a large part of the growth could be attributed to
product liability issues. Their statistical analyses indicate that if
those issues are set aside, the number of civil court cases among
businesses has in fact grown less than could be expected solely on
the grounds of economic growth. On top of this, the ndings from
later replications of these types of studies in different countries
hardly seem to deviate from the picture as originally sketched by
Macaulay (Blegvad, 1990; Beale and Dugdale, 1975; van Houtte
et al., 1995; see also the overview by Deakin and Michie, 1997).
The picture remains that it is usually quite exceptional for
disputes between companies to end up in court. In the wording of
Sarat: Litigation can be thought of as the tip of the iceberg or the
apex of a rather complicated process through which disputes
emerge, develop, and are resolved (Sarat, 1984, p. 331).
Following Galanters suggestion to analyze the dispute process
in more detail, Sarat uses the metaphor of a dispute pyramid for
the steps of conict resolution, whereby the rst step is the stage
of problem recognition, which occurs when one or another event
or transaction is perceived to be injurious or undesirable. The
second step is deciding to take up the issue with the other party
(instead of just accepting the problem and moving on, a
grievance). The third step is to decide to blame the other party
for at least part of the problem, confronting an adversary rather
than in the legal sense of starting a lawsuit. Disputes (all Sarats
terminology) can be said to exist only after the grievance has been
perceived and acknowledged and after the claim has been made,
but only if that claim is resisted. Sarat underlines that unless a
claim is made, a dispute cannot occur (Sarat, 1984). Once a
dispute has emerged, choices must be made about how it can be
processed, whereby at least in theory, a wide range of alternative
procedures might be employed, ranging from direct bilateral
negotiations to third party mediation or arbitration, and ultimately to adjudication and litigation (Sarat and Grossman, 1975).
The pyramid represents the stages through which events pass in
the disputing process. One can think of each of the stages below
litigation as establishing a benchmark or baseline against which a
litigation rate can be calculated. Thus we can compare litigation to
all potentially litigable grievances, claims, or disputes (Felstiner
et al., 19801981; Sarat, 1984). Another possibility will be to view
the incidence of litigation as a percentage or proportion of
potential legal claims (cf. Miller and Sarat, 19801981). One way

F. Tazelaar, C. Snijders / Journal of Purchasing & Supply Management 16 (2010) 221229

223

Fig. 1. The dispute pyramid (cf. Sarat, 1984).

or the other, this opens the door to view litigation in perspective.

Fig. 1 shows the dispute pyramid.
The focal starting point of our article is that viewing litigation
in perspective and in a representative way is imperative, but
literature on the construction industry generally fails to meet
these requirements. In the quarter of a century after Galanter
(1983) and Sarat (1984) suggested to always compare litigation to
the potential for litigation there have only been some suggestions
along these lines. Most notably, in the 2008 Annual Report of the
Technology and Construction Court it is now stated (on p. 4) that
y whilst the number of new claims issued is one measure of
activity, a number of cases are dealt with by way of pre-action
applications which do not currently feature within the (2008)
gures but will be separately identied from January 2009. This
shows at least the explicit intention not only to count the absolute
number of court cases, but also cases one step beneath the top
level of the pyramid that could have arrived in court but did not.
The closest we could nd to a representative study, although not
really a quantitative one, is the extended case study by Jettinghoff
(2001), who interviewed owners and managers from 16 companies in three countries in the construction industry, and compared
his ndings in the construction industry with those in plastics and
freight carriers. His ndings are that in plastics legal measures are
rarely taken, whereas this is more likely in the freight industry
and still more likely in construction, backing up the general
intuitions about the construction industry. We feel the discipline
cannot progress without some rigorous quantitative measurement of what is taking place in the construction industry,
representatively taking into account the probability of conict
occurrence and resolution, and comparing these numbers between sectors and countries.

3. Measuring conict resolution in the Dutch construction

industry
3.1. Data collection and methodology
We analyze the construction industry on the basis of a data set
consisting of 448 contractorsubcontractor transactions in The
Netherlands (WELL06), as introduced in Kamann et al. (2006) and
Welling (2006). We briey reiterate the basic characteristics of this
data set. Contractors were contacted by telephone and invited to

Table 1
Overview of the construction data (WELL06, N 448).
Firm size (contractor)

Infrastructural worksa
Housing, Ofces and Factoriesb
Other
Total: 448
a
b

Small
o 20

Medium
(20100)

Large
(4100)

Unknown

55
122
4
181

52
83
10
145

42
69
5
116

1
1
4
6

In Dutch: GWW-sector (Grond-, Weg en Waterbouw).

In Dutch: B&U (Burgerlijke en Utiliteitsbouw).

participate in an online survey on cooperation between contractor

and subcontractors in the construction industry. The telephone
numbers were sampled from the Dutch Chamber of Industry and
Commerce database 2004, and we stratied our sample with
respect to the type of construction project and the rm size. In total
3612 phone calls were made with contractors. Contractors were
asked whether they could think of the most recently completed
transaction of which they were fully aware, with a value of at least
5000 Euro, and in which they had used a subcontractor and/or
supplier. From the 3612 contractors who were called, 1984 (55%)
could come up with such a transaction and declared to be willing to
participate. All 1984 were sent an email invitation to participate in
the online survey. The survey consisted of seven parts, roughly
following a typical contractorsubcontractor transaction chronologically, ranging from information on the kind and content of the
transaction, the way in which the subcontractor was chosen, the
agreement or contract that was used, the result of the transaction,
as well as some questions about the respondent him/herself. Filling
out a survey took about 30 min on average. Eventually, 448
participants lled out the complete questionnaire for an email
response rate of 23% and a total response rate of 12%. Table 1,
copied from Kamann et al. (2006, p. 33) shows an overview of the
complete sample data.
3.2. Measurement of conicts
A substantial part of the survey was devoted to measuring the
kind and severity of the problems that were encountered during
the transaction. Respondents could indicate on a 5-point scale

224

F. Tazelaar, C. Snijders / Journal of Purchasing & Supply Management 16 (2010) 221229

(did not occur, was a minor issue, was somewhat of a problem,

was a problem, was a big problem), whether 23 different kinds of
problems occurred during or after the transaction. About 16% of
Table 2
Elaborate dispute pyramid for the construction data. The percentages show the
probability of a given behaviour when compared to the total number of
transactions.
Arbitration, suspension, legal steps
Demand additional personnel from partner
Breach of relation with partner
Involve extra own personnel
Demand discount or claim damages
Involve higher level management
Delay payment
Set deadlines
Negotiate about problem
Discuss problem
There is a problem
Total number of transactions

1.6%
4.9%
7.8%
11.2%
16.7%
21.2%
27.7%
33.0%
53.8%
67.4%
83.7%
448

( 7/448)
( 22/448)
( 35/448)
( 50/448)
( 75/448)
( 95/448)
( 124/448)
(148/448)
( 241/448)
( 302/448)
( 375/448)

Overall percentage of arbitration, suspension or legal steps 1.6% (7/448;

CI95 0.6%3.1%)

Table 3
Restricted version of the dispute pyramid for the construction data. Percentages
show the probability to reach the next step of the pyramid, given that a lower step
was reached.
Arbitration, suspension, legal steps
Delay payment, claim damage
Refer to contract, set deadlines
Discuss problem
There is a problem
Total number of transactions

10%
53%
49%
81%
84%
100%

(7/78)
(78/148)
(148/302)
(302/375)
(375/448)
(N 448)

Overall percentage of arbitration, suspension or legal steps 1.6% (7/448;

CI95 0.63.1%)

the transactions had no problems whatsoever (did not occur for

all 23 problem categories). When we only consider those cases in
which the respondent indicated that there was somewhat of a
problem, then 62% of the cases has at least one problem. Using
that same denition, the average number of problems is 2.3
(st.dev3.1) with a maximum of 15. For those with at least one
problem, the average number of problems is 3.2 (st.dev3.2). The
problems mentioned most often are discussion about the
preciseness and tidiness of the work (24% of transactions),
discussion about the price of extra/less work (22%), and
discussion about when work should start and nish (21%). For
more detailed information on the conicts that arose we refer the
reader to Kamann et al. (2006) and Welling (2006).
In line with Miller and Sarat (19801981) and similar to the
approach in Rooks and Snijders (2001) we use the dispute
pyramid to visualize the subsequent steps after a problem
between contractor and subcontractor has been identied. In
our data collection we considered 10 different reactions to the
problem(s), ranging from discussing the problem with the
partner, to mild sanctions, to legal steps and arbitrage. Our
analyses show these 10 steps form a reasonably strong Mokken
scale (H0.49, Mokken, 1971, 1996), which is evidence of the fact
that the steps on the pyramid are indeed taken in the order from
light to more severe without skipping steps (if that would not be
the case, Mokkens scalability coefcient H would be smaller).
Table 2 shows an overview of the dispute pyramid in these data.
In fact, Sarats original dispute pyramid as shown in Fig. 1 is
coarser than the one we give in Table 2. Because we shall compare
the dispute pyramid for these construction data with other data
later, we reduce the pyramid by focusing only on the categories
that are most closely related with Sarats setup (see Fig. 1 and
Table 3). In this case Mokkens H increases to 0.71, reinforcing the
idea that the steps on the pyramid are taken subsequently (no
steps are skipped). In addition, we add the percentage to reach a

Table 4
Overview of the different data collections.

Subject of research

Unit of analysis
Data collection
Sample type:
 rms

 transactions

Sampling frame
Indicating
willingness to
participate
Participating rms
with 1 or more
questionnaires
Total response rate
Completed
questionnaires
Operationalization of
problems
% problematic
transactions
Scalability dispute
pyramid: Mokken
H-coefcient.

WELL06

MAT95

EDV99

NEVI97

Cooperation between contractors/buyers and

subcontractors/suppliers in the Dutch
construction industry
Contractorsubcontractor transaction
2004

Cooperation between buyers

and IT-suppliers in The
Netherlands
BuyerSupplier transaction
1995

Cooperation between buyers

and IT-suppliers in Germany

Cooperation between buyers

and suppliers in The
Netherlands
BuyerSupplier transaction
19931998

Stratied sample of rms (contractors rm

size in 3 categories)
Stratied sample of transactions
(infrastructural vs. housing, ofces and
factories)

Random sample

BuyerSupplier transaction
1999

Convenience sample of rms

Stratied sample of
transactions (hardware vs.
software; standard vs.
complex)

Stratied sample of rms

(2 regions)
Stratied sample of
transactions (hardware vs.
software; standard vs.
complex)

3612 contractors
1983 contractors

1335 buyers
902 buyers

1702 buyers
1094 buyers

23 buyers

448 contractors

788 buyers

832 buyers

23 buyers

12%
448

59%
971

49%
1019

318

23 types of potential problems

84%

11 types of potential problems 11 types of potential problems Various types of problems,

depending on transaction
kind
73%
43%
67%

H 0.74 (strong scale)

H 0.72 (strong scale)

H 0.51 (strong scale)

Stratied sample of
transactions (volume/price;
uncertainty; past experience)

H 0.57 (strong scale)

F. Tazelaar, C. Snijders / Journal of Purchasing & Supply Management 16 (2010) 221229

higher step on the pyramid, given that a previous step was

reached.
Table 3 (and the subsequent Tables 57) should be read as
follows, from bottom to top. The total number of transactions
equals 448. There is a 84% probability that at least some problem
occurs. Given that one or more problems occur, there is an 81%
probability that the problem will be discussed with the other party
(the other 16% of the cases are never communicated to the other
party). Given that the problem has been discussed, there is a 49%
probability that explicit reference to the contract will be made or
explicit deadlines will be communicated to the other party. If that
happens, there is a 53% probability that payment will be delayed or
damage will be claimed. Given delayed payment or claimed
damage, there is a 10% probability that matter will escalate further
and lead to arbitration, suspension, or further legal steps.
As these data show, the litigation part of the pyramid (the top
layer) is a rather small part of the total set of problems. Our best
estimate is 1.6% ( 7/448), with a 95% condence interval ranging
from 0.6% to 3.1%. Note that even from those cases where the
problems have not been solved after several steps have been tried
including delaying payment and claiming damage, only 10% lead
to litigation. In general, stepping up the pyramid becomes less
likely as the pyramid is climbed. Still, one could rightfully argue
that 1.6% of many transactions still is a very large number. To get
a better feel for how high or low these percentages actually are,
we now compare these results to results from other domains.

Table 5
Dispute pyramid for MAT95. Percentages show the probability to reach the next
step of the pyramid, given that a lower step was reached.
Arbitration, suspension, legal steps
Delay payment, claim damage
Refer to contract, set deadlines
Discuss problem
There is a problem
Total number of transactions

6%
51%
72%
82%
73%
100%

(12/204)
(214/418)
(418/582)
(582/710)
(710/971)
(N 971)

Overall percentage of arbitration, suspension or legal steps 1.2% (12/971;

CI95 0.6%2.1%)

Table 6
Dispute pyramid for EDV99. Percentages show the probability to reach the next
step of the pyramid, given that a lower step was reached.
Arbitration, suspension, legal steps
Delay payment, claim damage
Refer to contract, set deadlines
Discuss problem
There is a problem
Total number of transactions

5%
38%
73%
63%
43%
100%

(4/ 77)
(77/ 201)
(201/ 273)
(273/ 435)
(435/1019)
(N 1019)

Overall percentage of arbitration, suspension or legal steps 0.3% (4/1019;

CI95 0.11.0)

Table 7
Dispute pyramid for the NEVI97 data. Percentages show the probability to reach
the next step of the pyramid, given that a lower step was reached.
Arbitration, suspension, legal steps
Delay payment, claim damage
Refer to contract, set deadlines
Discuss problem
There is a problem
Total number of transactions

8%
46%
31%
90%
67%
100%

(2/26)
(26/57)
(57/186)
(186/206)
(206/308)
(N 308)

Overall percentage of arbitration, suspension or legal steps 0.6% (2/308;

CI95 0.072.3%)

225

4. A comparison with dispute resolution and litigation in

other domains
We consider three alternative data sets, whose analogous
setup allows for a rather direct comparison of our results from the
construction industry. First we compare the construction data
with a data set on IT-transactions so that we can compare the
construction data with data from the same country but a different
sector (MAT95). Then we compare the IT-data to the same IT-data
as collected in Germany (EDV99) so that we can compare country
differences within a given sector. Finally we compare our results
with a more general data set of buyersupplier transactions from
a variety of organizations and sectors in The Netherlands and
Germany (NEVI97).
For an overview of the various research projects, see Table 4.
We rst consider the MAT95 data (cf. Batenburg, 1997; Batenburg
and van de Rijt, 1998; Buskens and Batenburg, 2000). Conict
resolution results related and similar to the ones reported here
about MAT95 have been reported earlier in Rooks and Snijders
(2001).
MAT95: The External Management of Automation 1995
(MAT95) is a large-scale survey on the purchase of IT-products
by Dutch SMEs (5200 employees; Batenburg and Raub, 1995;
Batenburg, 1997). The (stratied) sampling frame was a businessto-business database of Dutch SMEs that contained information
about the characteristics of these SMEs with respect to automation. Key informants of buying rms were rst briey interviewed
through a structured Computer Assisted Telephone Interview
(CATI). In the CATI-interview, cooperation was asked from an
employee responsible for automation in the rm and a transaction was selected randomly from all IT-investments of the rm in
the previous 5 years and about which the respondent was well
informed. A main sample of 547 IT-transactions was obtained.
Subsequently, the data set was extended with an additional
sample of innovative and complex IT-products. Another 241
questionnaires were collected within this additional sample.
About 25% (463 out of 1798) of the rms contacted turned out
not to be suitable respondents. From the main sample and the
additional sample, data are obtained from 788 (547+ 241) ITbuying rms. About 25% (183 out of 788) of the respondents were
willing to ll out a second questionnaire regarding the purchase
of a different IT-product, in most cases from a different supplier.
In these cases, another questionnaire was left at the site of the
rm and returned by mail. In total, the data set thus consists of
971 (547+ 241 +183) transactions, of which 183 are second
transactions from the same buyer. The bulk of the questionnaires
were lled out between January and June 1995. The average
response rate to the CATI-interview was 67% (902 out of 1 335).
Multiplied with the eld response rate of 87% (788 out of 902), the
total response rate equalled 59% (788 out of 1335).
Problems were measured in a similar manner as in Kamann
et al. (2006).2 Respondents could indicate on the same ve-point
scale whether and which of 11 possible problems had occurred.
About 28% had no problems whatsoever (did not occur on all 11
items). Dening a problem as having scored at least one item with
a 3 or higher on the ve-point scale, we nd that in MAT95 58% of
transactions encounters one or more problems. Because there is
less detailed information about the kinds of conict resolution in
these data, we use the same courser set of steps as in the
construction data. The results are shown in Table 5.

2
In fact, the survey in Kamann et al. (2006) was highly inspired by the MAT95
survey in design and wording of the survey questions, to allow for precisely these
kinds of comparisons.

226

F. Tazelaar, C. Snijders / Journal of Purchasing & Supply Management 16 (2010) 221229

Table 5 shows that the overall percentage of transactions that

leads to arbitration, suspension or legal steps is a bit lower but in
the same order of magnitude as in the construction data (1.2%,
condence interval from 0.6% to 2.1%, versus 1.6% in the
construction data). One of the few clear differences that can be
seen is that the probability with which a next step is taken
decreases one step later in the IT-data (72% versus 49%, p o0.01).
Apparently, at least when compared to these IT-data, problems
and litigation are not that extreme in the construction sector
when compared to the IT-sector.
In our next comparison, we consider country differences.
Either because of differences in culture or in legal context,
countries might differ in the extent to which they follow the steps
up the dispute pyramid. Because the above IT-data from the
Netherlands have been collected in a similar way in Germany as
well, we can get a handle on the difference between The
Netherlands and Germany, keeping the sector constant.
EDV99 (in German: Elektronische DatenVerarbeitung 1999) :
Four years after MAT95, similar data on IT-purchasing were
collected in two regions in Germany: the Halle/Leipzig region in
the former East Germany and the Munich region in the former
West Germany). At the time of the survey both of the regions were
economically prosperous regions. The sample consisted of small
and medium sized rms with 4500 employees. The general setup
of the data collection was identical to MAT95. For EDV99, the
yellow pages were used as a sampling frame. The data collection
was conducted in two phases. First, a member of the research team
contacted the rm in the sample by phone to determine whether
the rm met the requirements to be part of the survey, and if this
was the case, whether the rm was willing to cooperate in the
survey. If a rm agreed to cooperate with the survey and met the
requirements, then a knowledgeable contact person who had been
responsible for purchases of information technology was selected
and an appointment was made for a face-to-face interview (if rms
refused a face-to-face interview a questionnaire was mailed to
them). The telephone interviews started in March 1999 and were
concluded in August 1999 (Berger et al., 2001/2, 2002). In total,
1702 appropriate rms were reached, and 1094 (64%) were willing
to participate in either a face-to-face or a mail survey. The EDV data
set contains complete and detailed data on 1019 IT-transactions
from 832 buying rms. From these 832 buying rms, 645 (78%)
provided data on a single IT-transaction and 187 (23%) provided
data on a second IT-transaction as well. Compared to rms in the
region of Halle/Leipzig, rms in the region of Munich provided
more second questionnaires (29% vs. 15%).
The number and kind of problematic transactions were
measured in a similar manner as in MAT95. We nd that in 435
out of the 1019 transactions ( 43%), at least one problem
occurred at least to a certain degree. Table 6 shows the dispute
pyramid for EDV99.
It is striking that the German data show a much lower percentage
of problematic transactions (43% versus 73% in MAT95 and 84% in
the construction data, po0.01 for both). This in itself shows that, at
least here, a comparison within sectors between countries in terms
of litigation leads to a larger difference than a comparison between
sectors within a country. However, as soon as a problem does occur,
the subsequent behaviour up the pyramid seems roughly equal to
the previous data sets. We get an estimate of 0.3% (95% condence
interval from 0.1 to 1.0) for arbitration, suspension, and legal steps.
This is still lower than in the construction data and in MAT95, even
signicantly so (both at p 0.04). The signicant difference
disappears if we calculate the proportion of cases in the top step
when compared to the cases in which a problem occurs.
Finally, we compare the construction industry results with a
data set with a somewhat different choice of transactions, albeit
using a similar survey.

NEVI97: Between 1993 and 1998 data were collected at

Utrecht University on the management of partnerships: purchasing transactions, development transactions and R&D alliances.
The sample was a convenience sample of generally larger rms in
The Netherlands and Germany in a variety of sectors, ranging
from telecom, medical equipment, electronics, military equipment, well engineering, railway transport, air trafc, computer
equipment, metal bres, metal tubes, steal manufacturing, food
manufacturing, and plastics. In total 23 rms participated
(Tazelaar et al., 1995; Blumberg and Vaessen, 1999). For each
participating company a stratied sample was taken from their
transaction base, basically ranging from relatively cheap and
simple to expensive and complex transactions. A total number of
318 transactions were extensively studied, so we have on average
about 14 transactions per rm. The questionnaire itself once again
followed a transaction or alliance chronologically, from specication, search, screening and selection of supplier/partner, negotiation and contracting, performance, problems encountered, and
conict resolution. Given that this is a convenience sample, we
have no response characteristics. The main advantage of the data
is that it once again had a setup that was similar to the other data
collections. Mokkens H equals 0.57, so we can again think of the
conict management in these data as following subsequent steps
on the dispute pyramid. Table 7 shows the results.
Overall, we see that the NEVI97 data shows a percentage of
arbitration, suspension, or legal steps of 0.6%, lower than the
construction data (1.6%) and the MAT data (1.2%), but higher than
in EDV99 (0.4%). However, none of these differences are signicant,
which is probably at least partly due to the sample size that is
necessary to be able to nd differences that are this small
(typically, being able to nd a signicant difference between 0.6%
and 1.6% would necessitate a sample size of about 5 000, and for
being able to discern between 0.6% and 0.4% already about 50 000).

5. Dispute resolution in perspective: discussion of results

For further clarication, we show several of our key indicators
(based on Tables 3,57) simultaneously in Table 8, now comparing conict management with the total number of transactions in
the sample (instead of calculating the probability to go up another
step), once again in line with suggestions by Sarat and Grossman
(1975), Miller and Sarat (19801981), and Sarat (1984).
Table 8 reveals that the conict resolution in the construction
industry shows dispute percentages that are comparable with
those in IT (MAT95) but are generally on the high side, albeit often
not up to the level of statistical signicance. The percentage of
problems and the discussion about them shows the largest
difference with the other data sets. However, this difference
becomes smaller as we reach higher levels up the pyramid. Still,
the measured percentage of arbitration, suspension, and legal steps
is highest in the construction data, which supports (at least to some
extent) the general intuition that the construction industry is
Table 8
Percentages of transactions reaching more severe stages in the dispute resolution
process, as compared to the overall number of transactions.

Arbitration, suspension, legal steps

Delay payment, claim damage
Refer to contract, set deadlines
Discuss problems
There is a problem
Total number of transactions

WELL06

MAT95

EDV99

NEVI97

1.6
17.4
33.0
67.4
83.7
(100)

1.2
22.0
43.0
59.9
73.1
(100)

0.4
7.6
19.7
26.8
42.7
(100)

0.6
8.4
18.5
60.4
66.9
(100)

F. Tazelaar, C. Snijders / Journal of Purchasing & Supply Management 16 (2010) 221229

Table 9
Percentages of arbitration, suspension or legal steps, compared to other steps on
the pyramid (i.e., ratio of the frequency of arbitration, suspension or legal steps to
frequency of other steps on the pyramid).

Delay payment, claim damage

Refer to contract, set deadlines
Discuss problems
There is a problem
Total number of transactions

WELL06

MAT95

EDV99

NEVI97

9.0
4.7
2.3
1.9
1.6

5.6
2.9
2.1
1.7
1.2

5.2
2.0
1.5
0.9
0.4

7.7
3.5
1.1
1.0
0.6

harsher. This can be seen more clearly in Table 9, where we

compare the top level of the pyramid to the lower levels.
To clarify the results in Table 9, rst note that the number of
cases of arbitration, suspension or legal steps constitutes 1.6% of
the total number of transactions, as mentioned before. Furthermore, 1.9% of the cases in which there was a problem lead to
arbitration, suspension, or legal steps. Comparing across data sets,
we see that in the construction data (WELL06) the probability to
end up with arbitration or worse is highest, irrespective of the
base level with which it is compared (in each row, the highest
percentage is in the WELL06 column). Again, because the
differences are small and the frequencies are small as well,
differences are typically not signicant and some care should be
taken when interpreting these data as universally in favour of
showing the problematic nature of the construction industry.
Nevertheless, this does give at least some rigorous empirical
backing to the claim that the construction industry is more
litigation-prone than other sectors. All in all, our data lend some
support for the claim that the construction industry is more
litigation prone, although not as extremely as one might think
based on previous reports.
There are however several issues that should be taken into
account when considering these data. First, the (slightly) higher
litigation percentages might be the consequence of the fact that
transactions in the construction industry have different characteristics. For instance, they might be generally more risky, or more
often between partners that do not know each other. A logical
follow-up analysis would be to control for such characteristics.
First steps along these lines are in principle possible with the
presented data sets.
A second and related issue is the time lag between the four
surveys in our analysis. The data collections have been carried out
in different time periods, ranging from 1994 to 2006, which might
inuence the results given the technological and socio-economic
changes in that period. Earlier investigations of this kind suggest
that there might not be such an obvious difference. In an earlier
paper we analyzed transaction data from three separate surveys
more than 2000 transactions in total- showing that the average
amount and degree of problems encountered in Dutch ITtransactions was stable over the period 19872003 (Snijders and
Tazelaar, 2005). The suggested reason for this stability was that
although complexity and technological uncertainty has increased,
managers expertise and monitoring capacity has likewise
increased, leading to a zero net effect over time.
A third issue is that although the different samples are
stratied in order to be able to arrive at representative estimates
with smaller overall samples, this makes the results dependent on
the way in which these strata are weighed. The presented results
are unweighted, but one might arrive at different results given a
different weighting scheme (although the weighting schemes we
tried did not lead to intrinsically different ndings). To some
extent this highlights once again that a logical follow-up analysis
would be to control for characteristics of the transaction so that
effects of different weighting schemes can be minimized.

227

As a fourth issue one should note that a main difculty in

conict resolution and litigation research remains that the
percentage of severe conicts is small when compared to the
number of transactions (here in the order of magnitude of 1%).
This makes the minimum sample sizes for standard survey
sampling of transactions within organizations large (typically in
terms of several thousands rather than hundreds) and it makes
nding differences when sample sizes are in the hundreds
unlikely. Another design would be to count formal litigations
and to estimate the number of transactions out of which these
formal litigations have materialized. However, this is likewise
problematic how do you for instance estimate the number of
transactions per rm? With hindsight, it is perhaps not that
surprising that not many empirical contributions on the topic
exist. Here, we have tried to make a rst step in the right
direction, but more conclusive empirical evidence will need even
larger sample sizes and a careful sampling scheme.
Finally, our empirical work is restricted to the Dutch and
(partly) German context. Although our empirical evidence is to
our knowledge one of the few rigorous data collections on the
topic, the question remains what the scope of our ndings is.
Although this remains an empirical question, we feel that one of
the more general conclusions we can draw is that next to sectoral
differences, cross-national differences play a role of similar
importance. Countries with less efcient legal systems make it
harder for buyer and supplier to make agreements legally
enforceable (North, 1990). Moreover, in some countries more
than others, progress has been made during the last decades in
formulating and establishing sector-specic systems of rules,
regulations and procedures, thus providing a higher level of
institutional embeddedness (Johannisson et al., 2002). Such
systems of rules and regulations provide rms with extra
opportunities for managing ex post problems in a less harsh
way, for example in the form of mediation or alternative dispute
resolution instead of litigation. There might be country differences with respect to less formal reasons as well. For instance,
Henrich et al. (2006) nd that the degree of sanctioning
uncooperative behaviour varies substantially among human
societies, and that culture plays a crucial role in economic action
and sanctioning of uncooperative behaviour. More closely related
to our own analyses, Rooks and Matzat (2010) examined the
moderating role of culture by comparing effects of social
embeddedness on trust and conict regulation in Germany and
The Netherlands, relating these to Hofstedes cultural femininity
dimension (cf. Hofstede, 1980; Ulijn et al., 2004; Leung et al.,
1990). Further empirical research on an even larger and broader
scale will be necessary to gain more insight in the exact reasons
for cross-sector, cross-regional and cross-national differences.

6. Conclusion
Conicts and litigation have been claimed to proliferate in the
construction industry. A closer inspection of the literature reveals
that the empirical basis of this proliferation claim is at best
circumstantial. Empirical generalizations are frequently formulated based on measurements of a limited number of interesting
rather than representative cases. In addition, larger scale
empirical data collections usually measure practitioners perceptions about the industry. However useful these might be, these
perceptions might not coincide with the actual state of affairs in
the eld (cf. Bryde, 2008).
If one considers the literature on conict resolution and
litigation in a somewhat broader business context, it becomes
clear that during the last ve decades many researchers have been
impressed by what has been called the litigation explosion in

228

F. Tazelaar, C. Snijders / Journal of Purchasing & Supply Management 16 (2010) 221229

both the US and the UK. The main problem with the perceived
litigation explosion is that it should be considered relative to the
growing number of transactions. If one does so, it appears that it is
usually quite exceptional for disputes between companies to end
up in court, even after controlling for the increased competition in
business, increased internationalization, and the increased litigiousness in the world of business (cf. Deakin and Michie, 1997).
We used the metaphor of a dispute pyramid for the various steps of
conict resolution (Sarat and Grossman, 1975; Miller and Sarat,
19801981; Sarat, 1984). Such a pyramid represents the stages
through which events pass in the dispute process.
To add to the (representative) empirical evidence on conict
resolution, we rst analyzed the construction industry on the
basis of a data set consisting of 448 contractorsubcontractor
transactions in The Netherlands, as introduced in Kamann et al.
(2006) and Welling (2006). A substantive part of that survey was
devoted to measuring the kind and severity of the problems that
were encountered during the transaction and to measuring
conict resolution. Secondly, we confronted the results of our
analyses of these construction data with similar analyses of data
from three other surveys. One with conict resolution data from
the IT-sector in The Netherlands, a similar data set with conict
resolution data from the German IT-sector, and nally a more
general data set of cooperative interactions of Dutch and German
rms from various sectors. In total, we analyzed more than 2750
transactions in detail.
The results show some interesting and subtle differences in the
extent to which transactions lead to conicts and conict
management across industries and countries. Firstly, all four data
sets support the pyramid model: the next phases in conict
resolution are taken step-by-step, without skipping steps in
between. This signies that where problems occur within
transactions, conict resolution in general is characterized by a
consistent pattern: most of the time lighter steps are taken before
heavier forms of conict resolution.3 Those who associate harsh
business with wilder and less predictable conduct in conict
resolution do not nd much empirical evidence in our data.
Secondly, it is indeed the case that the percentage of transactions
that leads to some kind of problem is highest in the construction
industry across our data sets: 84% versus 73 and 43 in IT and 67 in
the general data set. However, as we progress up the pyramid, the
differences become less pronounced or disappear altogether.
Thirdly, it is striking that there is a big difference between the two
data sets on IT-transactions. The German IT-data show a much
lower percentage of problems. Note that the size of these
differences is about as large as the difference between the
construction industry and the other data sets in general. In other
words, it is not obvious that the main nding here lies in the
prevalence of conicts and litigation in the construction industry
country differences, either culturally or institutionally, are likely
to play just as big a role. On the other hand, as soon as problems
are encountered, the subsequent conict resolution gures are
not that different from the other data sets. This also shows the
usefulness of the pyramid metaphor: it allows one to show more
precisely where differences in conict likelihood and resolution
occur. Finally, we nd that about 1.6% of construction transactions
leads to serious conicts in the sense that they lead to arbitration,
suspension and/or legal steps. That is a larger percentage than in

3
Strictly speaking the data from the four surveys do not present the
opportunity to formulate statements about specic sequences in time. One can
only formulate statements about the specic combinations of the various steps
that are taken. Ideally, the study of transformations of disputes should be
longitudinal (cf. Felstiner et al., 19801981). Only a pilot study conducted before
the four surveys were carried out reveals that most of the time lighter steps in
conict resolution were indeed taken before the heavier steps.

the other data sets (there the estimates are 1.2, 0.4, and 0.6).
When we compare only with the transactions that lead to some
kind of conict, largely the same image appears: 1.9% versus 1.7,
0.9, and 1.0. Taken together, these results are one of the few
rigorous statistics on conict resolution and litigation, putting
litigation and conict resolution in the appropriate perspective.
Above all, our results show that one should remain cautious and
prudent about the harshness of the construction industry.

Acknowledgements
We acknowledge nancial support from the Netherlands
Organization for Scientic Research (PGS 50-370), the German
Science Foundation (DFG AZ. Vo 648/2-1), and the Netherlands
Association for Purchasing Management (NEVI/NRS).

References
Annual Report of the Technology and Construction Court, 2008. HMRC, London
(UK).
Barton, J.H., 1974. Behind the Legal Explosion. Stanford Law Review 27, 567584.
Batenburg, R., 1997. The external management of automation 1995: the codebook
of MAT95. ISCORE Paper 58. Department of Sociology, Utrecht University,
Utrecht (NL).
Batenburg, R.S., Raub, W., 1995. The External Management of Automation 1995
[Data Set]. Utrecht University, Utrecht (NL).
Batenburg, R., van de Rijt, A., 1998. The external management of automation 1998:
The codebook of MAT985. ISCORE Paper 155. Department of Sociology, Utrecht
University, Utrecht (NL).
Beale, H., Dugdale, T, 1975. Contracts between businessmen. Planning and the use
of contractual remedies. Journal of Law & Society 9 (2), 4560.
Berger, R., Kropp, P., Voss, T., 2001/2. Das Management des EDV-Einkaufs. Soziale
Einbettung und Gestaltung wirtschaftlicher Transaktionen. Eine empirische
Untersuchung am Beispiel der Beschaffung informationstechnischer Leistungen und Produkte durch Klein und Mittelbetriebe. Abschlussbericht and die
deutsche Forschungsgemeinschaft. [Publication in German. Translation: The
management of IT-purchasing. Social embeddedness and transaction char Soziologie (2001), Leipzig (Ger.)/
acteristics]. Arbeitsbericht des Instituts fur
ISCORE paper No. 198 (2002), Utrecht University, Utrecht (NL).
Berger, R., Kropp, P., Voss, T., 2002. Das Management des EDV-Einkaufs 1999.
Codebook [Publication in German. Translation: The management of ITpurchasing] ISCORE paper No. 199, Utrecht University, Utrecht (NL).
Blegvad, B.M., 1990. Commercial relations, contract and litigation in Denmark. A
discussion of Macaulays theory. Law and Society Review 24 (2), 397411.
Blumberg, B.F., Vaessen, P.M.M., 1999. Samenwerking tussen inkoper en
leverancier [Publication in Dutch. Translation: Cooperation between buyer
and supplier: Codebook of The Management of Partnerships (MOP)]. ISCORE
paper, Utrecht University, Utrecht (NL).
Bresnen, M., Marshall, N., 2000. Partnering in construction: a critical review of
issues, problems and dilemmas. Construction Management and Economics 18,
229237.
Brooker, P., Lavers, A., 1997. Perceptions of alternative dispute resolution as
constraints upon its use in the UK construction industry. Construction
Management and Economics 15, 519526.
Bryde, D.J., 2008. Is construction different? A comparison of perceptions of project
management performance and practices by business sector and project type.
Construction Management and Economics 26, 315327.
Buskens, V., Batenburg, R., 2000. The external management of automation 1998:
codebook for the combined data from the Netherlands and Germany. ISCORE
Paper 175, Department of Sociology, Utrecht University, Utrecht (NL).
Colledge, B., 2005. Relational contracting creating value beyond the project. Lean
Construction Journal 2 (1), 3045.
Cox, A., Thompson, I., 1997. Fit for purpose contractual relations; determining a
theoretical framework for construction projects. European Journal of Purchasing and Supply Management 3, 127135.
Davis, K., Letbetter, W.B., Burati, J.L., 1989. Measuring design and construction
quality costs. Journal of Construction Engineering and Management 115 (3),
385400.
Deakin, S., Michie, J. (Eds.), 1997. Studies in Economics, Management and Law.
Oxford University Press, Oxford (UK).
Dunworth, R., Rogers, J., 1996. Corporations in court. Big business litigation in the
U.S. federal courts: 19711991. Law and Social Inquiry 21, 497592.
Fellows, R.F., 1992. Karming conict. In: Fenn, P., Gameson, R. (Eds.), Construction
Conict Management and Resolution. E & FN SPON/Chapman & Hall, London
(UK), pp. 122127.
Felstiner, W.L.F., Abel, R.L., Sarat, A., 19801981. The emergence and transformation of disputes: naming, blaming, claiming. Law and Society Review 15 (3/4),
631654.

F. Tazelaar, C. Snijders / Journal of Purchasing & Supply Management 16 (2010) 221229

Fenn, P., 1991. Managing Corporate Conict and Resolving Disputes on Construction Projects. In: Proceedings of the 7th Annual Conference Association of
Researchers in Construction Management ARCOM, University of Bath, Bath
(UK), pp. 2233.
Galanter, M., 1983. Reading the landscape of disputes: what we know and dont
know (and think we know) about our allegedly contentious and litigious
society. UCLA Law Review 31 (1), 471.
Galanter, M., Rogers, J., 1991. The transformation of American business disputing.
Some preliminary observations. Institute of Legal Studies, University of
Wisconsin. Working Paper DPRP 10-3, Madison (Wisc.).
Galanter, M., Macaulay, S., Palay, T., Rogers, J., 1991. The transformation of American
business disputing. A sketch of the Wisconsin Project. Institute of Legal Studies,
University of Wisconsin. Working Paper DPRP 10-6, Madison (Wisc.).
Gibbons, E.N., 2007. Effects of litigation in the construction industry: stratication
and insolubility. In: Proceedings for the construction and building research
conference of the Royal Institution of Chartered Surveyors, Georgia Tech.,
Atlanta USA: 67 September 2007. RICS in association with College of
Architecture Georgia Institute of Technology, London/Atlanta.
Harmon, K.M.J., 2003. Resolution of construction disputes: a review of current
methodologies. Leadership Management Engineering 3 (4), 187201.
Henrich, J., McElreath, R., Barr, A., Ensminger, J., Barrett, C., Bolyanatz, A., et al.,
2006. Costly punishment across human societies. Science 312, 17671770.
Hofstede, G., 1980. Cultures Consequences: International Differences in WorkRelated Values. Sage, Beverly Hills (CA).
van Houtte, J., Duysters, A., Thieleman, V., 1995. Gebruik van het recht in de
ondernemingen [translation:
zakenwereld. Over juristen in grote industriele
The use of law in business world. On lawyers in big industrial enterprises].
Acco, Leuven (B) (in Dutch).
Jettinghoff, A., 2001. Het komt zelden voor. Beheersing van klachten en geschillen
in relaties tussen bedrijven [Publication in Dutch, with a summary in English.
Translation: Where legal steps are rare: governing complaints and disputes in
business relations]. Metajuridica Publications, in cooperation with Universitaire Pers Maastricht, Nijmegen (NL).
Johannisson, B., Ramrez-Pasillas, M., Karlsson, G., 2002. The institutional
embeddedness of local inter-rm networks: a leverage for business creation.
Entrepreneurship and Regional Development 14, 297315.
Kadefors, A., 2004. Trust in project relationships inside the black box.
International Journal of Project Management 22 (3), 175182.
Kamann, D-J., Snijders, C., Tazelaar, F., Welling, D.Th., 2006. The ties that bind:
buyersupplier relations in the construction industry. Journal of Purchasing
and Supply Management 12 (1), 2838.
Khalfan, M.M.A., McDermott, P., Swan, W., 2007. Building trust in construction
projects. Supply Chain Management: an International Journal 12 (6), 385391.
Latham, M.S., 1994. Constructing the Team: Final Report of the Government/
Industry Review of Procurement and Contractual Arrangements in the UK
Construction Industry. HMSO, London (UK).
Lavers, A., 1992. Construction conict: management and resolution analysis and
solutions. In: Fenn, P., Gameson, R. (Eds.), Construction Conict Management
and Resolution. E & FN SPON/Chapman & Hall, London (UK), pp. 320.

229

Leung, K., Bond, M.H., Carment, D.W., Krishnan, L., Liebrand, W.B.G., 1990. Effects
of cultural femininity on preference for methods of conict processing: a
cross-cultural study. Journal of Experimental Social Psychology 26, 373388.
Macaulay, S., 1963. Non-contractual relations in business: a preliminary study.
American Sociological Review 28 (1), 61.
Macaulay, S., 1985. An empirical view of contract. Wisconsin Law Review 1985 (3),
465482.
Manning, B., 1976. Hyperlexis, our national disease. Northwestern University
Law Review 71, 767782.
Miller, R.E., Sarat, A., 19801981. Grievances, claims, and disputes: assessing the
adversary culture. Law and Society Review 15 (3/4), 525566.
Mokken, R.J., 1971. A Theory and Procedure of Scale Analysis. With Applications in
Political Research. De Gruyter, Mouton, New York/Berlin.
Mokken, R.J., 1996. Nonparametric models for dichotomous responses. In: van der
Linden, W.J., Hambleton, R.K. (Eds.), Handbook of Modern Item Response
Theory. Springer, New York, pp. 351368.
North, D.C., 1990. Institutions, Institutional Change and Economic Performance.
Cambridge University Press, Cambridge (UK).
Rooks, G., Matzat, U., 2010. Cross-national differences in effects of social
embeddedness on trust: a comparative study of German and Dutch business
transactions. The Social Science Journal 47, 4568.
Rooks, G., Snijders, C., 2001. The purchase of information technology products by Dutch
SMEs: problem resolution. The Journal of Supply Chain Management 37 (2), 3442.
Saad, M., Jones, M., James, P., 2002. A review of the progress towards the adoption
of supply chain management (SCM) relationships in construction. European
Journal of Purchasing and Supply Management 8, 173183.
Sarat, A., 1984. The litigation explosion, access to justice, and court reform:
examining the critical assumptions. Rutgers Law Review 37, 319336.
Sarat, A., Grossman, J.B., 1975. Courts and conict resolution: problems in the
mobilization of adjudication. American Political Science Review 69 (4), 12001217.
Snijders, C., Tazelaar, F., 2005. Five counterintuitive ndings in IT-purchasing.
Journal of Purchasing and Supply Management 11, 8396.
Stipanowich, T.J., 1997. At the cutting edge: conict avoidance and resolution in the
US construction industry. Construction Management and Economics 15, 505512.
Tazelaar, F., Vaessen, P.M.M., Blumberg, B.F., Raub, W., 1995. Samenwerking tussen
inkoper en leverancier. Verslag van een vooronderzoek naar het management
van inkooptransacties [Publication in Dutch. Translation: Cooperation
between buyer and supplier. A pilot study report on the management of
purchase transactions in the Netherlands]. ISCORE paper, Universiteit Utrecht
i.s.m. NEVI Research Stichting, Utrecht (NL).
Ulijn, J.M, Lincke, A., Wynstra, F., 2004. The effect of Dutch and German culture on
negotiation strategy: an exploratory study to compare innovation and
operations contexts. International Negotiation 9 (2), 201228.
Welling, D.Th., 2006. Bouwen op een gemeenschappelijk verleden aan een
succesvolle toekomst. Een onderzoek naar samenwerkingsproblemen in
interorganisationele relaties vanuit een sociaal-economisch netwerkperspectief [Publication in Dutch, with a summary in English. Translation: Building a
successful future on a shared past]. Dissertation. Rijksuniversiteit Groningen,
Groningen (NL).