The relationship between board characteristics and diversification: do the differences between one- and two-tier boards have an effect? M.P.H.J. (Michel) Witte Final Version: 19-01-2012
The relationship between board characteristics and diversification: do the
differences between one- and two-tier boards have an effect?
M.P.H.J. (Michel) Witte
Final Version: 19-01-2012
The relationship between board characteristics and diversification: do the
differences between one- and two-tier boards have an effect?
Master Thesis of the department of Organization & Strategy
Tilburg School of Economics and Management
Tilburg University
ANR: 115960
Name: Michel Witte
Supervisor: Dr. E. Golovko
Second Reader: Drs. J.M. Dumas
Date of defense: 09-02-2012
I
Management Summary
What influences the decisions to diversify the product portfolio or stick to the core product has been the
subject of research by strategic management scholars for some time. Since this decision is made by the
board, it can be seen as not solely based on incentives on firm level, as most managers personally tend to
profit as well from diversifying. It can reduce risk and increases job security (Baysinger & Hoskisson, 1990).
Furthermore, some literature states that strategic choices are partially predicted by managerial background
characteristics (e.g. Hambrick & Mason, 1984). Therefore board characteristics as, for instance, age,
education and gender can have an effect on diversification.
This study tried to determine the exact role of the management and supervisory boards as well as the
characteristics of the people on these boards, in the strategic choice for diversification. Special attention
was given to the difference between one-tier and two-tier boards. The decision to diversify has been
looked at from several different perspectives. This paper merely used a strategic management perspective
and has combined arguments from the Resource-Based View, Agency theory and Stewardship theory.
These theories were used to formulate five hypotheses that were tested cross-sectional using a sample of
165 corporations from Germany, the Netherlands, the UK, Ireland, Belgium and Luxemburg.
This paper analyzed the differences between boards with respect to diversification by giving the answer to
the two research questions: first, which board characteristics affect the decision to engage in product
diversification and secondly, does the difference in board structure have an effect on this decision?
The results of the empirical study of the sample, which included the statistical testing of the hypotheses, as
well as t-tests, gave a clear answer to the first question. The data gave no support for the influence on
diversification by any of the board characteristics, although these were proposed by other scholars and as
such could be reliably retested in this paper.
The answer to the second question, the main topic of this research, the relationship between Board type
and diversification, is less simple. The correlation matrix and t-test showed an, albeit weak, positive
relationship. The regression analysis, however, turned out insignificant. Therefore it can be concluded that
there are indeed differences between one-tier and two-tier boards. But they cannot be explained by the
regressions used in this research, which is why in this study it is suggested how this relationship can
possibly be re-tested.
II
Foreword
This master thesis is the final product of over six years of studying at Tilburg University. It was a memorable
period, to say the least. The experiences I had over the years, the labor I put in and the friends I have made,
all lead to the final goal: graduation.
First, I have a more serious note. The inspiration for making this Master thesis mainly came from one man:
Dr. A. van Oijen. His inspirational classes on Corporate-Level Strategy made me think about the possibilities
to write my thesis on a topic in this field. After I consulted him whether my initial idea was good enough, he
told me straight up that I would be nearly impossible to make a decent thesis from my first research
proposal. Nevertheless he stimulated me to look further in the field of diversification. The end-result lies in
front of you. I hope it speaks for itself.
It is thanks to many people that have come this far. First of all, much gratitude goes to my parents. Not only
did they provide the (much-needed) financial support, but their wisdom and occasional words of
motivation and encouragement were the main reason I could eventually complete my studies. Special
thanks go to my brothers Philippe, for stealing my idea and using it in his own Master thesis and Didier for
the continued interest in what I was doing. For keeping me busy, while not working on this thesis I thank
my friends of R.A.F. and CCT. Without them my time in Tilburg would have been a lot more boring.
Next I also owe many thanks to my university supervisor Dr. E. Golovko for her advice and support. Without
it I would probably be still (re-)writing. Finally I would like to thank Drs. J.M. Dumas for taking seat in the
examining committee.
Michel Witte
Tilburg, January 2012
III
Table of Contents
MANAGEMENT SUMMARY .......................................................................................................................................... I
FOREWORD ................................................................................................................................................................ II
TABLE OF CONTENTS ................................................................................................................................................. III
1. INTRODUCTION ................................................................................................................................................. 1
2. THEORY AND BACKGROUND ............................................................................................................................. 3
2.1 THE MOTIVES TO DIVERSIFY ......................................................................................................................................... 3 2.1.1 Resource-Based View .................................................................................................................................. 3 2.1.2 Agency Theory ............................................................................................................................................. 5 2.1.3 Stewardship Theory .................................................................................................................................... 5
2.2 EMPIRICAL EVIDENCE ON THE FACTORS THAT INFLUENCE DIVERSIFICATION ............................................................................ 6 2.2.1 What affects diversification? ..................................................................................................................... 6 2.2.2 The influence on the level of diversification by different characteristics of the board ............................... 7
3. DATA AND METHODS .......................................................................................................................................15
3.1 SAMPLE SELECTION AND DATA COLLECTION .................................................................................................................. 15 3.2 VARIABLE SPECIFICATION AND MEASUREMENT............................................................................................................... 16 3.3 EMPIRICAL MODEL ................................................................................................................................................... 20
4. RESULTS ...........................................................................................................................................................22
4.1 DESCRIPTIVE STATISTICS ............................................................................................................................................ 22 4.2 PRELIMINARY T-TESTS ............................................................................................................................................... 24 4.3 REGRESSION RESULTS ............................................................................................................................................... 25
5. CONCLUSIONS AND RECOMMENDATIONS .......................................................................................................30
5.1 SUMMARY AND CONCLUSION ..................................................................................................................................... 30 5.2 LIMITATIONS .......................................................................................................................................................... 30 5.3 SUGGESTIONS FOR FUTURE RESEARCH ......................................................................................................................... 31
REFERENCES ..............................................................................................................................................................32
APPENDICES ............................................................................................................................................................ VIII
APPENDIX 1: INDUSTRY DUMMY LABEL CATEGORIZATION ...................................................................................................... VIII APPENDIX 2: THE COMPANIES IN THE SAMPLE ....................................................................................................................... IX
1
1. Introduction
The strategic decision whether to diversify the product portfolio or focusing on the core business has been
since long a source of heated debates. The decision to diversify ultimately lies in the hands of the board.
However, this decision can be seen as not solely based on incentives on firm level, as most managers
personally tend to profit as well from diversifying, as it reduces risk for them regarding bonuses and
increases job security (Baysinger & Hoskisson, 1990). Furthermore, strategic management literature states
that strategic choices are partially predicted by managerial background characteristics (Hambrick & Mason,
1984). Therefore board characteristics, such as age, education and gender can have an effect on
diversification.
Westphal and Frederickson (2001) found that supervisory board members exert relatively little influence
over major decisions such as corporate diversification. Still, their influence should not be neglected as
Anderson, Bates, Bizjak and Lemmon (2000) found. They did research on whether an inadequately
performing supervisory board led to an increased incentive to diversify, but found no evidence that this
was the case. Furthermore, Goranova, Alessandri, Brandes and Dharwadkar (2007) even suggested that a
good supervisory board, one that is strict and has a ‘hands-on’ mentality, even results in increased
incentive alignment and therefore is negatively related to corporate diversification. Therefore, if, for
instance, people on the management board were best friends in university, this could be potentially
harmful for the company. These findings are of particular interest for the so-called two tier board system as
is in place in countries as Germany and the Netherlands. Ruigrok, Peck and Keller (2006) found that
supervisory boards greatly mitigate the harmful behavior that can go hand in hand with unnecessary
diversification.
This study will try to determine the exact relationship between the strategic choice for diversification and
the role of boards, as well as the influence of the characteristics of the people on these boards in this
decision. This topic was recently investigated to some extent by Chen, Dyball and Wright (2009). Special
attention will be given to the difference between one-tier and two-tier boards. In one-tier boards there is a
division between executive directors and independent directors, the latter forming the ‘control’ mechanism
in the board. Most of the time, these independent directors have a lot of experience with other large
companies. In the two-tier board system the controlling element is a supervisory board, which, in most
cases, has law-mandated representation from the employees. Social and legal developments have
influenced that the one-tier board system in prevalent in the so called Anglo-Saxon countries, such as the
UK and Ireland. The two-tier board system is mainly used in Rhineland countries, such as Germany and the
Netherlands. These differences could potentially heavily influence the role of management regarding the
diversification decision.
2
To summarize, there are two central problem statements that will be used to come to conclusions:
Which board characteristics affect the decision to engage in product diversification?
Does the difference in board structure have an effect on this decision?
To extract the answer to this problem statement, some research questions will be posed to systematically
come to sub-conclusions, which in turn can give a wide understanding of the problem area.
First, which characteristics have a significant impact on eventual diversification decision?
In order to understand research on diversification, Ramanujam and Varadarajan (1989) suggested that four
influences induce a firm to diversify: the general environment, the industry environment, firm
characteristics and overall performance. The specific characteristics of the firms are an interesting
influence. It covers everything from how the company is governed to demographics, such as the average
age. Especially in this last category there are variables that this study will try to investigate and will focus on
the demographics of the people on the board in detail. Furthermore, top management teams’ demographic
characteristics can help to predict changes in diversification (Wiersema & Bantel, 1992). From this starting
point the influence of board characteristics will be explored further.
Secondly, what is the influence of the type of board system on diversification decisions?
Hoskisson and Hitt (1990) established that managerial motives for diversification may exist independent of
resources and incentives. Therefore these may serve as a motive for increased diversification. They also
found that these motives could be limited by governance. Mechanisms such as a strict board of directors
and corporate control limit tendencies to over diversify. As already mentioned, the difference between
one-tier and two-tier boards is interesting in this case. And from a broader perspective this also means
difference between countries, namely Anglo-Saxon countries and Rhineland countries. Ground-breaking
work in the clarification of the distinctions between these two models was done by Albert (1993). He
described the Anglo-Saxon model as market-oriented, more focused on short-term profits. In contrast the
Rhineland model is more society-oriented and opts for a more sustainable, future-oriented approach. In
essence, the only goal of an Anglo-Saxon company is to maximize profits. This is why board remuneration in
these countries often is linked to bonuses which, as already mentioned, could lead to misaligned
managerial motives. This as opposed to the Rhineland modeled companies, which opt for a more
sustainable way of doing business. Striking differences between the Rhineland and Anglo-Saxon are for
instance the obligation for large companies to have employee representation on the board (Koen, 2005)
and the increased involvement of banks (Aguilera & Jackson, 2003). As such, the board system will receive
extra attention in this study.
3
2. Theory and Background
This chapter will give a first introduction on why firms can decide to diversify. Three perspectives from
management literature will be looked at. Next, empirical studies will be analyzed to establish which
variables have proven to influence the diversification decision. Then also the specific focus of this study will
be discussed in-depth. The differences in one-tier and two-tier management boards, differences in
governance and also the parallels between this difference and the division between Anglo-Saxon and
Rhineland countries will be described.
2.1 The motives to diversify
The decision to diversify has been looked at from several different perspectives. Scholars in industrial
organization, strategic management and finance have all looked at the upsides and downsides of
diversification. This paper merely uses a strategic management perspective, stemming greatly from ground-
breaking work by Rumelt (1974).
There are two main motives suggested in the literature on why firms diversify. They will do so for either
synergistic or financial reasons (Amit & Livnat, 1988). However, the rationale behind diversification focuses
on three mayor stances that have dominated the strategic management literature on this subject: the
Resource-Based View, Agency theory and, more recently, Stewardship theory.
2.1.1 Resource-Based View
The resource-based view (RBV) has been an important theory in the strategic management literature for
some time now. It argues that firms diversify in response to excess capacity in productive factors
(Montgomery, 1994). Though this description does not completely cover how and why this excess capacity
is utilized. There are several lines of reasoning that scholars have used to argument for this statement. For
instance, some argue that resources are to some degree transferable across products and industries.
Because these are valuable, it may not be optimal for a firm to go slowly out of business as sales of its
products decline. Rather it may be better to move into areas where potential demand for new products is
greater than existing output, by diversifying (Chenhall, 1984; Matsusaka, 2001). This notion is supported by
Hoskisson and Hitt (1990), who found that excess capacity of tangible assets, such as plants and
warehouses, can be utilized for very closely related products. However, sometimes this reasoning can prove
to be not valid anymore. Namely, in a situation where the firm is financially in such dire straits that
divesting parts of the company to raise cash is the only option for the company. In which case decreases in
diversification can also be associated with financial distress (Denis, Denis & Sarin, 1997).
4
An extension of this theory claims that firms not only use resources to diversify out of self-preservation but
also to exploit operating synergies (Amit & Livnat, 1988; Ramanujam & Varadarajan, 1989). A tactic that is
often used by multi-business organizations, which try to achieve economies of scope by sharing rare and
costly to imitate strategic assets among their businesses (Wang & Barney, 2006). Examples of this are
sharing customer data and marketing information, or joint production by two entities at the same location
(Klein & Saidenberg, 2000).
Others argue that firms have different incentives if excess capacity occurs. They propose that firms diversify
in order to utilize excess resources which could not be otherwise sold or leased because of the high
transaction costs that this would entail (Teece, 1982; Chatterjee & Wernerfelt, 1991; Fox & Hamilton,
1994). Especially companies in niche markets or with highly specific assets are likely to diversify for this
reason. Companies can choose to sometimes only break-even on the non-core product, as it enables them
to amortize their assets over more units.
The final argument RBV proposes is that diversification is used to reduce risk (McDougall & Round, 1984,
Hoskisson, Hitt & Hill, 1991). It can be a defensive move by management to mitigate undesirable
characteristics in a firm's dominant industrial environment (Wiersema & Bantel, 1992). For instance, firms
in an industry that is highly dependent on the state of the economy, such as luxury goods. These might be
tempted to also invest in something more secure, such as foods. However, firms can also have periods that
induce and reduce risk taking, depending on the level diversification and associated control system
attributes (Hoskisson, Hitt & Hill, 1991). A positive forecast for the economy might tempt companies to
differentiate less in comparison what would have happened in a normal situation.
Reducing risk can also be associated with the value of core firm resources (Wang and Barney, 2006). An
example of this reasoning is the situation for firms that are greatly affected by seasonality (Penrose, 1995
[1959]). For example an ice-cream shop can sell ice-cream during the summer, but sell skis from the same
building during the winter. This way the core firm resource, the building, is optimally used. Besides reducing
risk, another positive consequence of this strategy is that firms have more stable cash flow as seasonality
effects are smoothed. As a result firms can attempt to gain financial benefits from their ability to increase
leverage (Amit & Livnat, 1988). Furthermore, firms can use so-called cross-subsidization, meaning that the
firm uses its profits from one market to support predatory pricing activities in another (Penrose, 1995
[1959]). The losses that normally would have occurred can be covered by the extra income. Firms can use
this tactic so to keep their market share or try to push competitors out of the market, after which they can
bring back the price to a normal level.
5
Concluding, the resource-based view proposes four main reasons on why and how firms diversify. They do
so to reduce risk, to overcome transaction costs, to exploit operating synergies and sometimes out of self-
preservation of the firm to keep market-share in existing markets or get a toehold in new markets.
2.1.2 Agency Theory
Agency theory mainly focuses on the relationship managers have with the firm and the effect this has on
the company (Jensen & Meckling, 1976). A popular explanation of agency theory emphasizes the ‘dark side’
of diversification: firms are plagued with agency problems that allow managers to enter new businesses,
getting benefits in the process that they may reap at the expense of its shareholders (Montgomery, 1994;
Ramanujam & Varadarajan, 1989; Matsusaka, 2001). These private benefits may come from a variety of
sources. They may arise from prestige or better career prospects associated with running a more diversified
firm. Private benefits may arise because running a more diversified firm increases managers’ pay (Aggarwal
& Samwick, 2003). As already mentioned in section 2.1.1, diversification may allow cross-subsidization of
unprofitable divisions. This not only can stimulate the core product of the firm, but also can be a way to
cover up losses, as they might not appear on consolidated balance sheets (Klein & Saidenberg, 2000). This
practice can be harmful since the use of financial controls becomes more common as firms diversify
(Hoskisson, Hitt & Hill, 1993). This means that board members can influence the height of a bonus or
increase the likelihood of one being paid by means of creative accounting.
Beside direct benefits, managers can also benefit themselves indirectly. Managers may diversify to protect
their specific human capital from firm risk (Amihud & Lev 1981; Anderson, Bates, Bizjak & Lemmon, 2000).
If, for example, the current manager is about to be replaced by someone who would run the firm better
than him, he has an incentive to diversify into areas where he has a comparative management advantage
(Shleifer & Vishny, 1989). These scholars go even further by claiming that in some cases managers can
diversify to industries where their skills are essential. In this way they entrench themselves in the company
and increase job security.
Concluding, agency theory claims that managers have several reasons to diversify. They do so out of self-
interest: gaining prestige, to increase pay-checks, protecting their human capital or increasing job security
by entrenching themselves.
2.1.3 Stewardship Theory
The newest of the three stances is stewardship theory, that proposes that managers are not motivated by
individual goals, but rather are stewards whose motives are aligned with the objectives of their principals
(Davis, Schoorman & Donaldson, 1997). In other words, managers and staff work in the best interest of the
company and act accordingly. This creates some interesting reasons for diversification. For instance, Hyland
6
and Diltz (2002) suggested that diversifying firms have not engaged in as much research and development
as their non-diversifying counterparts. In order to grow or perhaps even maintain their current status, they
must buy growth in areas outside of where they are currently operating. Following somewhat the same
reasoning Montgomery (1994) found that young and growing businesses have plenty of profitable
opportunities in which to re-invest earnings. However, as businesses mature, these opportunities become
scarce, and managers begin to use cash flows from earlier innovative efforts to invest in other areas.
Another argument stemming from stewardship theory is that diversification can be used by managers to
increase profitability (McDougall & Round, 1984; Fox & Hamilton, 1994). However, this is a highly contested
argument since the increase in profitability caused by diversification has both been confirmed (Palepu,
1985; Amit & Livnnat, 1988) as well as rejected (McDougall & Round, 1984; Montgomery, 1994; Matsuaka,
2001). Furthermore it has been shown that the effect diversification has on performance can differ
between countries (Mayer & Whittington, 2003). Concluding, stewardship theory suggests that companies
diversify for several reasons: to ensure growth, to invest free cash or to increase profitability.
2.2 Empirical evidence on the factors that influence diversification
While the theory in the previous sections already gives a good overview on what exactly drives
management to diversify their company, there have also been some studies that have been testing these
theories statistically. Therefore, in this paragraph an overview will be given on the factors that have been
proven by management literature to have an influence on the diversification decision. In addition, this
study focuses particularly on the different characteristics of the board and proposes to add a variable,
namely the way the board is structured.
2.2.1 What affects diversification?
Strategic management literature, and to a lesser extent the financial literature, identified numerous factors
that affect diversification decision. Among these factors are also a lot of variables that have a link with how
and by whom the company is governed. There are two streams in the literature on governance: one that
focuses especially on the CEO and a stance that looks at governance in a much broader perspective.
Some scholars see the role of the CEO, as key decision maker, as vital in the process of change. Jin (2002)
and Field and Keys (2003) did research on CEO characteristics and decisions that affect firm risk. They found
that a higher level of non-diversifiable wealth the CEO has invested in the firm, the more likely that an
acquisition will be diversifying in nature. Thus, they provided evidence for the argument from resource-
based view theory that diversification is part of protecting their personal risk. Ruigrok, Peck and Keller
(2006) found even more evidence for this problem as they showed that the level of diversification, for
personal reasons, increases when a company combines the roles of the CEO and chairman.
7
Other researchers however, mainly focus on the role of governance. A subject that is heavily discussed
from this stance is the relationship between remuneration of board members and diversification. In
conjunction with the agency theory discussed earlier, increases in payment and incentive schemes have a
great correlation with diversification. Higher levels of pay, for instance, lead to increased diversification
(Anderson, Bates, Bizjak & Lemmon, 2000). Also the proportion received as a bonus will increase the
diversification in firms (Napier & Smith, 1987). But remuneration policies do not only cover direct monetary
incentives. Larger companies often offer packages with shares and options as a part of the total
remuneration. Board members in diversified firms tend to have lower stock ownership (Anderson, Bates,
Bizjak & Lemmon, 2000) which suggests support for the employment risk-reduction perspective (Goranova,
Alessandri, Brandes & Dharwadkar, 2007). This is a reason widely regarded as one of the problems
associated with diversification from an agency theory view. The level of diversification is not only negatively
related to managerial equity ownership, but also the equity ownership of outside block holders (Denis,
Denis & Sarin, 1997). Still, even if management does not have a stake in the company, ownership can
influence diversification. For instance, it was proven that dominant family ownership also leads to less
diversification (Chenhall, 1984). Most of the companies that are family-owned want to continue the
tradition and the line of trade of the family for as long as possible and do not want to spread their chances
but protect the heritage instead.
This study is seeking to cover both streams to some extent. It will look at how the level of diversification is
influenced by a number of variables. These will vary from the personal characteristics of the all the
executives on the board to the governance system that is used by the company.
2.2.2 The influence on the level of diversification by different characteristics of the board
In this section an overview will be given on the characteristics of the board that already have been proven
by other scholars to have an influence on the diversification decision. Examples of such characteristics are
age, tenure in the organization, education and socioeconomic roots (Hambrick & Mason, 1984). This paper
re-assesses these first three variables and additionally looks at the variables gender and, in particular,
board structure. Most scholars see the influence of the age of board members on diversification as positive.
It influences diversification positively as a higher age provides more incentives for entrenchment. Other
effects include risk-aversion of older people and increase of pay. Education also provides some direction for
the level of diversification. Educated people will be more knowledgeable on diversification and thus will be
better equipped to make risk assessments. The average number of years of tenure in some ways combines
the good and bad points of age and education. A long tenure means directors have a better overview on
the good points of a firm, but also the weaknesses. These can be potentially exploited for personal gains.
Lastly, Gender is also brought forward, as women tend to be more risk-averse and in accordance to social
changes become more involved in the boardroom. The effect this has on diversification has already been
8
researched to some extent. Early work by Sexton and Bowman-Ufton (1990) showed that this influence
might be exaggerated in strategic management studies. Later studies (Hillman, Shropshire & Cannella,
2007; Miller & Del Carmen Triana, 2009) predict that women are believed to have a positive effect on
diversification. Therefore this variable will be re-examined. In addition to these variables this study
proposes to add a variable, namely the way the board is structured. The variables will be explained further
in the next section.
Age
As already mentioned, risk management is an important argument in the discussion on the influence of age
on diversification. Young managers can be willing to take more risk in comparison to their older peers
(Hambrick & Mason, 1984), but at the same time they will be more willing to establish themselves in the
company. They have the drive to work hard and be innovative, which can lead to the company looking at
things from a different perspective. During the process the might find, for instance, new growth markets for
the current product or to streamline the current operation. This can mean that diversification on grounds
suggested by RBV and stewardship theory, such as ensuring growth or finding synergies to operate more
efficiently, may not be necessary. On the other hand, the experience and know-how of the older board
members makes them both indispensable and a risk to the business. Since they know every detail of the
company they can provide excellent insight in where opportunities for potential synergies with other
companies are. However manager age also gauges stamina for a demanding job. Many believe that
management is so demanding that the negative impact of age on stamina leads to poorer performance
(Golec, 1996). This can lead to a situation where the company is looking to bring fresh into the board in
order to ensure stable growth. When this is the case the entrenchment (Schleifer & Vishny, 1989) argument
presented earlier can present itself, because age also measures time until retirement and, hence, the
importance of future job income to the manager (Golec, 1996). The old manager in question will want to
continue his job until the end of his current term to be sure of a steady income, as well as aware of a
situation where he is forced to look for a new job in a climate where it is very difficult for older people to
find a new job. Therefore, taking the entrenchment argument into account, this study proposes the
following hypothesis:
H1: A higher average age of the management team leads to a higher degree of diversification.
Education
The next variable that will be considered is education. Wiersma and Bantel (1992) proposed that
management teams with higher education level were more open to change. In fact, managers in larger
multinationals tend to at least have completed a university education while a considerable number also has
a MBA or PhD. It is this extra knowledge that this study proposes makes a difference for the level of
9
diversification. Professional education in management is associated with moderation. MBA candidates by
their nature probably are less risk prone, as the analytic techniques learned in an MBA program are geared
primarily to avoiding big losses or mistakes (Hambrick & Mason, 1984). Their peers that lack this education
will more willing to take risks of invest in opportunities that in hindsight can prove to be wasteful.
Therefore, a board member with such a degree should know some basic tenets of investing as well as how
to recognize firms with good management (Golec, 1996). This knowledge makes them able to make better
educated choices. In addition, as all highly educated managers are trained to take more risk-averse
decisions they are more able to look at things with the bigger picture in mind. This stance is associated with
diversification. Therefore, regarding education, this study proposes:
H2: A higher level of average education of the management team leads to a higher level of diversification.
Tenure
The level of education is not the only variable that will lead to better informed choices, the level of tenure
as well influences the decision. Managers are better equipped to assess the dynamics of the organization
and the strengths and weaknesses when they have been employed at the company for a long time. As such,
tenure is a better measure of experience than age, since it measures the manager’s survivorship at the job.
Long tenure implies that the management company finds the manager’s ability and performance
satisfactory (Golec, 1996). This argument rules out the stamina argument presented earlier and can be seen
as a legitimate argument against entrenchment. Additionally it strengthens the argument that young
managers can have more of a “gung-ho” mentality, which leads them to try and find growth fast, without
proper consideration of the facts. Indeed, several scholars found that management teams with shorter
organizational tenure have more diversification (Michel & Hambrick, 1992; Wiersema & Bantel, 1992).
However, there is also a negative argument for a long tenure, since it may also indicate that the manager
has few better opportunities because of specialized skills or an unspectacular performance record (Golec,
1996), which brings back the entrenchment argument (Schleifer and Vishny, 1989) yet again. This can also
be difficult if the influence of seniority, especially in hierarchical organizations, is considered. People who
are longer in the organization can push decisions their way, based on the fact that they are more
experienced and as such could be respected by younger managers. However, it can be said that since most
companies are increasingly held accountable by their actions by the public and investors that this scenario
gets more unlikely by the day. Therefore, following the aforementioned scholars, this study proposes:
H3: Longer average organizational tenure leads to a higher level of diversification.
10
Gender
A variable that has become increasingly important as companies try to fulfill equal opportunities policies is
the role of gender in boards. Whether the presence of women in the board has an influence on
diversification has been researched to some extent. However, this research gave some mixed results.
Therefore the role of gender in relation to the level of diversification can be open to interpretational
disputes. This suggests that this variable can have both positive and negative effects. In the section below
the reasoning behind this suggestion is explained further.
As mentioned before, RBV proposed that one of the arguments for diversification is that it is used to reduce
risk (McDougall & Round, 1984). Risk aversion is an important argument in the differences in diversification
posture between men and women. With respect to the risk management, women tend to take lower risks
and try to avoid losses. Also females are less willing than males to become involved in situations with
uncertain outcomes (Sexton & Bowman-Ufton, 1990). The need for insurance is therefore low and
diversification appears a much more appropriate strategy for women (Dwyer, Gilkeson & List, 2002;
Schubert, 2006). This would implicate that more women on the board would lead to more diversification.
But there are also other arguments, stemming from agency and stewardship theory. For instance the
growth, free cash and profitability arguments proposed by stewardship theory are considered. Sexton and
Bowman-Ufton (1990) for example, found that females have less stamina that is needed to maintain a
growth-oriented business. Males are therefore more likely to pursue the growth opportunities presented
by diversifying. The quest for higher profitability also applies this method and would suggest that the lack
of stamina by women would implicate that men are more eager to look for, diversified, profit
opportunities. When following the reasoning of low insurance (Dwyer, Gilkeson & List, 2002; Schubert,
2006), males are more likely to spend their free cash on diversifying acquisitions. When these arguments
are considered, it would mean that less women on the board lead to more diversification.
It seems that there is evidence that a difference in diversification could be attributed by gender, but it
should be noted that studies directly researching the relationship between gender and diversification level
(Hillman, Shropshire & Cannella, 2007; Miller & Del Carmen Triana, 2009) found only a weak, lowly
significant, correlation. The problem, according to Donnell and Hall (1980), is that women do not differ
from men in the ways in which they administer the management process. Therefore, the arguments
stemming from agency theory are applicable to both men and women. In addition, strategic management
literature overuses gender related managerial differences (Sexton & Bowman-Ufton, 1990). Taking all
aforementioned arguments into account, studying gender seems worthwhile, so this paper proposes that:
H4a: A higher percentage of women on the board leads to a lower level of diversification
11
H4b: A higher percentage of women on the board leads to a higher level of diversification
Board Structure
The aforementioned variables are all tested before and will be used to see whether they are also supported
by this research. However, an increasingly important theme in comparative cross-country research is the
differences in organizational form in diverse national settings (Kogut, Walker & Anand, 2002; Wan &
Hoskisson, 2003). This has also an influence on diversification (Wan & Hoskisson, 2003). Therefore, this
study will thoroughly discuss the influence of the differences between several countries in the way the
board is structured. With board structure this paper refers to how boards are organized. For instance, how
higher management cooperates, which control mechanisms are in place and how decision making
procedures work (Aguilera, 2005). This is important because they influence the quality of directors’
deliberation and decisions, the ability of directors to protect shareholder interests and the ability to provide
strategic direction (Pearce & Zahra, 1992). Furthermore, as diversification increases the span of control of
corporate executives, they are no longer able to fully understand the operations of the multiple divisions
(Hoskisson, Hitt & Hill, 1993).
The board structure cannot be determined by preferences of the company alone. Demands from
governments, legal obligations and social advances have created a way in how decisions are being dealt
with. These distinctions can lead to totally different business environments per country and therefore each
company can have a unique board structure. Though, academic literature acknowledges that, generally
speaking, European companies can have one of two systems in place: a one-tier board system or a two-tier
board system.
One-tier boards
Anglo-Saxon countries can be described as ‘liberal market capitalism societies’ (Albert, 1993), which is the
foundation for some essential characteristics of the system. First of all, since the market is supposed to
regulate and support business life, financing is mostly done with equity. Furthermore, the management in
these countries acts out of one sole purpose and that is to make profit. As a consequence shareholders and
institutional investors are relatively passive and ownership can be quite dispersed. Therefore boards are
not always independent of management. Lastly, there are active markets for corporate control and flexible
labor markets (Gedajlovic & Shapiro, 1998; Aguilera & Jackson, 2003) to stimulate entrepreneurial
behavior.
The one-tier boards especially prevail in these countries. In this model, executive directors and non-
executive directors operate together in one organizational layer. Some one-tier boards are dominated by a
majority of executive directors while others are composed of a majority of non-executive directors
12
(Maasen, 1999). Leadership is divided between the CEO and the chairman. This can be the same person.
One-tier boards often use board committees like audit, remuneration and nomination committees
(Maassen, 1999) to cope with problems that the board alone cannot cope with.
Two-tier boards
Rhineland countries adhere to ‘social capitalism’ (Albert, 1993). The increased involvement of banks,
institutions and the government (Kogut, Walker & Anand, 2002) means that shareholders and stakeholders
participate actively and dynamically in the economy. That is why investments are mostly done using long-
term debt. More control elements are present in the system, such as boards that are more independent of
management ownership by large block holders and weak markets for corporate control. The increased
protection of stakeholders and the workers in particular results in rigid labor markets (Gedajlovic & Shapiro,
1998; Aguilera & Jackson, 2003).
A striking element is the influence of banks in the Rhineland system. Especially German banks typically hold
both large debt and equity positions in companies. This can be explained by the fact that shareholders
typically deposit their shares with these financial institutions. This contributes to a relatively high degree of
ownership concentration (Gedajlovic & Shapiro, 1998). Another example is the strong role of employees
(Olie & van Iterson, 2004) in companies. In Germany (Aufsichtsrat) and the Netherlands
(Ondernemingsraad) the supervisory board includes employee representatives (50% in companies with
more than 2000 employees) (Aguilera, 2005).
Due to these social elements, the two-tier board is prevalent in Rhineland countries. It is composed of a
Board of Management (or decision-making unit) and a Supervisory Board (or monitoring unit) (Aguilera,
2005). The management board is usually composed of executive managing directors. Law forbids that
directors combine the CEO and chairman roles in two-tier boards. Because the CEO has no seat in the
supervisory board, its board leadership structure is formally independent from the executive function of
the board (Maassen, 1999). The supervisory board constitutes entirely of non-executive supervisory
directors who protect interests in a company for unions, government or investors.
The functions of the supervisory board are three-fold: counseling, ratifying decisions made by the
managing board and monitoring the managing board (Douma, 1997). The role of monitoring is a central
element of agency theory and fully consistent with the view that the separation of ownership from control
creates a situation conducive to managerial opportunism (Daily, Dalton & Canella Jr., 2003). One of the key
goals of this board structure is to ensure the independence of the two boards by making sure that
executives are not too powerful (Goold, 1996).
13
In both systems there are several incentives present that either stimulate or discourage to diversify, as well
as measures to deal with unwanted diversification. First there will be looked at one-tier boards. The
greatest risks companies with these boards face are related to agency theory. Highly diversified firms often
display agency problems where governance has been ineffective and the agents (top executives) diversified
the firm in their own self interests (Johnson, Hoskisson & Hitt, 1993). The lack of a governing body makes it
easier for managers to entrench themselves or diversify for other personal reasons (Schleifer and Vishny,
1989), such as financial gains or protecting their human capital. However, agency theory suggests that the
market will resolve this problem as investors have some form of control as they have voting rights when
they feel the board is performing inadequately. Though, boards generally prefer to promote firm efficiency,
and hence help shareholder wealth preservation, before letting the market impose discipline (Singh,
Mathur & Gleason, 2004). As described before, in countries where the one-tier board system is prevalent
the main goal of doing business is making profit. This means that unwanted diversification can continue as
long as investors are not disappointed at the business end: when decent growth, profits and dividends are
ensured they will have little reason for complaints. However the heavily discussed profitability argument,
put forward by stewardship theory, also provides a countermeasure for unwanted diversification. As said
before some scholars argue that diversification has a negative influence on performance. When this is
indeed the case the investors in a one-tier board situation will be quicker to act against poor performance
by management (Westphal & Frederickson, 2001).
Two-tier boards, as explained in earlier sections, have an element of control build into the system, which
mitigates effects suggested by agency theory. Therefore companies with this structure are more likely to
diversify for reasons put forward by RBV and stewardship theory. Especially the role of the employees in
two-tier boards will increase the odds that the decision to diversify is made to reduce risk. But also the free
cash flow argument presented earlier can, when employees are involved, have a negative influence for this
system. Rather than spending this cash for new growth opportunities, they can feel that these funds can
better be invested in other areas. Furthermore, the counseling and monitoring functions of the supervisory
board can lead to slower decision making. Nevertheless there appears to be overwhelming support, in
particular among financial researchers, for supervisory boards providing beneficial monitoring and advisory
functions to firm shareholders (Fields & Keys 2003). Their argument is that people on the supervisory
boards are most of the times older businesspeople, whose experience makes them valuable to companies
from and sometimes vital to improve performance. For this study this implies, following stewardship
theory, that a higher age, longer tenure and higher level of education can also positively influence the
diversification level for two-tier boards.
Other reasons, such as overcoming transaction costs, exploiting operating synergies, preservation of the
firm or ensuring growth apply to both systems. Agency problems are associated with bad behavior by
14
individuals, while risk-reduction is a strategy that is made for the good of the firm and as such will lead to
quicker consensus on the righteousness of the diversification decision. Taking this and all the arguments
presented earlier into account, this paper proposes that companies with two-tier boards will be more
diversified.
H5: Companies with a two-tier board system are more diversified than those that have one-tier boards
All the aforementioned variables and their according hypotheses can be found in Table 1.
TABLE 1
Hypotheses and direction
Variable Hypothesis Effect
Age H1: A higher average age of the management team leads to a higher degree of diversification +
Education H2: A higher level of average education of the management team leads to a higher level of
diversification.
+
Tenure H3: Longer average organizational tenure leads to a higher level of diversification. +
Gender H4a: A higher percentage of women on the board leads to a lower level of diversification -
H4b: A higher percentage of women on the board leads to a higher level of diversification +
Board Structure H5: Companies with a two-tier board system are more diversified than those that have one-tier
boards
+
15
3. Data and Methods
3.1 Sample Selection and data collection
The hypotheses are tested cross-sectional using a sample of corporations for the financial year 2010. A
cross-sectional approach was chosen to control for the many factors external to the corporation, but
related to its diversification levels that vary over time (Chen, Dyball & Wright, 2009). A cross-sectional
approach is justified for this research as the composition of the board changes somewhat every year, but
these changes are marginal as most directors are appointed for certain terms. Only in rare situations these
terms are not completed. Therefore can be argued that the level of diversification in 2010 in the result of
the decisions by boards that had similar characteristics some years back. A sample was obtained using the
Orbis company database.
A random sample was selected by searching companies that matched several preconditions. First of all,
companies should be located in one of the 6 research areas: Germany, the Netherlands, the UK, Ireland,
Belgium or Luxemburg. These countries were chosen as they minimized a host of exogenous influences,
such as regional economic shocks and geographical remoteness (Wan & Hoskisson, 2003), which made it
easier to interpret the results within the variation of board structure. Also these countries are prime
examples of the different board structures in the same economic zone. The study tries to be as recent as
possible, so the companies should have been active in the last financial year (2010). Financial companies
were not included in the sample and as such only companies classified by Orbis as industrial companies
were selected. This was done especially since financial companies are obliged to adhere to strict laws that
forbid them to heavily diversify. The law was also used positively, since other laws mandate public
companies to make more information available, mainly for investors. However, all this information can also
be used for scientific purposes. Another positive ruling used to the advantage of this study was the
existence of international accounting standards. Especially German and British companies can provide
different accounting numbers for the same financial results. For instance the way in which assets can be
amortized is totally different and can heavily affect the results. Therefore companies using the same
accounting standard were used. Furthermore, the companies had at least €2.500.000 in assets in 2010 and
therefore could be considered a large company. Very large companies were chosen because their
management has more discretion in the choice of whether to operate as a single or diversified business,
compared to smaller corporations (Chen, Dyball & Wright, 2009).
After this selection, this initial database consisted of 174 companies. Then data, concerning financial year
2010, for this sample were gathered using Orbis, company websites and the Bloomberg Investor Website.
These data concerned financial performance, company size, board system and characteristics of people on
16
the board, such as gender, age, education and tenure. Hyland and Diltz (2002) also highly recommended
the inclusion of the height of the research and development budget, as it proved to be a significant
predictor of the level of diversification. However, this would have severely restricted the sample size
(Hoskisson, Hitt, Johnson & Moesel, 1993). Some companies lacked data needed for this study, as they
could not be found using aforementioned and other sources. Therefore 9 companies were taken out of the
sample, after which 165 companies were in the sample. Of these 165 companies 95 had a one-tier board
and 70 had a two-tier board.
3.2 Variable specification and measurement
Dependent variable
Diversification
The dependent variable for this study is diversification. There are several measurements for product
diversification that can be taken into consideration. The measures can be classified into three different
types. Firstly, the measure proposed by Rumelt (1974), which is a categorical measure. This measure
focuses on how the revenues of the different businesses are distributed. Secondly, Chatterjee and
Wernerfelt (1991) used a measure based on the Resource-Based View. The focus of these measures is on
the spread of strategic assets and competencies between different units (Markides & Williamson, 1996).
The third measure uses product-count measures and can be divided into two measures: simple product
count measure and the weighted product count measure.
The measure of Rumelt (1974) is based on the distribution of the revenues of the firm. A company is
classified as a single business if 95% of the sales are caused by one business of the firm. A business is said to
be a dominant one if between 70% and 94% of the company’s sales is generated by one business. When
the turnover of a company is below 70% Rumelt is also distinguishing a company as being related or
unrelated diversified.
Markides and Williamson (1996) first measured relatedness in the traditional (Rumelt) way, using a
dichotomous dummy variable: firms classified as related took the value one, and firms classified as
unrelated or dominant took the value zero. Single-business firms were excluded from the analysis. They
then replaced the related variable with structural indicators of relatedness to estimate the equation. As
such they calculated the level of diversification.
The method of Varadarajan and Ramanujam (1987) takes and easier approach and is an example of the
simple product count measure. Entropy measures, such as the one by Palepu (1985), are an example of the
weighted product count measure. Varadarajan and Ramanujam (1987) studied US companies in order to
17
find values for their model. They proposed two categories to measure the degree of related or unrelated
diversification; the ‘Mean Narrow Spectrum Diversification’ (MNSD) to measure the degree of related
diversification and the ‘Broad Spectrum Diversification’ (BSD) to measure the degree of unrelated
diversification. Companies were put in the respective categories based on the number of industries they
were active in. Palepu (1985) takes somewhat the same approach, but controls for the distribution of sales
between business units. Due to the absence of these data for most sample companies, the entropy
measure was not taken into consideration.
However, this study preferably needs a variable that is measured on a quantitative scale, to make it easier
to compute the regression equation. Therefore, diversification in this study will be operationalized by using
an un-weighted product-count measure. These are reliable, simple and easy to compute (Lubatkin,
Merchant & Srinivasin, 1993). Also, when circumstances prevent the use of entropy measures, as is the
case, the use of a product-count measure is appropriate (Hoskisson, Hitt, Johnson & Moesel, 1993). The
product count is based on SIC typology. The Standard Industrial Classification (SIC) system is a numerical
system developed by the US government for classifying all types of economic activity and is based on
establishment classifications, which are classified according to its primary activity (Montgomery, 1982). In
this way companies can be assigned codes according to the industries they operate in. There is high degree
of correspondence between the SIC-based diversification measures and Rumelt’s (1974) categorical
measures (Montgomery, 1982; Lubatkin, Merchant & Srinivasan, 1993). Furthermore, the most significant
studies relating structure to diversity have used business count measures to prove this (Pitts & Hopkins,
1982). Therefore the dependent variable will be a continuous variable, namely, the number of 4-digit SIC
categories the company in the sample was active in, in the year 2010.
Independent variables
In the previous chapter variables that seem to have an influence on diversification were already presented.
Amongst others, Wiersema and Bantel (1992) reported that top management teams’ demographic
characteristics help to predict changes. In this paragraph the demographics used in this study will be
introduced as well as will be explained how they will be operationalized.
Board Age
Age also measures time until retirement (Golec, 1996), but a lower age also means an increased willingness
to risk (Hambrick & Mason, 1984). This paper follows Wiersma and Bantel (1992) and has operationalized
the variable Board Age as the average age of the board members. In the average age all board of directors
that have executive power (Marlin, Lamont & Geiger, 2004) will be included. The average age of board
members will be calculated as the sum of the age of these people, divided by the number of people. In this
case the age will be the number that follows from 2010 minus the year of birth.
18
In sections 4.2 and 4.3 variables will be operationalized differently. In these sections it is necessary to
transform the variables to dummies in order to do t-tests and re-testing the regression to make the data
conclusions more rigid. Therefore Board Age will be recalculated as follows: all the data below the mean
age will be assigned dummy value zero and all the data points above mean will be assigned value one.
Board Tenure
Management teams with shorter organizational tenure have more diversification (Michel & Hambrick,
1992; Wiersema & Bantel, 1992) but could also point to entrenchment (Schleifer and Vishny ,1989).
Therefore, based on research by Michel and Hambrick (1992), Board Tenure will be measured as the mean
number of years the members of board have spent with a firm in their current position. As such the
average will point towards the normal level of diversification. Variation will point towards behavior
proposed by the theories presented earlier. Again, only the people on the board of directors that have
executive power (Marlin, Lamont & Geiger, 2004) are included. The calculation will be made with the
following formula: 2010 minus the year the board member started working in their current position.
As already mentioned, in sections 4.2 and 4.3 Board Tenure will also be operationalized differently. This
means that all the data below the mean tenure will be assigned dummy value zero and all the data points
above mean will be assigned value one.
Board Education
Also Board Education will be part of this study. For his study Golec (1996) used both the total number of
year of education as well as a dummy variable to account for whether a MBA was held. However, this paper
reasoned that while the total years of education measures only measures accumulated general knowledge,
a MBA measures business-specific knowledge. Therefore, in order stress the importance of this extra
education, this variable will be operationalized by the percentage of people on the board, that have
executive power, with a MBA, PhD, or equivalent, titles such as granted from a Doctorate or Professorship.
Again, only the people on the board of directors that have executive power (Marlin, Lamont & Geiger,
2004) are included.
In sections 4.2 and 4.3 also Board Education will be operationalized differently as well. This means that all
boards that have no-one at executive level with an advanced education will be assigned dummy value zero
and all the boards that do have executives with an advanced education will be assigned value one.
Board Gender
The Gender variable has been present in a large number of studies already presented (Sexton & Bowman-
Ufton 1990; Dwyer, Gilkeson & List, 2002; Schubert, 2006). But every time the variable was adapted the
19
specific needs of the study. Thus this study will do the same and follow previous studies (Hillman,
Shropshire & Cannella, 2007; Miller & Del Carmen Triana, 2009). The Board Gender variable will be
operationalized by the percentage of people on the board that is female. Also for this variable only the
people on the board of directors that have executive power (Marlin, Lamont & Geiger, 2004) are included.
Board Gender will be operationalized differently in sections 4.2 and 4.3, in a similar way as Board
Education. All boards that have no women in executive positions will be assigned value zero and all the
companies that have boards with women will be assigned value one.
Board Type
The Board type variable is the one that this study wants to introduce and consequently add to the
regression equation that models the relationship with the level of diversification. The rationale behind it
has been discussed in detail in the previous chapters. This variable will be operationalized by introducing a
dummy variable that will denote whether the company has an one-tier, in which case the dummy takes
value zero, or a two-tier board, in which case it takes value one.
Control variables
To make the statistical analysis more solid, this study also proposes some control variable to account for
effects other than those created by board characteristics.
Firm Size
The first control variable will be firm size, as it heavily affects differentiation and as such is very common in
the diversification literature (Chatterjee & Wernerfelt, 1991). A lot of large multinationals are in fact heavily
diversified. Therefore this relationship will be controlled by the natural logarithm of the assets, a widely
used measurement in the strategic management literature (Pearce & Zahra, 1992; Gedajlovic and Shapiro,
1998; Anderson, Bates, Bizjak & Lemmon, 2000; Mayer & Whittington, 2003).
Industry Effect
The type of industry firms operate in may provide incentives for firms to change their diversification level
(Hoskisson & Hitt, 1990). In addition, the economic concentration of an industry may influence the
likelihood of strategic change (Wiersema & Bantel, 1992). Indeed, some industries display higher levels of
diversification than other industries. As shown in Table 2 the case can be made that this study is no
exception. The industry dummy groups, their mean diversification and accompanying frequencies in the
sample can be found in this table. Controlling for the industry effect is not uncommon in strategic
management literature (Palepu, 1985) and so also for this study industry dummies were taken into account.
These were created by transforming the 2-digit SIC codes of the core business of the companies into
20
dummies of the sectors these codes belonged to. To which dummy groups, according to their 2-digit codes,
the companies in the sample were assigned can be found in Appendix 2. Since companies in the
manufacturing sector dominate this sample they provide the dummy base level.
TABLE 2
Diversification and Industry Dummies
Industry Dummy Mean Diversification Level % of total sample
Mining 2,31 7,9%
Construction 3,09 6,7%
Transport 1,67 20,0%
Wholesale 2,63 4,8%
Retail 1,71 8,5%
Services 1,60 9,1%
Manufacturing 2,34 43,0%
Total 2,15 100,0%
Firm Age
The last control variable will be firm age. Singh, Mathur and Gleason (2004) suggested that observed board
differences between heavily and less diversified focused firms are due to their being at different stages of
corporate evolution. Furthermore firm age must be controlled since young organizations have a lower
boundary on team tenure than old organizations (Michel & Hambrick, 1992). It will be operationalized as
the number that follows from the equation: 2010 minus the year of incorporation (found in the Orbis
database).
3.3 Empirical model
Because of the nature of the hypotheses and the characteristics of the data, several analytical procedures
were used to test this study's hypotheses. First the descriptive statistic s will provide an insight in the
correlations between the variables and possible outliers. Then t-tests are used as a preliminary test to see
whether there are the differences between the subsamples, which are based on the differently
operationalized variables of interest. These tests can also point out whether the individual variation in the
variable can be significant, but is possibly affected by the other data in the regression. Lastly, this
regression analysis will prove how the several variables interact with each other. To be able to draw
conclusions from this analysis the Hypotheses 1 through 6 were tested using multiple least-squares
regression. The overall model that applies to this paper is the following:
(1) Diversification level = β1+ β2 Age + β3 Tenure + β4 Education + β5 Gender + β6 Board
Type + β7 Firm Size + β8 Firm Age + β9 Firm Industry
21
However, to re-test the individual relationship between diversification and each variable the reduced
models (2) up to and including (6) will be used. This can confirm whether the direction of the un-
standardized regression coefficient is indeed correct and is not influenced by multi-collinearity within the
complete model. These models will have the variable of interest as only independent variable while there
will be controlled for firm size, firm age and industry. So models will be:
(2) Diversification level = β1+ β2 Age + β7 Firm Size + β8 Firm Age + β9 Firm Industry
(3) Diversification level = β1+ β3 Tenure + β7 Firm Size + β8 Firm Age + β9 Firm Industry
(4) Diversification level = β1+ β4 Education + β7 Firm Size + β8 Firm Age + β9 Firm Industry
(5) Diversification level = β1+ β5 Gender + β7 Firm Size + β8 Firm Age + β9 Firm Industry
(6) Diversification level = β1+ β6 Board Type + β7 Firm Size + β8 Firm Age + β9 Firm Industry
22
4. Results
4.1 Descriptive statistics
In this section will be looked at the characteristics of the companies in the sample. The descriptive statistics
of these companies are presented in Table 3. As already mentioned the sample consists of 165 companies,
of which 58% have a one-tier board and 42% has a two-tier board. Larger companies tend to have higher
levels of diversification. Therefore the sample studied consisted mostly of larger companies to be able to
pinpoint significant variances. This was also the reasoning behind the sample selection threshold of
€2.500.000 in assets. Furthermore, the companies are at different stages of corporate evolution. On
average they exist for nearly 60 years and vary from recent start-ups and spin-offs that are in business for
just a year to companies with a rich history spanning 197 years. On average they have 55.905 people
employed.
Looking at the variables of specific interest to this study, several remarkable results are worth mentioning.
Only around 1 in 20 board members studied in the sample are female. Even though this variable is not
really part of this study and as such was not studied in depth, it turned out while the data were gathered,
that most female director positions were in either the supervisory board or in non-executive function in
one-tier boards.
TABLE 3
Descriptive Statistics (N=165)
Diversification
Level
Board
Type
Board
Gender
Board
Tenure
Board
Education
Board
Age
Firm
Size
Firm Age
Mean 2,15 ,42 ,0493 3,8445 ,2253 53,030 16,0923 58,95152
Std. Error of Mean ,112 ,039 ,00749 ,20226 ,02027 ,3112 ,09109 3,907
Median 2,00 ,00 ,0000 3,5 ,17 52,500 15,7804 38
Std. Deviation 1,432 ,496 ,09615 2,59810 ,26039 3,9969 1,17001 50,183
Minimum 1 0 0 0 0 42,3 14,73 1
Maximum 10 1 ,50 14,00 1,00 65,5 19,30 197
Variables:
Board Type = dummy variable to difference between one-tier (D=0) and two-tier (D=1) boards
Board Gender = percentage of women in executive positions
Board Tenure = average tenure of people in executive positions for their current position
Board Education = the percentage of people in executive positions with an advanced (MBA/PhD) degree
Board Age = average age of people in executive positions
Firm Size = natural logarithm of the total assets of the company
Firm Age = the number of years since incorporation
The men and women that are working in executive positions have a maximum tenure in their current
position of 14 years, while on average they are working in that position for almost 4 years. The companies
employ on average 22,5% of their executives with people who followed an advance education such as a
23
MBA of PhD, but in some companies all the executives have an advanced education. The average board
age for the sample is 53 years.
Table 4 summarizes the correlations between the independent variables. The correlations between the
level of diversification and the independent variables already give a first insight into how strong the
relationship between them is. Furthermore it shows the strength of the relationship. Therefore it provides
the expected direction of the values that will be obtained with the regression analysis. As can be seen in the
table the relationship between diversification and the independent variables Gender, Tenure and Age show
negligible correlation and are statistically insignificant as well. The control variables Firm Size and Firm Age
are both highly significant, but at the same time are only lowly correlated with the diversification level.
However, the very high significance justifies the use of these variables as control variables in this research.
The last two variables, Board Type and Board Education, have a pretty high significance, but are weakly
correlated to diversification.
The mutual correlations between the independent variables also gave some interesting results. As shown in
the second column, fourth row of the table, the relationship between education and board system is the
only relationship that is both highly significant and has a moderate correlation. At first sight this seems
quite odd. However, an explanation can be found in the characteristics of the sample. Namely, in German
companies it is very common to have doctors or professors leading the company. These titles have been
assigned to matter in this study. Especially for high-tech companies this is the case. As Germany is an
important research nation for this study, this could have affected this relationship. This argument can be
proven with a t-test. A comparison between board type, as grouping variable, and education, as dependent
TABLE 4
Correlations Matrix (N=165)
Diversification Board Type Board Gender Board Tenure Board Education Board Age Firm Size
Board Type ,179**
Board Gender -,020 -,084
Board Tenure -,076 -,115* -,145**
Board Education ,136** ,477**** -,096 ,090
Board Age -,012 -,230*** -,008 ,279**** -,004
Firm Size ,250**** ,082 ,208*** ,033 ,120* ,219***
Firm Age ,243**** ,381**** -,082 -,060 ,147** -,057 ,002
Significance:
* = p < .10
** = p < .05
*** = p < .01
**** p < .001
Variables:
Diversification = level of diversification based on SIC 4-digit product-count
Board Type = dummy variable to differentiate between one-tier (D=0) and two-tier (D=1) boards
Board Gender = percentage of women in executive positions
Board Tenure = average tenure of people in executive positions for their current position
Board Education = the percentage of people in executive positions with an advanced (MBA/PhD) degree
Board Age = average age of people in executive positions
Firm Size = natural logarithm of the total assets of the company
Firm Age = the number of years since incorporation
24
variable, showed at 99% confidence level that in one-tier boards only 12% of people has an advanced
education versus 37% of people on two-tier boards.
Also some other highly significant relationships, albeit with a lower correlation are worth mentioning.
Board type does namely not only significantly correlate with education but also with board age and firm
age. Regarding the former, there is not really a logical explanation why this correlation exists. An
independent t-test confirmed that one-tier boards have on average older boards, but this test was not very
significant. Therefore this correlation is probably caused by sample selection bias. The correlation between
Firm Age and Board Type can be explained. However, this explanation has an argument based on sample
selection. Larger German companies are namely often a continuation of old German family-firms.
Therefore it is well possible that these firms have adjusted the results somewhat. The main topic of this
research, however, is the relationship between Board Type and the level of diversification. The correlation
between these variables was low and also not extremely significant. This gives doubts for usefulness of the
variable board type to explain diversification. Although the regression analysis will offer the full extent of
this relationship, a preliminary t-test confirmed that there are indeed differences in diversification between
one-tier and two-tier boards. At a 98% confidence-level the means of the one- and two-tier boards were
1,93 and 2,44 respectively.
The variable Firm Size has as well two highly correlating variables: Board Gender and Board Age. For the
former can be argued that larger companies feel more responsible to react to pressure put on them by
governments and public opinion to employ a balanced and diverse workforce. For the latter can be argued
that larger firms tend to appointed older, more experienced executives. Lastly, the correlation between
Board Age and Board Tenure can be explained by the reasoning that people get older while they are
accumulating years of their tenure at a company.
Collinearity checks gave no reasons for concern and it is therefore assumed that the models are not
affected by potential multi-collinearity in the regressions.
4.2 Preliminary t-tests
In order to get a first impression of the probable results of the regression and to check the direction of the
coefficients, all independent variables will be t-tested. This means comparison will be done using the
following procedure: all the independent variables that used a continuous or proportional scale were
operationalized differently. All data points were assigned dummy values. For the variable Age this meant
that the dummy took variable zero if the data point was below the mean and value one if it was above the
mean. The same went for the variable Tenure. For the variable board gender the dummy took value zero if
there were no women on the board and value one if there was female representation on the board.
25
Following somewhat the same principle the Education variable was treated. It took value zero if no people
on the board had followed more education after graduating from university and value one is this was the
case. This measurement for education was already proposed in the first place by Golec (1996).The dummy
variable for board type remained. All these dummy variables were treated with an independent variables t-
test, were diversification was the dependent variable and the assigned dummies provided the compared
groups. The results can be found in Table 5.
TABLE 5
Independent variables t-test
N Mean Std. Deviation t sig
Board System 0 95 1,93 1,331 -2,320 0,011
1 70 2,44 1,519
D Board Gender 0 123 2,08 1,239 -,984 0,163
1 42 2,33 1,896
D Board Education 0 73 1,90 1,169 -1,944 0,027
1 92 2,34 1,592
D Board Age 0 89 2,24 1,365 ,878 0,191
1 76 2,04 1,509
D Board Tenure 0 94 2,26 1,579 1,135 0,129
1 71 2,00 1,207
As can be seen, the t-tests show that the means of the variables Age, Gender and Tenure do not
significantly differ between groups. This is in line with what was concluded from the correlations matrix
(Table 4) where already was shown that these variables had no relation with diversification. The expectancy
is that this will be confirmed by the regression analysis. For both Age and Tenure the t-test, as well as the
correlation matrix, predict a negative direction of the relationship with diversification. For Gender the
correlation matrix predicted a negative direction of the relationship, while the t-test is shows a positive
sign. This could be caused by the transformation of the variable and will be checked with the regression.
Again, these tests are not significant, but could further strengthen the conclusions that follow from the
regression. The t-tests for Board System and Education give more hopeful results as they are both
significant at least at a 97% confidence level. The direction of the tests confirms what the correlation matrix
also brought forward, namely that the relation should be positive and as such the regression should give
the same result.
4.3 Regression results
The results of the six regression models that are used to test the hypotheses can be found in Table 6. Each
of these models is used to explain total diversification. First of all, the full model (1) will be considered.
Around 20,6% of the variation in the level of diversification of the companies in the sample can be
explained by this model. The F-value of 3,021 confirms the model is significant. The un-standardized
regression coefficients (B) of the independent variables give some interesting results.
26
Board age has a negative coefficient, both in the full as in the reduced model. This is in line with what was
expected based on the correlation matrix and t-test. However, this is not in line with the effect proposed in
Hypothesis 2. However, these results are not significant, so there can be concluded that no effect is
present.
Models (1) and (3) predict a negative coefficient for the relation between Board Tenure and the level of
diversification. This confirms the expectation based on the correlation matrix and t-test, but again it goes
against what was proposed in Hypothesis 2, which proposed a positive relationship between tenure and
age. All the statistical results prove otherwise. The coefficient is not significant and provides no support for
Hypothesis 2.
The coefficient between Board Education and the level of diversification is positive, which is in line with
what was proposed by both model (1) and (4) and congruent with the statistical procedures in earlier
sections. However, yet again the coefficient does not have a high significance level, which means there is
no support for Hypothesis 3 either.
The relation between Board Gender and the level of diversification is remarkable. Because the correlation
matrix and the regression seems to point out that relationship between them is negative. The t-test on the
other hand, perhaps due to how the dummy was constructed, turned out positive. All these results are
somewhat in line with what was expected for Hypotheses 4a and 4b. All results were not significant though,
so no effect is found.
Lastly, the main topic of this research: the relationship between Board type and diversification. The
correlation matrix proved an, albeit weak, positive relationship. This was backed up by the t-test that
showed a positive, and more significant, relationship. It is also confirmed to be positive by model (1) and (6)
but yet again these results are not statistically significant. As such, Hypothesis 5 cannot be confirmed and
no effect is found.
To account for problems arising from how the independent variables were operationalized, the models
were re-tested. In this test some variables were subjected to regression analysis again, while they were
operationalized differently. The variables Gender and Education, which were measured on a proportional
scale, were operationalized by the dummy variables created to do the t-tests in section 4.1. Also the
variables Age and tenure were calculated differently. After that the regressions were re-run, for which the
results can be found in Table 7. The complete model (1) improved ever so slightly. The R-square of 0,208
and an F-value of 3,059 are slightly higher than in the original model. Also the models (2) and (3) improved
slightly in comparison to the original model. One explanation is that this difference can probably be
27
attributed to the fact that the dummies have less variance. Therefore the role of the, already highly
significant, Firm Size variable becomes even more powerful as it is the only variable with a lot of variance.
Individual significance of the coefficients in all the re-tested models was not different than the original
model. The variable models (4) and (5) did not improve either. There is however an interesting finding in
this retest. The direction of the coefficient of the variable Gender changed from negative, in the original
regression, to positive. This is quite strange, but it was already hypothesised the coefficient in theory could
have gone both ways. Though this finding is not significant, the fact that the retest gave other results could
point towards evidence for this proposition.
28
Variables:
Board Type = dummy variable to differentiate between one-tier (D=0) and two-tier (D=1) boards
Board Gender = percentage of women in executive positions
Board Tenure = average tenure of people in executive positions for their current position
Board Education = the percentage of people in executive positions with an advanced (MBA/PhD) degree
Board Age = average age of people in executive positions
Firm Size = natural logarithm of the total assets of the company times number of employees
Firm Age = the number of years since incorporation
Industry Dummies = dummy variable to differentiate between several industries, assigned as the dominant
2-digit SIC code sector. Base value = ID Manufacturing
TABLE 6
Regression results (N=165)
Diversification
B Model (1) B Model (2) B Model (3) B Model (4) B Model (5) B Model (6)
Constant -2,932 -2,269 -3,625** -3,867** -2,669 -3,808**
Board Age -,019 -,031
Board Tenure -,051 -,059
Board Education ,403 ,461
Board Gender -,371 -,031
Board Type 0,085 ,261
Firm Size ,373**** ,385**** ,359**** ,350**** ,385**** ,349****
Firm Age ,004 ,005** ,004* ,005** ,005** ,004
Industry Dummies
ID Mining ,169 ,152 ,054 ,130 ,152 ,157
ID Construction 1,084** 1,102** 1,052** 1,130** 1,102** 1,099**
ID Transport -,503 -,449 -,484 -,410 -,449 -,435
ID Wholesale ,587 ,567 ,563 ,643 ,567 ,605
ID Retail -,411 -,433 -,523 -,281 -,433 -,331
ID Services -,083 -,055 -,197 -,095 -,055 -,133
R2 ,206 ,192 ,195 ,192 ,192 ,192
Adjusted R2 ,138 ,145 ,149 ,145 ,145 ,145
F 3,021 4,092 4,183 4,092 4,092 4,100
Significance:
* = p <.10
** = p < .05
*** = p < .01
**** p < .001
Models:
(1) Diversification Level = β1+ β2 Age + β3 Tenure + β4 Education + β5 Gender + β6 Board Type + β7 Firm Size + β8 Firm Age
+ β9 Firm Industry
(2) Diversification Level = β1+ β2 Age + β7 Firm Size + β8 Firm Age + β9 Firm Industry
(3) Diversification Level = β1+ β3 Tenure + β7 Firm Size + β8 Firm Age + β9 Firm Industry
(4) Diversification Level = β1+ β4 Education + β7 Firm Size + β8 Firm Age + β9 Firm Industry
(5) Diversification Level = β1+ β5 Gender + β7 Firm Size + β8 Firm Age + β9 Firm Industry
(6) Diversification Level = β1+ β6 Board Type + β7 Firm Size + β8 Firm Age + β9 Firm Industry
29
Variables:
DBoard Type = dummy variable to differentiate between one-tier (D=0) and two-tier (D=1) boards
DBoard Gender = dummy variable to differentiate boards with (D=1) and without (D=1) women
DBoard Tenure = dummy variable to differentiate between boards with mean below (D=0) and above (D=1)
mean tenure
DBoard Education = dummy variable to differentiate between boards without (D=0) and with (D=1) people with
advanced education in executive positions
DBoard Age = dummy variable to differentiate between boards with mean below (D=0) and above (D=1)
mean age
Firm Size = natural logarithm of the total assets of the company times number of employees
Firm Age = the number of years since incorporation
Industry Dummies = dummy variable to differentiate between several industries, assigned as the dominant 2-digit
SIC code sector. Base value = ID Manufacturing
TABLE 7
Re-Test: Regression with variable dummies, results (N=165)
Diversification
B Model (1) B Model (2) B Model (3) B Model (4) B Model (5)
Constant -3,767** -4,293*** -3,653*** -3,772** -3,676**
DBoard Age -,297 -,367*
DBoard Tenure -,181 -,287
DBoard Education ,000 ,087
DBoard Gender ,127 ,100
Board Type ,167
Firm Size ,362**** ,393**** ,356**** ,348**** ,343****
Firm Age ,004 ,005* ,004* ,005** ,005
Industry Dummies
ID Mining ,114 ,142 ,005 ,097 ,077
ID Construction 1,123** 1,124* 1,064** 1,113** 1,108**
ID Transport -,489 -,424 -,507 -,406 -,408
ID Wholesale ,559 ,556 ,551 ,591 ,604
ID Retail -,493 -,435 ,492 ,347 -,408
ID Services -,051 -,010 ,175 ,131 -,152
R2 ,208 ,200 ,195 ,186 ,186
Adjusted R2 ,140 ,154 ,148 ,139 ,139
F 3,059 4,317 4,165 3,947 3,947
Significance:
* = p < .10
** = p < .05
*** = p < .01
**** p < .001
30
5. Conclusions and Recommendations
5.1 Summary and conclusion
This paper analyzes the differences between characteristics of boards with respect to diversification, with a
special focus on the differences between one-tier and two-tier boards. Particularly by giving the answer to
the two research questions: first, which board characteristics affect the decision to engage in product
diversification and secondly, does the difference in board structure have an effect on this decision?
The results of the empirical study of a sample with 165 companies provide a clear answer to the first
question. They gave no support for any of the board characteristics, although these were proposed by
other scholars and could be safely retested in this paper.
The hypothesis for Board age proposed a positive relationship between age and the level of diversification,
but a negative relationship arose from the regression, correlation matrix and t-test. However, the variable
Board age was insignificant. The same problem was seen for the relationship between Board Tenure and
the level of diversification.
Board Education and the level of diversification were confirmed to have a positive relationship by the t-test,
which was even highly significant. Still, no result arose from the regression analysis, which could mean that
for this variable there is no variance in the sample. Board Education was insignificant. The relation between
Board Gender and the level of diversification gave mixed results in all the statistical procedures. Therefore
the presumption that it could provide both a positive as a negative relationship between the two may be
correct. However, also this variable proved insignificant, so a definite conclusion cannot be given.
Lastly, the main topic of this research: the relationship between Board type and diversification. The
correlation matrix and t-test showed an, albeit weak, positive relationship. The regression analysis,
however, turned out insignificant. Nevertheless, a conclusion is that there are indeed differences between
one-tier and two-tier boards. But they cannot be explained by the regression used in this research. In
section 5.3 there will be suggested how this relationship can possibly be re-tested.
5.2 Limitations
There are some limitations to this study that have to be taken into account. First, the insignificance of the
models is potentially caused by the sample. Only a small sub-set of countries is researched and also only
the companies that fulfilled certain requirements were introduced into the sample. The high presence of
manufacturing companies in the sample might also have influenced the results. These companies
31
accounted for almost half of the sample. Since in the manufacturing dummy both companies that are
sensitive as well as less sensitive to changes in the economy this could have made the sample too
homogenous. The sample might have also been too small to properly investigate differences in variation.
Secondly, construct validity could also pose a problem. For product diversification an un-weighted product
count was used. This measure is less refined than weighted measures, like the entropy measure (Palepu,
1985), and far less refined then the measures based on RBV (Markides & Williamson, 1996). A business
count measure is deemed particularly suitable for research comparing diversified and non-diversified firms,
but less for investigating differences among diversified firms (Pitts & Hopkins, 1982). Another objection to
the use of the product count measures has to do with the SIC system which underlies this product count
measure. As already mentioned, manufacturing firms accounted for a large proportion of the sample. This
is caused by two flaws in the SIC system, which inevitably could influence this study. First, if an industry is
classified according to market-based criteria, products produced through radically different processes could
be classified in one category (Montgomery, 1982). Secondly, the manufacturing category groups an
enormous amount of different products, from snacks to paints and furniture. The markets for these
products can have a totally different stance towards diversification, though the SIC system puts them in the
same category. Lastly, the proportional measurement of Board Education and Board Gender may not be
enough to fully explain the variations in these variables between companies. Although the measurement
with a dummy variable was also tested, this could not improve the model. Therefore the insignificance of
these variables can probably be attributed to the absence of enough variation in the sample.
Thirdly, the relationship between board characteristics and diversification may be sensitive to factors other
than the variables that were used. These influences are indeed very complex and maybe some explanatory
variables were missing.
Fourthly, the cross-sectional approach might not have been appropriate. Some other studies used panel
data to properly account for the characteristics some firms have developed as they are in further state of
their corporate evolution (Singh, Mathur & Gleason, 2004). The cross-sectional approach is a snapshot of
the situation in a certain year. Therefore companies can have made rigorous decisions based on factors
that lie out the range of normal influences. The cross-sectional approach disregards these influences, but
the use of panel data approach does account for these influences as multiple years are under investigation
and as such a more stable behavioral pattern can be perceived.
5.3 Suggestions for Future Research
As shown in the limitations as well as some sub-conclusions in the section 4.1 this research suffered from a
selection bias. The inclusions of selection threshold of €2.500.000 in assets to which companies should
32
comply to enter the sample made it a not-random sample. Therefore other companies which could
potentially increase the variation and therefore could lead to better results were potentially excluded.
Future research should, therefore, re-examine the variables used in this study in a larger sample with
possibly a larger geographical scope. That could mitigate the socio-cultural effects that hampered this
study, as for instance the tendency described in section 4.1 that for German companies it is very common
to have doctors or professors leading the company. These titles have been assigned to matter in this study.
Especially for high-tech companies this is the case. As Germany is an important research nation for this
study, this could have affected this relationship. More countries under investigation could have an effect
on this. The research could also be extended to see whether differences in diversification for the board
systems are also present for related and non-related product diversification. A longitudinal study or panel
data study looking how an evolving board composition changes the stance towards diversification would
also heavily complement this study. Especially the influence of equal opportunities schemes should provide
an interesting topic. Another suggestion is that this research is redone, without the use of un-weighed
product-count measure, but with weighted measures product measures or the entropy measure instead.
Also the inclusion of some other variables could have improved the research. For instance, American
literature that investigated the relationship between board and diversification only form a one-tier
perspective included more differences. For example, internal differences between the one-tier boards,
concerning whether board members were possibly independent or outside directors. These are not
accounted for by this study.
The use of sampling, while collecting the data and the inability to generalize the results from the use of only
one year give reason for more testing. The results of a longitudinal study could therefore complement this
study.
The most important implication of this study for the research into product diversification is that the role of
firm size is validated again. In all the models this variable proved most of the variation in the data. Also the
inclusion of industry dummies can be mentioned as a valuable addition to this study. Both are therefore
recommended as control variables in future research.
Besides that fact that this paper aimed to fill a gap in the strategic management literature concerning the
influence of boards on diversification, also a practical answer could be obtained. As explained in section 2.2
the largest danger for diversifying companies lies in the fact that board members do so for personal
reasons. This assumption is based around agency theory and as such, this paper tried to find out whether
the appointment of people with certain characteristics could overcome these problems. This study has,
however, not found a solid answer for this question, and perhaps future research could validate this claim.
IV
References
� Aggarwal, R., & Samwick, A. (2003). Why Do Managers Diversify Their Firms? Agency Reconsidered. The
Journal of Finance, volume 58, pp. 71–118.
� Aguilera, R. & Jackson, G. (2003). The Cross-National Diversity of Corporate Governance: Dimensions
and Determinants. The Academy of Management Review, volume 28, nr.3, pp. 447-465.
� Aguilera, R. (2005). Corporate Governance and Director Accountability: an Institutional Comparative
Perspective. British Journal of Management, volume 16, pp. S39–S53.
� Albert, M. (1993). Capitalism against Capitalism. London: Whurr Publishers.
� Amihud, Y., & Lev, B. (1981). Risk reduction as a managerial motive for conglomerate mergers, Bell
Journal of Economics, volume 12, pp. 605–617.
� Amit, R, & Livnat, J. (1988). Diversification strategies, business cycles and economic performance.
Strategic Management Journal, volume 9, pp. 99–110
� Anderson R., Bates, T., Bizjak, J., & Lemmon, M. (2000). Corporate governance and firm diversification.
Financial Management, volume 29, pp. 5-22
� Baysinger, B., &Hoskisson, R. (1990). The composition of boards of directors and strategic control.
Academy of Management Review, volume 15, pp. 72-87.
� Chatterjee, S., & Wernerfelt, B. (1991). The link between resources and type of diversification: Theory
and evidence. Strategic Management Journal, volume 12, pp. 33–48.
� Chen, R., Dyball, M., & Wright S. (2009). The link between board composition and corporate
diversification in Australian corporations. Corporate Governance: An International Review, volume 17,
208-223.
� Chenhall, R. (1984) Diversification within Australian manufacturing enterprise. Journal of Management
Studies, volume 21, nr. 1, pp. 23-60.
� Daily, C., Dalton, D., & Canella Jr., A. (2003). Corporate Governance: Decades of Dialogue and Data. The
Academy of Management Review, volume 28, nr. 3, pp. 371-382.
� Davis, J., Schoorman, F., & Donaldson, L. (1997). Toward a stewardship theory of management.
Academy of Management Review, volume 22, pp. 20–47.
� Denis, D.J., Denis, D.K., & Sarin, A. (1997). Agency Problems, Equity Ownership, and Corporate
Diversification. The Journal of Finance, volume 52, nr. 1, pp. 135-16.
� Donnell, S., & Hall, J. (1980). Men and Women as Managers: A Significant Case of No Significant
Difference, Organizational Dynamics, volume 8, nr. 4, pp. 60-77.
� Douma, S. (1997). The Two-tier System of Corporate Governance. Long Range Planning, volume 30, nr.
4, pp. 612-614.
V
� Dwyer, P., Gilkeson, J., & List, J. (2002). Gender Differences in Revealed Risk Taking: Evidence from
Mutual Fund Investors. Economic Letters, volume 76, nr. 2, pp. 151-159.
� Fields, M., & Keys, P. (2003). The Emergence of Corporate Governance from Wall St. to Main St.:
Outside Directors, Board Diversity, Earnings Management, and Managerial Incentives to Bear Risk.
Financial Review, volume 38, pp. 1–24.
� Fox, M., & Hamilton, R. (1994). Ownership and diversification: agency theory or stewardship theory.
Journal of Management Studies, volume 31, pp. 69–81.
� Gedajlovic, E., & Shapiro, D. (1998). Management and ownership effects: evidence from five countries.
Strategic Management Journal, volume 19, pp. 533-553.
� Golec, J. (1996). The Effects of Mutual Fund Managers’ Characteristics on Their Portfolio performance,
Risk and Fees. Financial Services Review, volume 5, nr. 2, pp. 133-148.
� Goold, M. (1996). The (Limited) Role of the Board, Long Range Planning, volume 29, nr. 4, pp. 572- 575.
� Goranova, M., Alessandri, T., Brandes, P., & Dharwadkar, R. (2007). Managerial ownership and
corporate diversification: a longitudinal view. Strategic Management Journal, volume 28, nr. 3, pp 211-
225.
� Hambrick, D., & Mason, P. (1984). Upper Echelons: The Organization as a Reflection of Its Top
Managers. Academy of Management Review, volume 9, nr. 2, pp. 193-206.
� Hillman, A., Shropshire C., & Cannella Jr., A. (2007). Organizational predictors of women on corporate
boards. Academy of Management Journal, volume 50, nr. 4, pp. 941–952.
� Hoskisson, R., & Hitt, M. (1990). Antecedents and performance outcomes of diversification: a review
and critique of theoretical perspectives. Journal of Management, volume 16, pp. 461-509.
� Hoskisson, R., Hitt, M., & Hill, C. (1991). Managerial risk taking in diversified firms: an evolutionary
perspective. Organization Science, volume 2, nr. 3, pp. 296-314.
� Hoskisson, R., Hitt, M., & Hill, C. (1993). Managerial Incentives and Investment in R&D in Large
Multiproduct Firms. Organization Science, volume 4, nr. 2, pp. 325-341.
� Hoskisson, R., Hitt, M., Johnson, R., & Moesel, B. (1993). Construct validity of an object (entropy)
categorical measure of diversification strategy. Strategic Management Journal, volume 14, pp. 215-235.
� Hyland, D. & Diltz, D. (2002). Why Firms Diversify: An Empirical Examination. Financial Management
volume 31, nr. 1, pp. 51-81.
� Jensen, M., & Meckling, H. (1976). Theory of the firm: Managerial behavior, agency costs and
ownership structure. Journal of Financial Economics, volume 3, nr. 4, pp. 305-360.
� Jin, L. (2002). CEO compensation, diversification, and incentives. Journal of Financial Economics, volume
66, pp. 29–63.
� Johnson, R., Hoskisson, R., & Hitt, M. (1993). Board of director involvement in restructuring: The effects
of board versus managerial controls and characteristics. Strategic Management Journal, volume 14
(special issue), pp. 33-50.
VI
� Klein, P., & Saidenberg, M. (2000). Diversification, Organization, and Efficiency: Evidence from Bank
Holding Companies, in Harker, P. & Zenios, S.: Performance of Financial Institution, Cambridge
University Press, Cambridge, pp. 153–173.
� Koen, C. (2005): Comparative International Management, McGraw-Hill, London.
� Kogut, B., Walker, G., & Anand, J. (2002). Agency and Institutions: National Divergences in
Diversification Behavior. Organization Science, volume 13, nr. 2, pp. 162-178.
� Lubatkin, M., Merchant, H., & Srinivasan, N. (1993). Construct validity of some unweighted product-
count diversification measures. Strategic Management Journal, volume 14, pp. 433–449.
� Maassen, G. (1999). An International Comparison of Corporate Governance Models. Amsterdam:
Spencer Stuart
� Markides, C., & Williamson, P. (1996). Related diversification, core competences and corporate
performance. Strategic Management Journal, volume 15, pp. 149–165.
� Marlin, D., Lamont, B., & Geiger, S. (2004). Diversification Strategy and Top Management Team Fit.
Journal of Managerial Issues, volume 16, nr. 3, pp. 361-381.
� Matsuaka, J. (2001). Corporate Diversification, Value Maximization, and Organizational Capabilities. The
Journal of Business, volume 74, nr. 3, pp. 409-431.
� Mayer, M., & Whittington, R. (2003). Diversification in context: a cross-national and cross-temporal
extension. Strategic Management Journal, volume 24, pp. 773–781.
� McDougall, F. & Round, D. (1984). A Comparison of Diversifying and Nondiversifying Australian
Industrial Firms. The Academy of Management Journal, volume 27, nr. 2, pp. 384-398.
� Michel, J., & Hambrick, D. (1992). Diversification posture and top management team characteristics.
Academy of Management Journal, volume 35, pp. 9-37.
� Miller, T., & Del Carmen Triana, M. (2009). Demographic Diversity in the Boardroom: Mediators of the
Board Diversity–Firm Performance Relationship. Journal of Management Studies, volume 46, nr. 5, pp.
755-786.
� Montgomery, C. (1982).The Measurement of Firm Diversification: Some New Empirical Evidence. The
Academy of Management Journal, volume 25, nr. 2, pp. 299-307.
� Montgomery, C. (1994). Corporate Diversification. The Journal of Economic Perspectives, volume 8, nr.
3, pp. 163-178.
� Napier, N., & Smith, M. (1987). Product diversification, performance criteria, and compensation at the
corporate manager level. Strategic Management Journal, volume 18, pp. 195-201.
� Olie, R., & Iterson, van, A. (2004). Top management teams in their national context. Advances in
International Management, volume 15, pp. 129–157.
� Palepu, K. (1985). Diversification strategy, profit performance, and the entropy measure. Strategic
Management Journal, volume 6, pp. 239-255.
VII
� Pearce, J., & Zahra, S. (1992). Board compensation from a strategic contingency perspective. Journal of
Management Studies, volume 29, pp. 411–438.
� Penrose, E. (1995 [1959]). The Theory of the Growth of the Firm, Oxford: Oxford University Press
� Pitts, R., & Hopkins, D. (1982). Firm Diversity: Conceptualization and Measurement. The Academy of
Management Review, volume 7, nr. 4, pp. 620-629.
� Ramanujam, V., & Varadarajan, P. (1989). Research on corporate diversification: A synthesis. Strategic
Management Journal, volume 10, pp. 523–551.
� Ruigrok, W., Peck, S., & Keller, H. (2006). Board Characteristics and Involvement in Strategic Decision
Making: Evidence from Swiss Companies. Journal of Management Studies, volume 43, pp. 1201–1226.
� Rumelt, R. (1974). Strategy, structure, and economic performance. Harvard Business School Division of
Research, Boston
� Schleifer, A., & Vishny, R. (1989). Management Entrenchment, the Case of Manager-Specific
Investments. Journal of Financial Economics, volume 25, pp. 123-139.
� Schubert, R. (2006). Analyzing and managing risks – on the importance of gender differences in risk
attitudes. Managerial Finance, volume 32, nr. 9, pp. 706-715.
� Sexton, D., & Bowman-Ufton, N. (1990). Female and male entrepreneurs: Psychological characteristics
and their role in gender-related discrimination. Journal of Business Venturing, volume 5, pp. 29-36.
� Singh, M., Mathur, I., & Gleason, K. (2004). Governance and Performance Implications of Diversification
Strategies: Evidence from Large U.S. Firms. Financial Review, volume 39, pp. 489–526.
� Teece, D. (1982). Towards an economic theory of the multiproduct firm. Journal of economic Behavior
and Organization, volume 3, pp. 39-63.
� Varadarajan, P., & Ramanujam, V. (1987). Diversification and Performance: A Reexamination Using a
New Two-Dimensional Conceptualization of Diversity in Firms. The Academy of Management Journal,
volume 30, nr.2, pp. 380-393.
� Wan, W., & Hoskisson, R. (2003). Home Country Environments, Corporate Diversification Strategies,
and Firm Performance. The Academy of Management Journal, volume 46, nr. 1, pp. 27-45.
� Wang, H., & Barney, J. (2006). Employee Incentives to Make Firm-Specific Investments: Implications for
Resource-Based Theories of Corporate Diversification. Academy of Management Review, volume 31, nr.
2, pp. 466–76.
� Westphal, J., & Fredrickson, J. (2001). Who directs strategic change? Director experience, the selection
of new CEOs, and change in corporate strategy. Strategic Management Journal, volume 12, pp. 1113–
1138.
� Wiersema, M., & Bantel, K. (1992). Top Management Team Demography and Corporate Strategic
Change. The Academy of Management Journal, volume 35, nr. 1, pp. 91-121.
VIII
Appendices
APPENDIX 1
Industry dummy label categorization
CONSTRUCTION TRANSPORTATION, COMMUNICATIONS, ELECTRIC, GAS, AND SANITARY
15 - - GENERAL BUILDLING CONTRACTORS 40 - - RAILROAD TRANSPORTATION
16 - - HEAVY CONSTRUCTION, EXCEPT BUILDING 41 - - LOCAL AND INTERURBAN PASSENGER TRANSIT
17 - - SPECIAL TRADE CONTRACTORS 42 - - TRUCKING AND WAREHOUSING
43 - - U.S. POSTAL SERVICE
MINING 44 - - WATER TRANSPORTATION
10 - - METAL MINING 45 - - TRANSPORTATION BY AIR
12 - - COAL MINING 46 - - PIPELINES, EXCEPT NATURAL GAS
13 - - OIL AND GAS EXTRACTION 47 - - TRANSPORTATION SERVICES
14 - - NONMETALLIC MINERALS, EXCEPT FUELS 48 - - COMMUNICATION
49 - - ELECTRIC, GAS, AND SANITARY SERVICES
MANUFACTURING
20 - - FOOD AND KINDRED PRODUCTS WHOLESALE TRADE
21 - - TOBACCO PRODUCTS 50 - - WHOLESALE TRADE - DURABLE GOODS
22 - - TEXTILE MILL PRODUCTS 51 - - WHOLESALE TRADE - NONDURABLE GOODS
23 - - APPAREL AND OTHER TEXTILE PRODUCTS
24 - - LUMBER AND WOOD PRODUCTS RETAIL TRADE
25 - - FURNITURE AND FIXTURES 52 - - EATING AND DRINKING PLACES
26 - - PAPER AND ALLIED PRODUCTS 53 - - GENERAL MERCHANDISE STORES
27 - - PRINTING AND PUBLISHING 54 - - FOOD STORES
28 - - CHEMICALS AND ALLIED PRODUCTS 55 - - AUTOMOTIVE DEALERS & SERVICE STATIONS
29 - - PETROLEUM AND COAL PRODUCTS 56 - - APPAREL AND ACCESSORY STORES
30 - - RUBBER AND MISC. PLASTICS PRODUCTS 57 - - FURNITURE AND HOMEFURNISHINGS STORES
31 - - LEATHER AND LEATHER PRODUCTS 58 - - EATING AND DRINKING PLACES
32 - - STONE, CLAY, AND GLASS PRODUCTS 59 - - MISCELLANEOUS RETAIL
33 - - PRIMARY METAL INDUSTRIES
34 - - FABRICATED METAL PRODUCTS SERVICES
35 - - INDUSTRIAL MACHINERY AND EQUIPMENT 70 - - HOTELS AND OTHER LODGING PLACES
36 - - ELECTRONIC & OTHER ELECTRIC EQUIPMENT 72 - - PERSONAL SERVICES
37 - - TRANSPORTATION EQUIPMENT 73 - - BUSINESS SERVICES
38 - - INSTRUMENTS AND RELATED PRODUCTS 75 - - AUTO REPAIR, SERVICES, AND PARKING
39 - - MISC. MANUFACTURING INDUSTRIES 76 - - MISCELLANEOUS REPAIR SERVICES
78 - - MOTION PICTURES
79 - - AMUSEMENT & RECREATION SERVICES
80 - - HEALTH SERVICES
81 - - LEGAL SERVICES
82 - - EDUCATIONAL SERVICES
83 - - SOCIAL SERVICES
84 - - MUSEUMS, BOTANICAL, ZOOLOGICAL GARDENS
86 - - MEMBERSHIP ORGANIZATIONS
87 - - ENGINEERING & MANAGEMENT SERVICES
88 - - PRIVATE HOUSEHOLDS
89 - - SERVICES, (NOT ELSEWHERE CLASSIFIED)
Retrieved 14-10-2011 from: http://www.gti.net/njchamber/index-sic.htm
NOTE: This appendix refers to how the
industry dummies are categorized. They
are based on the US government
guidelines, where the first two digits of
the SIC-code of the core operation of the
company is used to assign the sub-
category. To which category the
company belongs can be found in
Appendix 2.
IX
APPENDIX 2
The companies in the sample
Company Name Core Business Core 4-digit SIC Code Industry Dummy Label Diversification Product Count
ROYAL DUTCH SHELL PLC Crude petroleum and natural gas 1311 Mining 10
BP P.L.C. Petroleum refining 2911 Manufacturing 2
VOLKSWAGEN AG Motor vehicles 3711 Manufacturing 5
E.ON AG Electric services 4911 Transport 4
DAIMLER AG Motor vehicles 3711 Manufacturing 5
SIEMENS AG Electrical apparatus 3825 Manufacturing 8
TESCO PLC Grocery stores 5411 Retail 1
METRO AG Department stores 5311 Retail 4
BASF SE Miscellaneous chemical products 2899 Manufacturing 2
DEUTSCHE TELEKOM AG Communications services 4899 Transport 2
BAYERISCHE MOTOREN WERKE AG - BMW Motor vehicles 3711 Manufacturing 1
VODAFONE GROUP PUBLIC LIMITED COMPANY Communications services 4899 Transport 1
ARCELORMITTAL S.A. Iron and steel foundries 3325 Manufacturing 1
RWE AG Electric services 4911 Transport 3
RIO TINTO PLC Miscellaneous metal ores 1099 Mining 1
EUROPEAN AERONAUTIC DEFENCE AND SPACE COMPANY EADS N.V. Aircraft and parts 3728 Manufacturing 1
THYSSENKRUPP AG Steel works 3312 Manufacturing 4
BHP BILLITON PLC Coal mining services 1241 Mining 3
AUDI AG Motor vehicles 3711 Manufacturing 2
BAYER AG Drugs 2834 Manufacturing 6
SCOTTISH AND SOUTHERN ENERGY PLC Electric services 4911 Transport 3
GLAXOSMITHKLINE PLC Drugs 2834 Manufacturing 1
IMPERIAL TOBACCO GROUP PLC Cigarettes 2111 Manufacturing 3
KONINKLIJKE AHOLD NV Grocery stores 5411 Retail 1
DEUTSCHE LUFTHANSA AG Air transportation 4512 Transport 1
ANHEUSER-BUSCH INBEV Beverages 2082 Manufacturing 1
CONTINENTAL AG Tires and inner tubes 3011 Manufacturing 2
J SAINSBURY PLC Grocery stores 5411 Retail 1
ASTRAZENECA PLC Drugs 2834 Manufacturing 4
X
Company Name Core Business Core 4-digit SIC Code Industry Dummy Label Diversification Product Count
KONINKLIJKE PHILIPS ELECTRONICS NV Household appliances 3639 Manufacturing 2
BAE SYSTEMS PLC Aircraft and parts 3721 Manufacturing 2
BT GROUP PLC Communications services 4899 Transport 1
CELESIO AG Drugs 5122 Wholesale 2
XSTRATA PLC Miscellaneous metal ores 1099 Mining 1
HOCHTIEF AG Highway and street construction 1611 Construction 3
WM MORRISON SUPERMARKETS PLC Grocery stores 5411 Retail 1
ENBW ENERGIE BADEN-WÜRTTEMBERG AG Electric services 4911 Transport 2
BRITISH AMERICAN TOBACCO P.L.C. Cigarettes 2111 Manufacturing 3
NATIONAL GRID PLC Gas production and distribution 4923 Transport 1
COMPASS GROUP PLC Eating and drinking places 5812 Retail 1
CRH PUBLIC LIMITED COMPANY Concrete 3272 Manufacturing 2
TUI AG Travel 4725 Transport 4
HEINEKEN NV Beverages 2082 Manufacturing 1
FRESENIUS SE & CO. KGAA Medical instruments 3841 Manufacturing 1
WOLSELEY PLC Hardware wholesale 5074 Wholesale 4
HENKEL AG & CO. KGAA Soap and toilet preparations 2841 Manufacturing 4
AKZO NOBEL NV Drugs 2834 Manufacturing 3
DIAGEO PLC Beverages 2085 Manufacturing 1
KONINKLIJKE KPN NV Communications services 4899 Transport 1
ROLLS-ROYCE HOLDINGS PLC Aircraft and parts 3724 Manufacturing 1
LINDE AG Industrial inorganic chemicals 2813 Manufacturing 4
BG GROUP PLC Crude petroleum and natural gas 1311 Mining 3
SAP AG Computer related services 7372 Services 1
KINGFISHER PLC Variety stores 5331 Retail 1
ADIDAS AG Rubber and plastics footwear 3021 Manufacturing 5
JOHNSON MATTHEY PLC Miscellaneous chemical products 2899 Manufacturing 1
HEIDELBERGER ZEMENT AG Cement, hydraulic 3241 Manufacturing 4
MARKS AND SPENCER GROUP P.L.C. Department stores 5311 Retail 3
ASSOCIATED BRITISH FOODS PLC Food 2099 Manufacturing 3
BALFOUR BEATTY PLC Heavy construction 1629 Construction 3
THOMAS COOK GROUP PLC Travel 4724 Transport 1
XI
Company Name Core Business Core 4-digit SIC Code Industry Dummy Label Diversification Product Count
DIXONS RETAIL PLC Consumer electronics 5731 Retail 4
AURUBIS AG Copper ores 1021 Mining 1
RECKITT BENCKISER GROUP PLC Soap and toilet preparations 2841 Manufacturing 1
N.V. UMICORE S.A. Copper ores 1021 Mining 2
MERCK KGAA Drugs 2834 Manufacturing 2
DCC PUBLIC LIMITED COMPANY Petroleum 5172 Wholesale 3
KONINKLIJKE DSM N.V. Miscellaneous chemical products 2899 Manufacturing 1
G4S PLC Miscellaneous business services 7382 Services 1
VEDANTA RESOURCES PLC Nonferrous foundries (castings) 3365 Manufacturing 3
X5 RETAIL GROUP N.V. Grocery stores 5411 Retail 1
SALZGITTER AG Steel works 3312 Manufacturing 1
BILFINGER BERGER SE Heavy construction 1622 Construction 3
BAYWA AG Machinery 5083 Wholesale 4
STMICROELECTRONICS N.V. Electronic components 3674 Manufacturing 2
KONINKLIJKE BAM GROEP NV Contractor 1521 Construction 4
LANXESS AG Plastics materials 2821 Manufacturing 3
HOME RETAIL GROUP PLC Department stores 5311 Retail 1
SOLVAY SA Drugs 2834 Manufacturing 3
BELGACOM SA Communications services 4899 Transport 1
BRITISH SKY BROADCASTING GROUP PLC Cable services 4841 Transport 1
INCHCAPE PLC Motor vehicles 5012 Retail 5
PEARSON PLC Books 2731 Manufacturing 1
SMURFIT KAPPA GROUP PLC Paper 2671 Manufacturing 1
SUDZUCKER AG Sugar and confectionery products 2063 Manufacturing 5
BEIERSDORF AG Soap and toilet preparations 2844 Manufacturing 3
GKN PLC Motor vehicles 3714 Manufacturing 2
BUNZL PUBLIC LIMITED COMPANY Paper 2671 Manufacturing 1
RTL GROUP SA Broadcasting 4833 Transport 3
REXAM PLC Metal cans and containers 3411 Manufacturing 2
KLÖCKNER & CO SE Metals and minerals 5051 Wholesale 1
K+S AKTIENGESELLSCHAFT Agricultural chemicals 2874 Manufacturing 3
XII
Company Name Core Business Core 4-digit SIC Code Industry Dummy Label Diversification Product Count
SERCO GROUP PLC Management services 8741 Services 1
CARILLION PLC Contractor 1522 Construction 2
KERRY GROUP PUBLIC LIMITED COMPANY Food 2099 Manufacturing 3
CARNIVAL PLC Cruises 4481 Transport 1
WACKER CHEMIE AG Miscellaneous chemical products 2899 Manufacturing 1
GEA GROUP AG Technology 8711 Services 4
IMTECH N.V. Electrical work 1731 Construction 2
LOGICA PLC Computer related services 7373 Services 2
INTERNATIONAL POWER PLC Electric services 4911 Transport 2
RHEINMETALL AG Vehicle parts 3714 Manufacturing 3
SCA HYGIENE PRODUCTS SE Paper personal car 2676 Manufacturing 1
RYANAIR HOLDINGS PUBLIC LIMITED COMPANY Airline 4512 Transport 2
TRAVIS PERKINS PLC Construction materials 5031 Wholesale 1
ANTOFAGASTA PLC Copper ores 1021 Mining 1
EASYJET PLC Airline 4512 Transport 1
AMEC P L C Contractor 1522 Construction 5
FREENET AG Telephone communications 4813 Transport 1
INFINEON TECHNOLOGIES AG Electronic components 3679 Manufacturing 1
BABCOCK INTERNATIONAL GROUP PLC Miscellaneous business services 7389 Services 1
AGFA GEVAERT NV Photographic equipment 3861 Manufacturing 2
BEKAERT SA/NV Steel works 3315 Manufacturing 2
TATE & LYLE PUBLIC LIMITED COMPANY Sugar and confectionery products 2062 Manufacturing 3
PETROFAC LIMITED Oil and gas field services 1389 Mining 1
SMITHS GROUP PLC Technology 3812 Manufacturing 1
CAPITA GROUP PLC (THE) Management services 8744 Services 2
EXPERIAN PLC Information services 7323 Services 1
TAYLOR WIMPEY PLC Contractor 1522 Construction 4
AXEL SPRINGER AG Publishing and printing 2711 Manufacturing 3
PROSIEBENSAT1 MEDIA AG Broadcasting 4833 Transport 1
CSM NV Sugar and confectionery products 2064 Manufacturing 3
COOKSON GROUP PLC Electrics 3679 Manufacturing 2
XIII
Company Name Core Business Core 4-digit SIC Code Industry Dummy Label Diversification Product Count
SMITH & NEPHEW PLC Medical instruments 3845 Manufacturing 2
MILLICOM INTERNATIONAL CELLULAR SA Communication services 4899 Transport 1
HEIDELBERGER DRUCKMASCHINEN AG Machinery 3555 Manufacturing 1
PREMIER FOODS PLC Food 2099 Manufacturing 2
RHOEN-KLINIKUM AG Clinics 8011 Services 2
CABLE & WIRELESS WORLDWIDE PLC Communication services 4899 Transport 1
MTU AERO ENGINES HOLDINGS AG Aircraft and parts 3721 Manufacturing 1
KONINKLIJKE BOSKALIS WESTMINSTER NV Heavy construction 1629 Construction 3
TOGNUM AG Machinery 5084 Wholesale 1
NATIONAL EXPRESS GROUP PLC Trains and buses 4131 Transport 2
DEBENHAMS PLC Department stores 5311 Retail 1
KAZAKHMYS PLC Copper ores 1021 Mining 1
ITV PLC Broadcasting 4833 Transport 1
FUGRO NV Surface data services 8713 Services 1
BARRATT DEVELOPMENTS P L C Operative builders 1531 Construction 3
SBM OFFSHORE N.V. Oil and gas field services 1389 Mining 3
MITCHELLS & BUTLERS PLC Eating and drinking places 5813 Retail 2
YELL GROUP PLC Advertising 7311 Services 1
COBHAM PLC Aircraft and parts 3728 Manufacturing 2
FRAPORT AG Airport 4581 Transport 1
SEVERN TRENT PLC Water supply 4941 Transport 2
WHITBREAD PLC Hotels and motels 7011 Services 3
PERSIMMON PUBLIC LIMITED COMPANY Contractors 1521 Construction 2
UNITED UTILITIES GROUP PLC Water supply 4941 Transport 2
SES S.A. Communication services 4899 Transport 1
MAINOVA AG Electric services 4911 Transport 4
AEGIS GROUP PLC Advertising 7319 Services 2
THE SAGE GROUP PLC. Computer related services 7372 Services 1
STADA ARZNEIMITTEL AG Drugs 2834 Manufacturing 2
TOMTOM NV Communications equipment 3669 Manufacturing 1
PUNCH TAVERNS PLC Eating and drinking places 5813 Retail 2
XIV
Company Name Core Business Core 4-digit SIC Code Industry Dummy Label Diversification Product Count
DYCKERHOFF AG Cement, hydraulic 3241 Manufacturing 1
INFORMA PLC Publishing and printing 2721 Manufacturing 1
PENNON GROUP PLC Sanitary services 4952 Transport 1
SOLARWORLD AG Electric services 4911 Transport 1
GREENE KING PLC Beverages 2082 Manufacturing 4
CAIRN ENERGY PLC Crude petroleum and natural gas 1311 Mining 2
LONMIN PUBLIC LIMITED COMPANY Gold and silver ores 1041 Mining 1
KONINKLIJKE VOPAK N.V. Public warehousing and storage 4226 Transport 1
MILLENNIUM & COPTHORNE HOTELS PLC Hotels and motels 7011 Services 1
QIAGEN NV Instruments 3826 Manufacturing 1
MARSTON'S PLC Beverages 2082 Manufacturing 4