1

Determinants of Board Structure in Microfinance Institutions: Evidence from

East Africa

This Version: May, 2011

Neema Moriab

Trond Randøyac

a. University of Agder, 422, 4604 Kristiansand, Norway

b.Email: neemagm @uia.no. c. Email: trond.randoy@uia.no

Abstract

This paper investigates the associations between three characteristics of microfinance institutions

(regulation, international and founder management) and their associated board structure (size,

independence, and diversity). The microfinance industry is unique in terms of its newness,

diverse organizational status, with for-profits and non-profits, its high level of international

influence and its social mission, commonly focusing on female and poor. We therefore find it

necessary to examine whether these unique characteristics determine how boards in the

microfinance industry are structured. We are guided by agency and resource dependence

theories, as we test these associations using a panel data set of 63 organizations and 465 board

members from three East Africa countries. We find that the presence of regulations and

international influence to be significantly associated with larger boards while that of founder

management to be associated with less board independence. We further evidence a higher level

of board gender diversity in microfinance organizations which are managed by founders and

those with international influence.

Keywords: Microfinance Institutions; Board independence; board size; board diversity.

JEL codes: G21, G30

2

INTRODUCTION

Microfinance Institutions (MFIs) offer financial services to poor people. Access to financial

services is provided to help poor people mitigate risks, build their assets, improve their income,

and furthermore contribute to development of the focal community (Cull et al., 2009). The

building of better governance mechanisms for MFIs is high on the policy agenda of many

developing countries (Labie & Mersland, 2009). It is suggested that better governance is among

the most important pillars for MFIs to sustainably offer financial services to the poor (Hartaska

& Mersland, 2009). Boards of directors provide an internal governance mechanism (versus the

external governance from shareholders and other external stakeholders) that is particular

important in order to oversee and advise the organization’s managers (Hillman & Dalziel, 2003).

For board members to efficiently oversee and advise, they need to have a structure that supports

such activities (Jackling & Johl, 2009; Pearce & Zahra, 1992). However, little is known about

MFIs’ boards and what determines their structure. This study addresses this research gap by

examining the relationship between MFIs unique characteristics and their board structure. We

therefore ask: What determines the board structure of an MFI?

The unique characteristics of MFIs relates to a number of aspects. First, most MFIs are

characterized by their combining commercial and social goals (Batilana & Dorado, 2010). These

organizations typically build on a novel approach to the provision of loans, utilizing regular

banking principles, while at the same time targeting social goals. These dual objectives have

brought the attention of international aid agencies and philanthropically investors (Mersland,

Randøy & Strøm, 2011). Second, the international affiliation of MFIs provides another unique

characteristic for the industry. Using an analogy from the international business literature e.g.,

(Morck & Yeung, 1991; Tallman & Li, 1996), we argue that internationalization of MFI leads to

better organizational results. This was also partly supported by (Mersland et al., 2011). For

example (Mersland et al., 2011) reports more than 51% of MFIs in their 73-country dataset have

a relationship with an international capital provider (subsidized debt). Third, MFIs vary in

ownership forms with some operate as either regulated shareholders-owned with a profit motive,

while others operate as unregulated non-profit MFIs (Mersland & Strøm, 2008). Lastly, the

microfinance industry is young and to a large extend still managed by its founders. For example,

in our East African data set 54% of all CEOs were also the historical founder of the MFIs. We

3

suggest that MFI’s board structure is partly determined by these unique characteristics of the

industry.

Steier (1998) note that board structure in general is supposed to be set in a way that supports

the unique nature of the industry. In the context of the microfinance industry, (Conger, Finegold

& Lawler, 1999), provide guidelines on desirable board composition, members’ representation

and the size of MFI boards. The authors suggest MFI boards to be composed of a mixture of

members with different skills, especially related to expertise on social and financial issues.

Hartarska, (2005) further suggest that MFI board size should be small enough to avoid free

riding problem. Boards are also required to be structured in a way that they are able to

accomplish their main roles (CMEF, 2005). Campion and Frankiewicz, (1999) further suggest

five roles that MFIs board should fulfill: act as a boundary spanner; provides a reputational

referent; attracts and mobilizes resources and assists in identifying the MFI’s core mission. These

roles can be integrated into two major roles of MFIs boards: monitoring and advising CEOs on

various matters (Hillman & Dalziel, 2003). These roles are grounded on agency and resource

dependency theories (Hillman & Dalziel, 2003). The theories suggest that the ability of board to

effectively practice its roles depend on how the board is structured and composed (Boone, 2007).

In this study we examine the determinants of board structure in East African MFIs (Kenya,

Tanzania and Uganda). We find this specifically interesting to analyze for two reasons. First,

Africa is recognized as the least developed continent in terms of microfinance development, but

it has the biggest growth potential market for microfinance (CGAP, 2009). In Africa, East Africa

is evidenced to be the most microfinance developed region (CGAP 2010). For example (CGAP,

2010) reports 30% of borrowers in Africa are from the East African region. Second, with a

significant influence from international donors and capital providers in East Africa, we expect

that this push MFIs towards an effective governance and board structure. Since microfinance in

East Africa is somewhat developed and there is more international influence comparing to other

African regions, we find it particularly interesting to analyze how governance and specifically

boards are structured in this region.

We examine three characteristics of MFIs: regulatory status, international influence and

founder management. We test the associations between these characteristics and three

dimensions of board structure: board size, board independence and board gender diversity.

4

Hypotheses were tested using a panel data that was obtained through a survey conducted by

researchers between December, 2009 and July, 2010. A total of 63 MFIs, 465 board members

were surveyed and panel data between 2004 and 2009 was collected.

We analyzed our data using a mixture of analytical procedures. First, we performed single

measures of each board structure variable in relation to MFI variables using random effects

estimations. Second, we perform joint estimations using seemingly unrelated (SUR) regression

analysis in order to consider possible combined effects among the board structure variables.

Third, we addressed possible endogeneity problems by using a two stage least square regressions

(SLS) with instrumental variables. Main instruments used for endogenous variables were

whether the MFI was internationally initiated and MFI size.

Our results indicate that MFIs’ board structure is indeed determined by MFIs unique

characteristics. In terms of board size, we evidence large board size to be determined by whether

the MFI is; regulated or exposed to international influences. These findings are in line with

resource dependency theory, that strong stakeholders have the power to be added to the board in

order to provide resources to the MFI. Furthermore, this is a way for such stakeholders to reduce

agency costs. On the other hand, we evidence that founder managed MFIs to be associated with

small boards–as such highly internally motivated founders reduce agency costs of monitoring.

This confirms literature on founder managed organizations that they have small boards that they

are manageable and controllable (Certo, Covin, Daily & Dalton, 2001). Results further indicate

that MFIs which have less board independence are those that are part of international networks

and are managed by founders. Our results suggest that the typical East African MFI is at a stage

of development where resource through insider board members is more important than the

potentially better outsider monitoring that comes with more board independence. We further

show that board gender diversity is determined by MFI characteristics. We specifically evidence

this for MFIs with international networks and founder managed. This is in line with a resource

access argument that diversity bring in variety of resources to organizations (Terjesen, Sealy &

Singh, 2009).

This paper contributes to the literature by showing how key MFI characteristics produce

variation in board structures in MFIs in Africa. Resource dependency theorists highlight the

importance of resources access to organizations (Hillman & Dalziel, 2003; Pfeffer, 1973), and

5

we show this in the context of the microfinance industry. We also draw from agency theory by

showing how the various MFI characteristics affect incentives and the potential for monitoring –

and the subsequent need for a certain board structure. Furthermore, variation in monitoring needs

depends on characteristics such as MFI age and size. MFIs, of which most are small and young,

have the greatest need for resources access. However, as they grow they may opt for a structure

that is better for monitoring, for example by the means of more board independence. To our

knowledge, this is among the first study that looks at the determinants of boards in MFIs. We see

that this paper provides a contribution in three areas; first as it relates to boards and governance

in MFIs, second as it relates to the role of boards in non-profit organizations/mission driven

organizations, and third, as the paper relates to corporate governance in emerging markets.

The paper proceeds as follows. The next section presents a brief background of microfinance

in East Africa. Next section presents theory, literature and hypotheses developed for the study.

We then elaborate on data and methods used. Next we present and discuss results followed by a

section presenting contributions and conclusion.

MICROFINANCE IN EAST AFRICA

The East African countries chosen for this study are Kenya, Uganda and Tanzania. These

counties represent significant similarities on legal and regulatory regimes (LaPorta, Lopez,

Shleifer & Vishny, 1997) but they are heterogeneous in terms of the development of the national

microfinance industry. This combination of cross-country institutional similarity with industry

variations provides East Africa a suitable environment for microfinance research (particularly in

relation to other English common law countries).

The region is recognized as the most microfinance developed region in Africa. Microfinance

operations in East Africa officially started in 1990s (Olomi, 2008; Randhawa & Gallardo, 2003),

but there were microfinance related activities before that. For example, Uganda Finance Trust

was established in 1984 while Kenya Female Finance Trust was established in 1982.

Microfinance activities and operators increased dramatically in the period of twenty years. MIX

Market (2011) reports 3.3 million East African active borrowers reached (MFIs reporting to the

MIX Market database as of December 2009). The current major providers of microfinance

services in East Africa have four major legal and institutional forms: (1) banks, (2) cooperatives,

6

(3) non-bank financial institutions (NBFIs) and (4) non-profit MFIs (NPOs) which are

commonly referred to as NGOs (Mersland & Strøm, 2008). However, we do not include

cooperatives in this study, as such institutions have a vastly different ownership and governance

structures (Hartarska, 2005).

Another interesting characteristic of MFIs in East Africa is the existence of both domestically

and internationally funded MFIs. MFIs with international partners and/or participants in

international networks have access to international funding and/or technical assistance in various

areas. Commonly, fund providers influence governance of MFIs. For example, some funders

have specifications on how the board of a MFI should be, before they provide funding (CGAP,

2008). Since the region has had extensive funding and international attention, one may expect

that the governance structure of MFIs in this region to be more “developed” and even potentially

more efficient than in regions with less such pressure.

The level of government regulations in East Africa is also relatively extensive in relation to

other regions (particularly in Africa) and this again makes regional MFIs particularly interesting.

Again, we make the argument that this adds to the pressure for an efficient governance structure.

In terms of regulatory regime, there are shareholders owned MFIs of which all of them are

regulated by their respective central banks. NPOs are not regulated by central banks, though

there is a certain political pressure to look for ways to regulate them. All MFIs in East Africa are

required to have boards. Furthermore, all regulated MFIs are required to form boards of a certain

structure. For example, boards of regulated MFIs in Kenya and Tanzania are required to have

two thirds of board members who are not affiliated to the MFI. In Uganda, the laws require

regulated MFIs to have at least five board members.

Despite regulation guidelines, there is still significant flexibility in terms of how MFIs’ board

members are appointed. Furthermore, based on the general CEO founder literature (Certo et al.,

2001a), such CEOs have typically significant ability to determine board structure. In Uganda,

five of the biggest MFIs become regulated between 2005 and 2007, with the founder remaining

CEOs in some cases. The similar experience is in Kenya and Tanzania, where by most MFIs are

managed by founders. This is different from other MFIs globally. Table 1 below shows that 38%

of MFIs worldwide are managed by founders. The corresponding number in our data is 54%.

7

Unregulated non-profit MFIs in East Africa are a mixture of small and large in sizes; also

with variation in terms of some local and some international founders. Locally founded NPOs

are in most cases managed by local founders, and since they are unregulated, they have much

influence on deciding who should be on the board (Sabana, 2006). Internationally founded NPOs

have board structures that vary depending on where the MFI is originally based. It is common to

find NPO-MFI in Tanzania having board members who are CEOs of the same NPO in other

countries. For example, board members of Brac Tanzanian are also board members and CEOs of

Brac Uganda.

Table 1 below compares MFIs in East Africa and MFIs globally. Data on global MFIs is

obtained from the rating agencies at www.ratingfund1.org and contains MFIs from 73 countries

worldwide, most recently used in a study by Mersland et al. (2011).

------------------------------------ Insert Table 1 about here ------------------------------------

As seen from table 1 above, on average, MFIs in East Africa are similar to other MFIs

worldwide in terms of age, experience and size. However, there are other characteristics which

are unique for East African MFIs. The industry has had more international attention as 66% of

MFIs have international affiliations. This resulted into more international influence on

governance and regulation processes. A larger proportion of MFIs in the region are managed by

founders, than in other parts of the world. Furthermore, a large proportion of MFIs in this region

are regulated by central banks compared MFIs globally.

LITERATURE AND HYPOTHESES

Dimensions of Board Structure

This study focuses on factors determining board structure. Based on past research on board

structure (Jackling & Johl, 2009; Linck, Netter & Yang, 2008; Pearce & Zahra, 1992) and boards

in microfinance specifically (CMEF, 2005; Conger et al., 1999; Mersland & Strøm, 2009), we

chose to address three dimensions of board structure: board size, board independence and board

diversity.

Board size is an important board structure variable because past research indicate that board

size effect boards’ effectiveness. According to agency theory, small boards operate more

8

effectively than large boards because of lower coordination costs and less free-rider problems

(Jensen & Meckling, 1976; Yermack, 1996). For instance, Lipton and Lorsch, (1992) argue that

“when a board has more than ten members it becomes more difficult for them all to express their

ideas and opinions.” Similarly, (Jensen, 1993) conjectures that “keeping boards small can help

improve their performance. This idea is consistent with the more general view that larger boards

may be less participative, less cohesive, and less able to reach consensus. On the other hand,

resource dependence scholars suggest that larger boards are associated with higher levels of

organizational performance (Hillman, et al., 2009; Pfeffer & Salancik, 1978). In this view, board

size is a measure of an organization's ability to form environmental links to secure resources.

With different expertise, CEOs are able to get better advice and resources from large boards

(Lorsch & Maclver, 1989). A larger board may offer an exceptional level of high quality advice

and counsel to a CEO.

With the nature of the microfinance industry, we argue, we see the same kind of trade-off

between monitoring and resource access. A large board may enhance MFI resources access while

being weaker at monitoring (particularly the CEO). Microfinance policy papers advocate for

bigger enough boards to provide better monitoring and resources to managers (CMEF, 2005;

Sabana, 2006). Hartarska and Mersland, (2009) studied which governance mechanisms promote

efficiency in MFIs. Their results indicate that there is some benefit to a larger board, but this

effect reverses after a certain size. They argue, the results are consistent with the literature on

boards in banks and non-profits, for which boards are found to be larger than conventional

boards. Their results are also consistent with the idea that excessively large boards may suffer

from free-riding. These studies provide evidence that MFIs can benefit from having large boards

but there is an optimal board size. We take the perspective that the choice of board size is

governed by the trade-off between the aggregate information that large boards possess and the

increased costs of decision making associated with them.

Board independence is an important board structure variable as it determines how

independent from management the board is. Jensen, (1993) argues that because inside board

members are likely to be ineffective in monitoring and evaluating the CEO, the only inside

member on the board should be the CEO. Hermalin and Weisbach (1988) develop a bargaining

model in which the board’s independence as a monitor evolves as a function of the bargaining

9

power of the CEO and the board. Raheja (2005) takes into account both the monitoring and

advising roles of the board, and finds that outsider dominated boards are of greater value in

organizations. On the other hand, Harris and Raviv (2008) demonstrate that when insiders have

important information relative to outsiders, insider-controlled boards are preferred. Inside board

members have been found to provide several benefits to organizations they serve. Baysinger, et

al., (1991), for example, found a positive and significant relationship between inside board

members and expenditures for research and development of organizations. Insiders can

contribute to boardroom discussions in ways outside directors cannot because the former have

direct knowledge of organization operations and capabilities (Baysinger & Hoskisson, 1990).

In the young and entrepreneurial industry like microfinance, various stakeholders value

clarity of strategic direction (Lapenu & Pierret, 2005). Inside board members, due to their

detailed organizational specific knowledge, are likely to be superior to outside directors in

assisting CEOs with defining and maintaining a clear strategic focus for their organizations

(Dallas, 1996). Also, MFIs could benefit from a more independent board through not only better

monitoring, but also variety of resources that could be brought by outsiders (Chen, 2011; CMEF,

2005). Outsiders are often valued for their ability to monitor and control management (Johnson

et al. 1996). This control function might be less important to young organizations - like MFIs -

than among more mature firms (Pearce & Zahra, 1992). But as MFIs or other organizations grow

old, they require boards that are better at monitoring. As one MFI CEO puts it; “Four years ago,

the Board’s role was largely advisory. Now, the stakeholders expect it to take a more active role,

especially in monitoring and risk control. We need to draw more on the specialized expertise of

the outside members” (CMEF, 2005). It then appears that MFIs with more independent (less

independent) boards benefit more (less) from monitoring than from resource access.

Diversity within boards is advocated to be an important board structure element (Erhardt,

Werbel & Shrader, 2003; Strøm, D'Espallier & Mersland, 2010; Terjesen et al., 2009). We focus

on board gender diversity in terms of female board members – which has also recently been done

in the general finance literature (Adams & Ferreira, 2009). Burgess and Tharenou (2002) list

numerous arguments for the recruitment of female board members: a) increased diversity of

opinions in the board room b) influence on strategic decision making and leadership styles of the

organization, c) improving organizational image with stakeholder groups and d) ensuring better

10

board room behavior. Female board members further contribute an independent view to the

board and demonstrate how one female member’s intervention can change the strategic direction

of an organization (Selby, 2000). A study by Daily, Certo & Dalton (1999), reports that female

board members have made significant progress in terms of assuming seats on boards. Bilimoria

(2000) reports that even though the number of female board members is increasing, few

organizations actively recruit females, and there is still a common gender bias, stereotyping and

tokenism on boards. Miller and Triana (2009) studied demographic diversity among fortune 500

organizations in US. They found that organizations with female on boards have higher rate of

innovation. Women also contribute to governance and reduce CEO dominance due to their

power sharing style. Adams and Fereira (2009) for example, show that female board members in

the US have a significant impact on boards’ monitoring ability. They conclude that gender-

diverse boards allocate more effort to monitoring the CEO.

Microfinance and a female focus are intrinsically linked. The objective of the Microcredit

Summit Campaign, which plays a central role in the promotion of microfinance, is “to ensure

that 175 million of the world’s poorest families, especially female, have access to financial and

business services (Strøm et al., 2010). Several MFIs in East Africa for example, are designed and

targeting female only. Mersland and Strøm (2009) report, that 73% of the customers in their

global dataset are female. From the female customer focus, it appears that there are benefits of

having female on boards. Their presence on board would bring not only a gender balanced board

but also female will be able to give information on what female customers need (Mersland and

Strøm, 2009). This was further suggested by (CMEF, 2005) that MFIs boards should be gender

balanced in order to benefit from both parts.

Factors influencing board structure

The review above highlights that board size, level of independence and gender diversity can

affect board effectiveness. Past corporate governance literature helps us to identify factors that

might determine the choice of board structure (Coles, Daniel & Naveen, 2008; Lehn, Patro &

Zhao, 2009). Factors that have widely been studied are organization size (Alonso, Palenzuela &

Merino, 2009; Linck et al., 2008), organizational age, product complexity (Coles et al., 2008;

Markarian & Parbonetti, 2007) and leverage (Pearce & Zahra, 1992). Since these factors are

11

generic and focused on corporate organizations, we chose to specifically focus on MFIs unique

factors that according to literature, we argue, contribute to the way MFIs boards are structured.

We specifically focus on three factors: regulatory status, international influence and founder

management.

Effect of Regulatory Status

The fact that MFIs can both be regulated or unregulated is one unique characteristic of this

industry. Regulated MFIs are required to have defined shareholders (which non-profits do not

have) and they are regulated by central banking authorities. Regulated MFIs operate on a for-

profit basis and they vary in terms of types (CMEF, 2005). Some are registered as banks while

others are registered as non-bank financial institutions. Unregulated MFIs operate on a non-profit

basis and they are registered as non-governmental organizations (NGOs) (Mersland, 2009b). The

difference between the two is that regulated MFIs are allowed to offer more services than

unregulated MFI– typically deposit taking.

Corporate finance literature evidences that banks have larger boards than other organizations

(Adams & Mehran, 2003, 2008). This is due to shareholders who, according to agency theory,

prefer to sit on boards in order to protect their interests. On the other hand, literature on non-

profit organizations evidence similarly at nonprofit boards are large comparing to other

conventional boards, but small compared to banks (Alonso et al., 2006; Hartaska, 2005). This is

because NGOs are ownerless, and they have many stakeholders involved that prefer to be

represented on the boards. For example, the average board size of corporate bank board was 18

members (Adams & Mehran, 2003), while that on NPO study was 11 members (Alonso et al.

2009). Both boards are large comparing to common corporate governance recommendations of

eight board members (Jensen, 2003). Comparing regulated MFIs with unregulated MFIs, we

expect to see larger boards in regulated MFIs. However, we argue that MFI CEOs are able to

benefit from resources provided by large boards, at a cost of being less monitored. Therefore:

Hypothesis 1a: There is a positive relationship between regulated MFIs and board size.

12

Regulated MFIs have greater flexibility on leveraging and seeking funding from various sources,

then unregulated MFIs. Different funding sources bring in various funders who need to monitor

the usage of their funds (Hartarska & Nadolnyak, 2007). Regulated MFIs are required by law to

have independent board members with prescribed skills. In East Africa for example, the law

prescribes that at least 75% of board members should not be current employees of the

organization (Bank of Tanzania, 2005). Adams and Fereira (2003) studied US corporate

organizations and report that banks have more independent board members than other

organizations. Adams and Mehran (2008) examined banks boards in the US and show banks are

dominated by outsiders. This suggest that for regulated MFIs, as they attract for outside funding

and complying to regulatory requirements, would opt for more independent boards. Thus:

Hypothesis 1b: There is a positive relationship between regulated MFIs and board

independence.

Mersland and Strøm (2009), report that unregulated MFIs have a high rate of female insider

board members (CEOs). Cull et al (2008) further show that unregulated MFIs have more female

customers than regulated MFIs, because of their social driven objectives. More female customers

in unregulated MFIs could be a reason for MFIs to have more female board members in order to

benefit from resource access. Steane & Michael (2001) provide further evidence that boards of

NPOs (similar to unregulated MFIs) possess greater gender diversity than corporate boards. It

then appears that unregulated MFI boards are more gender diverse compared to regulated MFIs.

Therefore:

Hypothesis 1c: There is a negative relationship between regulated MFIs and board

gender diversity.

International Influences

There is significant international influence in the microfinance industry. These influences come

in various forms such as international funding providers, international knowledge transfers and

international networks. For example, (Mersland et al., 2011) report 38% of microbanks in their

global dataset were internationally initiated. In addition, (CGAP, 2010) reports that East Africa is

the leading region with international influence in terms of funding and initiators. International

13

business research shows that internationalization of for-profit firms tends to produce

organizations with better performance (Oxelheim & Randøy, 2003; Riahi-Belkaoui, 1998). This

is contributed through among others, provision of resources from international partners and/or

better governance mechanisms. Similarly, Sanders and Carpenter, (1998) suggests that

international influence is a way of strengthening interdependencies among organizations. We

suggest that this means that MFIs with international influences are more likely to have larger

boards - as this facilitates the need for more resources. Thus:

Hypothesis 2a: There is a positive relationship between MFIs with international influence

and board size

The typical international partner in the microfinance industry is large international NGOs such as

CGAP, FINCA international, Oiko Credit, Accion international, BRAC international and so forth

(CGAP, 2008; Rhyne, 2005). Most of these partners play roles in initiating and owning MFIs.

They also contribute in selecting who should sit on boards. As mentioned earlier, it is common to

find same people sitting on various boards in several countries. These board members are

insiders as they are affiliated with the MFI. The governance literature further, asserts that

insiders are beholden to a CEO and may interfere with vigilant monitoring. However, Baysinger

and Hoskisson (1990) argued that the use of insiders on for-profit corporate board may positively

affect its ability process complex information and to make appropriate strategic decisions. In this

case, organizations with international influence possess unique information of which they need

to share within the board (Sanders & Carpenter, 1998). It is further argued that, the more boards

have access to information (information supply and processing), the more informed will be their

decisions (Sanders & Carpenter, 1998). We assert that an MFI with international influence will

largely influences the information-processing demands placed on it through having a less

independent board. Therefore:

Hypothesis 2b: There is a negative relationship between MFIs with international influence

and MFI board independence

Most international microfinance partners have a social driven mission of supporting MFIs to

reach women and the poor (Rhyne, 2005). Furthermore, since most international supporters of

14

microfinance come from the developed world, they value the importance of gender balanced

boards as advocated by the corporate literature. For example, (CMEP, 2005) advocated for a

gender balanced boards of MFIs. In addition, (Mersland et al., 2011) find international influence

to positively influence social results of MFIs of which among them is to serve female customers.

The authors argue that international partners are more driven with social motivations, and this we

argue, could lead to the promoting of more female on boards. Therefore:

Hypothesis 2c: There is a positive relationship between MFIs with international influence

and board gender diversity.

Founder Management

Anecdotal evidence, such as the case of Muhammed Yunus and the Grameen Bank–indicates

that MFIs founder CEOs are critically important organizational players in MFIs. The importance

of the founders can be due to the relatively newness of the industry. Since the industry is young,

founders are around and typically exert significant influence (Alonso et al., 2009). We argue that

founder CEOs have a tendency of treating the organization as their “baby” and want to hold on

to and control it as long as possible. Corporate research shows that founder CEOs have different

agency issues and they have influence over the choice of board structure (Wasserman, 2003).

This is due to both their formal position and their history with the organization. Wasserman

(2003) suggest that founder CEOs and co-founders exert power through board sets and over

strategic decision making. Agency theory advocate for small boards since they typically are

better CEOs monitors (Yermack, 1996). Increased monitoring might not be in the interest of

founder CEOs, but rather they would opt for small boards in order to have greater control of the

board’s decision making (Daily & Dalton, 1993). Furthermore, corporate literature evidence that

founder managed for-profit organizations are associated with small boards (Certo, Covin, Daily

& Dalton, 2001b). Based on the arguments above, we suggest that founder CEOs do not want to

lose influence and thus opt for small boards. Thus:

Hypothesis 3a: There is a negative relationship between founder managed MFIs and

board size.

Most pertinent research suggests that inside board members provide an important complement to

the founder CEOs (Certo et al., 2001a). Founder CEOs, as compared to non-founder CEOs, are

15

likely to be less objective in the assessment of their organizations (Cooper et al., 1988). They

believe that inside board members are able to provide an additional source of high-quality,

organization-specific information than by outside board members who are less familiar with the

organization (Certo et al., 2001). Literature further suggest that founder CEOs prefer insiders,

due to their detailed specific knowledge, and are likely to be superior to outsiders in assisting

CEOs with defining and maintaining a clear strategic focus for their organizations (Daily &

Dalton, 1993). Previous studies evidence that founding managed organizations provide special

kind of corporate governance that offers lower agency costs and higher performance (James,

1999; Mishra et al., 2001). Similarly, in microfinance industry, we expect founder CEOs to be

more concerned with having board members who have organization specific knowledge that can

enhance decisions made. This was also suggested by (CMEF, 2005) where they argue for inside

representation on MFIs boards in order to contribute to dual objectives. Thus:

Hypothesis 3b: There is a negative relationship between founder managed MFIs and board

independence.

As mentioned earlier, in our data set 54% of the CEOs are founder CEOs. 28% of these founder

CEOs are female. As (Johannisson & Huse, 2000) puts it, `it is the CEO, not the shareholders,

who recruit board members, when you recruit people, you recruit people in your own image,

people you trust and can easily communicate with’. Since this might also be the case with respect

to MFIs, we expect to see founder CEOs use their connections and invite people from their

network. Adams, (2009) suggested that male board members or CEOs use their connections with

female and invite them on boards. This can also be the same with female CEOs who could use

their connections with women and invite them on boards. Furthermore, gender balanced boards

are evidenced to be beneficial in terms having a positive work environment (Huse & Solberg,

2006), having more participative boards (Pearce & Zahra, 1991), and having access to a local

knowledge (Adams & Flynn, 2005). CEO founders can need to benefit from these advantages.

Thus:

Hypothesis 3c: There is a positive relationship between founder managed MFIs and board

diversity.

16

DATA AND DESCRIPTIVE RESULTS

Data

Data for this study is obtained from various sources. The major part comes from a hand collected

survey conducted by the researchers. This survey was conducted in East Africa between

December 2009 and July, 2010. We contacted umbrella associations of MFIs in respectful

countries and received MFIs directories. We used directories to identify the contacts and extract

information concerning all MFIs. Directories have information on names of MFIs and CEOs,

number of customers and contact addresses. We identified all MFIs listed in directories,

excluding cooperatives. Afterwards, we contacted and visited the identified MFIs with a check

list of questions on governance, board members demographics and MFI operations. We further

requested audited financial statements of between 2004 and 2009 for each MFI visited. Of 103

MFIs identified and contacted, 49 MFIs (47.6%) responded, and we obtained governance and

financial information for available years. Additionally, we used a snowball sampling (Saunders

et al., 2003), whereby after visiting one MFI, the CEO or the interviewed person gave us contacts

of another MFI that was, or was not, in our original list. We then visited these MFIs that fit our

categorization. In terms of information content, some MFIs in East Africa publish information on

governance and audited financial reports on their websites. We visited the available websites to

extract available information. To further check the quality of our information - we used the

information from these websites to compare with the data obtained from the survey. The above

data collection process led to a total of 63 MFIs with demographic information of 465 board

members and information of between 2004 and 2009.

These MFIs may represent between a third and a quarter of MFIs in the region. However, we

recognize that since most of the data was based on MFIs directories, it is possible that some

MFIs were left out and that we have sample biasness. We argue that our sample MFIs represents

the major and most established players in regional microfinance industry, since they are

members of umbrella associations and more visible on the web. Table 2 below shows MFIs

surveyed in the region. 22 MFIs are from Kenya and 22 MFIs are from Tanzania, while 19 MFIs

are from Uganda. Among these, 37 are unregulated MFIs and 26 are regulated MFIs.

------------------------------------ Insert Table 2 about here ------------------------------------

17

Descriptive Results

Table 3 below report definitions and summary statistics of variables and measures used in this

study. The average MFI in our sample has 7 board members, with a minimum of 2 and

maximum of 14 board members. 36% of all board members are independent which means most

of board members are insiders. Out of 465 board members, 127 of them (25%) are female. This

provides a preliminary picture that MFIs board are not gender balanced but have higher percent

of female board members. 39% percent of MFIs are regulated. 65% of MFIs in this sample have

international networks either as funders or for other relationships. Descriptive results also show

that 54% MFIs are managed by founders. The average age of MFI is 9 years, but few MFIs are

old to 31 years. Few MFIs (27%) are internationally initiated despite the fact that most of them

have international networks. In terms of country variables, we report that 34% of the MFIs are

from Kenya and 30% from Uganda, remaining being from Tanzania (36%).

------------------------------------ Insert Table 3 about here ------------------------------------

Table 4 presents correlations between all variables. First, we look at correlations among board

structure variables. Since they are all measures of board structure, their correlations are

significant. However, since they measure different aspects of board structure, we will run

separate regressions for each of the dimensions of board structure. The correlates provide a first

simple test of our hypotheses as we see that a number of them are consistent with our

hypotheses. For example, regulated MFIs are positively and significantly correlated with board

size. We also see that founder CEO is negatively correlated with board size, and board

independence, as we hypothesized.

As we prepare for multivariate tests, we now turn to the question of multicollinearity among

independent variables. However, correlation coefficients among our independent variables are

rather low. Kennedy, (2008) holds that correlations need to be above 0.7 to detect

multicollinearity among two variables. None of the correlation coefficients are of that magnitude.

However, significant coefficients are a warning signal that multicollinearity problems may arise

when all variables are included simultaneously. To reduce this problem, we run regressions with

few variables at a time (similar to Strøm et al., 2010). In addition, since panel data estimation

18

gives more data points, the multi-collinearity problem here is reduced even further (Wooldridge,

2006).

------------------------------------ Insert Table 4 about here ------------------------------------

FINDINGS

Regression Methods

We first perform single-equation regressions for each board structure variable followed by

system regressions that take joint estimations of all dependent variables. Our data consists of a

panel of six years and all three dependent variables are continuous in nature. In this case, we can

use either fixed effects model or random effects, depending on time invariant of variables. Fixed

effects regression is the model to use when one want to control for omitted variables that differ

between cases, but are constant over time. It allows the use of changes in the variables over time

to estimate the effects of independent variables on the dependent variable, and it gives consistent

results (Wooldridge, 2006). On the other hand, some omitted variables may be constant over

time but vary between cases, and others may be fixed between cases but vary over time. To

tackle this challenge, one may use random effects in order to take into considerations both effects

and to get efficient estimators. To choose whether we can use any of the two models, we run a

Hausman test (Hausman, 1978) which tests whether the null hypothesis that the coefficients

estimated by the efficient random effects estimator are the same as the ones estimated by the

consistent fixed effects estimator. The results of this test reject the null hypotheses, and therefore

we were able to use random effect model with the generalized least square (GLS) estimation

methodology.

Next we use system regressions. System estimation is applicable when we have related

measures of an underlying true variable that depends upon a set of explanatory variables. Table 4

shows that this is the situation here, among board structure variables. We therefore use panel

data estimation with the seemingly unrelated regression (SUR) methodology (Greene, 2008,

Mersland, et al., 2011), a procedure that takes account of possible correlations among the

dependent variables. We further perform a Breusch-Pagan test to check the extent to which the

residuals in the SUR regression are independent.

19

A common challenge in corporate governance research is the issue of possible reverse

causality. Specifically, we run robustness checks by re-running all models using instrumental

variables with a two stage linear regression (SLS) to address this issue. Another methodological

concern is the possible omitted unobservable variable that affects both independent and

dependent variables (Adams & Ferreira, 2009; Adams et al. 2010). For example, some MFIs are

more progressive than others, so they get international networks faster than others. This effect of

omitted variables could lead to spurious correlations among independent and dependent

variables, and thus we have to control for this possibility. Random effect model is argued to

reduce these concerns (Wooldridge, 2006).

Results and Discussion

Table 5 and 6 report our econometric findings for both random effects (RE) estimations and

seemingly unrelated (SUR) estimations, respectively. We use both tables to provide and explain

our results. We find support for hypotheses 1a in relation to board size and 1c in relation to board

diversity. Regulated MFIs have larger boards than unregulated MFIs. This is in support of what

(Adams & Mehran, 2003, 2008) found for banks in the US. This implies that despite unregulated

MFIs typically have a broader set of stakeholders than regulated MFIs, it looks like they need

larger boards to assist them with getting better resources access. We also evidence that regulated

MFIs are associated with less gender diverse boards. It seems that, gender balanced board is a

less concern for regulated MFIs, despite several call on this issue (CMEF, 2005). Agency theory

suggests that small and independent boards are better at monitoring (Wagner III et al., 1998;

Yermack, 1996). But for MFIs, it seems like regulated MFIs prefer to get more resources from

larger boards, despite that this also produce less board independent and less board diversity. This

we argue is contributed by the nature of regulated MFIs whereby they have defined shareholders

who may use their power to be stetted on the board, even at the “expense” of less board

independence and less board diversity.

We find evidence that international influence affect MFIs board structure. Tables 5 and 6

provide support for hypothesis 2a; that MFIs with international influence have larger boards.

This supports the theoretical argument that large boards provide enhanced resources access

(Hillman et al., 2009). Specifically for MFIs, when international connections are involved, their

20

aim is to make sure that MFIs have needed resources for survival (Schreiner, 2000). So they

would opt for a larger board that would bring those resources. Results further provide support for

hypothesis 2b; that international influence is associated with less independent boards. Again, for

resource reasons, that insiders have critical information that is important for MFIs. In line with

hypothesis 2c, we evidence that MFIs with international influence have greater gender diversity.

This implies that international sources advocate for gender biased boards and they believe that

female on boards bring in values and resources to MFIs or possibly, that this is a policy driven

by gender equally policies of governments and investors from the North. Our conclusion as to

the international influence on MFI board structure is that, international sources are more

concerned on the resource brought in by board members. In fact, there is a stream of literature

that supports the inclusion of more insiders on boards, and thus less board independence

(Osterloh & Frey, 2005; Wagner III et al., 1998). These researchers argue that insiders have

more knowledge and information from the market which can be useful for the organization

(Zahra, 1996). We suggest that some of the same motivations might apply in the microfinance

industry.

We find support for hypotheses 3a and 3b; that founder managed MFIs are associated with

smaller boards and less board independence. As Wasserman (2003) noted, Founder CEOs

typically hold a board seat (as well as significant influence on board selection) and this gives

them power over strategic decision making. Here we argue, founder CEOs opt for small boards

because it is easier for them to be accountable to fewer board members, as well as it easier to

control decision making within small groups. For example, when CEO have their agenda and

they want a board consent, it is easier for the CEO to speak informally to the individual board

member when there are few than many. Founder CEOs are also associated with less

independence board structure as shown in both tables. For resources reasons, Founder CEOs

might believe that inside board members provide additional source of MFI -specific information

than outside board members (as suggested in for-profit organizations by (Certo et al., 2001)). We

also argue that founder CEOs prefer less independent boards because it is easier to be

accountable to people who are more familiar to you that to less familiar ones. This may also

results into CEO control. We also find support that founder managed MFIs are associated with

more board gender diversity. This confirms Strøm’s et al (2010) assertion that female leadership

21

has an impact on governance of MFIs. There are many female CEO in MFIs and some of these

are founder CEOs who would influence board diversity by suggesting female to be on boards.

For example, Strøm et al.(2010) report 27% of CEOs in their global data set to be female which

is very close to our figure of 26%. In addition, we argue that since CEOs know that a large

proportion of their main customers are female, they would like to see their board composed of

female members. The argument is that female board members can have a deeper understanding

of female customers’ needs (Mersland & Strøm, 2009).

------------------------------------ Insert Table 5 about here ------------------------------------

------------------------------------ Insert Table 6 about here ------------------------------------

In summary, our results suggest associations between MFI unique characteristics and board

structure. We specifically evidence that East African MFIs put less emphasis on the monitoring

aspects of board structure (such as small boards and board independence) - as advocated by

agency theory. Alternatively, board structure that is better for resource and advice as advocated

by resource dependency theory appears to be preferred in this relatively young and

entrepreneurial industry. We evidence that board size increase when the MFI is; regulated and is

member of an international network. On the other hand, if a MFI is managed by a founder CEOs,

its board will be smaller and less independent. These suggest that board selection in MFIs is

more driven by resource access than monitoring concerning. This again might be a result of the

youngness of the industry. As also evident from the effect of our control variables, we see that

age is positively associated with board independence. This means as a MFI grow older, it would

tend to attract more independent board members. Since the average ages of MFI in our dataset is

9 years, most are still young and struggle to get access to resources, the recruitment of

knowledgeable internal board members is still the norm. We also evidence that board gender

diversity is becoming more common in the industry. We see that when MFIs are part of

international networks and founder managed; they tend to have more females on boards than

otherwise. This further suggests that female board members, similar to female customers, are

becoming important in the industry and that calls for more gender balanced board is also

gradually implemented.

22

Estimations that account for endogeneity

As part of robustness checks, we rerun our models by means of instrumental variables (IV)

methods. Among three independent variables, we argue that founder management is an

exogenous variable. When CEOs are founders, it means they have been there since the MFI

begins its operations and thus, the possibility of reverse causality between founder CEO and

board structure is low (Adams & Ferreira, 2009). We further confirm this by testing all

independent variables for endogeneity using Durbin-Hu Hausman test. The remaining two

variables, regulation status and international influence resulted to be endogenous. To solve this

problem, we first hypothesize that MFI size and international initiated could be valid instruments

for these endogenous variables. Based on previous literature, it is possible that MFIs which were

internationally initiated are the ones able and easily to get more international networks and

expand their services. Size in terms of assets is another variable we use as an instrument. Large

MFIs are in a better position to change their status into becoming regulated and secure more

funding through debt and shareholding. We test the correctness of these instruments and run a

two-stage instrumental variable (SLS) of Durbin-Hu Hausman (Wooldridge, 2006). Results are

presented in table 7 below.

The SLS analysis provides similar results for most of the hypothesized relations. We still find

support that regulated MFIs are positively associated with large boards and negatively with

board gender diversity and that MFIs with international influence have larger and less

independence boards. However, now there isn’t a significant association between international

influence and board gender diversity, as the case in our previous test shown in Table 5 and 6.

With founder management, the SLS results are closely consistent with previous estimations. This

could further support our argument that this variable is exogenous. Overall, we also highlight

that our results justify the need to address the three dimensions of MFIs’ board structure within

one framework.

------------------------------------ Insert Table 7 about here ------------------------------------

23

CONTRIBUTION AND CONCLUSION

This study examines the determinants of board structure in Microfinance Institutions (MFIs). We

specifically examine three unique characteristics of MFIs: regulation status, international

influence and founder management. Using a sample of 63 East African MFIs, we are able to

show the relationship between these characteristics and three dimensions of board structure:

board size, board independence and board gender diversity. We used a mixture of analytical

procedures in order to get valid and robust results. First, we perform single measure regression of

each board structure variable in relation to MFI characteristics, using random effects estimations.

We then perform joint estimations using a seemingly unrelated (SUR) regression analysis. We

handle endogeneity problems by using a two stage least square regressions (SLS) with

instrumental variables as our robustness checks. Main instruments used for endogenous variables

are whether the MFI was internationally initiated and MFI size.

Results show that there are associations between MFI characteristics and board structure. We

suggest board selection in this young and entrepreneurial industry (Strøm et al., 2010) appears to

be more motivated by a need for resource access and insider knowledge than for outside

monitoring. We find that MFIs which are regulated have large boards and are less gender

diverse. This supports the argument from the banking literature (Adams & Mehran, 2008).

Similarly, MFIs that are members of international networks (either in terms of funding and/or

technical assistance) have large, diverse and less independence boards. Past research highlight

that international partners associate themselves with MFIs in order to provide them with

resources such as funding, international experience through technical assistance and advice

(Mersland et al., 2011). Therefore when they have relationships with MFIs, they would want to

see their boards satisfy the need for resources. Founder managed MFIs bring similar evidence to

the corporate founder CEO literature (Certo et al., 2001). We evidence that founder managed

MFIs are associated with small and less independent boards. This is a result of the control that

founder CEOs have. Also, founder CEOs prefer less independent boards because insiders possess

organizational specific knowledge which is needed by CEOs in order to have better organization

results (Certo et al., 2001). Board diversity is becoming common in MFIs. We specifically

evidence that MFIs that are part of international networks, and founder managed MFIs, to be

24

positively associated with a high proportion of female board membership. This suggests that

MFIs are aware of the important information/knowledge that female board members provide.

From these results, we conclude that, MFI board selection is particularly motivated by

resources access. Apparently, MFIs, and ultimately the stakeholders that affect board

composition, are less concerned with the board’s ability to monitor the CEOs. They want MFIs

to get resource access that would enable them survive and serve poor people better.

Organizational and industry specific knowledge that is brought on boards by insider board

members appears to be a necessity for MFIs. Board independence, as a good corporate

governance mechanism of monitoring, seems to be less of a concern for MFIs selection of board

members. This, we argue is explained by the industry youngness and common social mission of

microfinance activities.

We contribute to the literature by showing that unique industry determinates affect board

structure in the microfinance industry. We apply resource dependence theory (Hillman &

Dalziel, 2003; Pfeffer, 1973) by showing how unique microfinance characteristics affect board

structure. We also apply agency theory to address monitoring aspects of MFI boards. However,

apparently not all MFI emphasize a board structure that is supposed to be better for monitoring

(particularly of CEO). It seems like small and young MFIs are in more need for resources; and as

they grow, they may opt for a board structure that is better for monitoring.

Our study is based in East Africa and within the time frame of 2004 to 2009. Though we

expect that the MFI characteristics of this study should be relevant in other parts of the world, we

see this as a limitation. There might be some other determinants of board structure that should be

considered when the study is conducted globally, for example the nature of the corporate

governance environment (such as the legal and accounting requirements), and the level of

financial development of the country. Furthermore, the chosen explanatory variables in this study

might be considered in more detail. For example, the cultural and institutional distance between

the MFI and its international partner/network, the specific background of the CEO founder, the

content of regulations, and finally, a more detailed picture of the MFIs customer environment.

However, we leave this for future research and highlight the need for similar studies in other

regions.

25

References

Adams, R., & Ferreira, D. 2009. Women in the boardroom and their impact on governance and performance. Journal of Financial Economics, 94(2): 291-309.

Adams, R., & Mehran, H. 2003. Is Corporate Governance different for bank holding companies? Economic Policy Review, 9(1): 191.

Adams, R., & Mehran, H. 2008. Corporate Performance, Board Structure,and Their Determinants in the Banking Industry. In F. R. B. o. N. Y. S. Reports (Ed.).

Adams, S. M., & Flynn, P. M. 2005. Local Knowledge Advances Women's Access to Corporate Boards. Corporate Governance: An International Review, 13(6): 836-846.

Alonso, A., Martín, C. N., & Romero-Merino, M. E. 2006. The governance of nonprofit organizations : Empirical evidence from nongovernmental development organizations in Spain. Nonprofit and Voluntary

Sector Quarterly, 35(4): 588.

Alonso, P., Palenzuela, V., & Merino, E. 2009. Determinants of Nonprofit Board Size and Composition : The Case of Spanish Foundations. Nonprofit and Voluntary Sector 38(5): 784 - 809.

Baysinger, B., & Hoskisson, R. E. 1990. The Composition of Boards of Directors and Strategic Control: Effects on Corporate Strategy. The Academy of Management Review, 15(1): 72-87.

Boone. 2007. The determinants of corporate board size and composition : An empirical analysis. Journal

of Financial Economics, 85(1): 66.

Burgess, Z., & Tharenou, P. 2002. Women Board Directors: Characteristics of the Few. Journal of

Business Ethics, 37: 10.

Campion, A., & Frankiewicz, C. 1999. Guidelines for the Effective Governance of Microfinance Institutions. Washington DC:: The Microfinance Network.

Certo, T., Covin, J., Daily, C., & Dalton, D. 2001b. Wealth and the effects of founder management among IPOs stage new ventures. Strategic Management Journal, 22(6-7): 641.

CGAP. 2008. Foreign Capital Investment in Microfinance: Balancing Social and Financial Returns. Focus Note 44: 24: CGAP.

CGAP. 2009. Africa Microfinance Analysis & Benchmarking Report Mix Market.

Chen, H.-L. 2011. Does Board Independence Influence the Top Management Team? Evidence from Strategic Decisions toward Internationalization. Corporate Governance: An International Review: 17.

CMEF. 2005. The Practice of Corporate Governance in Shareholder—Owned Microfinance Institutions. In C. o. M. E. Funds (Ed.): 35: Council of Microfinance Equity Funds: Washington, DC: Microfinance Gateway.

Coles, L., Daniel, D., & Naveen, L. 2008. Boards : Does one size fit all. Journal of Financial Economics,

87(2): 329-356.

26

Conger, J., Finegold, D., & Lawler, E. 1999. Boardcomposition:Member selection and appointment. In M. Network. (Ed.), Guidelines for the Effective Governance of Microfinance Organizations 9-16: Microfinance Network.

Cull, R., Morduch, J., & Demirguc-Kunt, A. 2009a. Microfinance Meets the Market. Journal of

Economic Perspective, 23(1): 25.

Daily, C., & Dalton, D. 1993. Board of Directors Leadership and Structure: Control and Performance Implications. Entrepreneurship: Theory & Practice, 17(3): 17.

Daily, C. M., Certo, S. T., & Dalton, D. R. 1999. A Decade of Corporate Women: Some Progress in the Boardroom, None in the Executive Suite. Strategic Management Journal, 20(1): 93-99.

Erhardt, N., Werbel, J., & Shrader, C. 2003. Board of director diversity and firm financial performance. Corporate Governance: An International Review, 11(2): 102.

Hartarska, V., & Nadolnyak, D. 2007. Do regulated Microfinance Institutions achieve better sustainability and Outreach? Cross-Country Evidence. Applied Economics, 39(10-12): 15.

Hartaska, V. 2005. Governance and Performance of Microfinance Institutions in Central and Eastern Europe and the Newly Independent States. World Development, Vol. 33(10): pp. 1627–1643.

Hartaska, V., & Mersland, R. 2009. Which Governance Mechanisms Promote Efficiency in Reaching Poor Clients? Evidence from Rated Microfinance Institutions. European Financial Management.

Hillman, A., & Dalziel, T. 2003. Boards of Directors and Firm Performance: Integrating Agency and Resource Dependence Perspectives. The Academy of Management Review, 28(3): 383-396.

Hillman, A. J., Withers, M., & Collins, B. 2009. Resource Dependence Theory: A Review. Journal of

Management: 25.

Huse, M., & Solberg, A. G. 2006. Gender - related boardroom dynamics : How Scandinavian women make and can make contributions on corporate boards. Gender in Management: An International Journal,

21(2): 113.

Jackling, B., & Johl, S. 2009. Board structure and firm performance : evidence from Indias top companies. Corporate Governance: An International Review, 17(4): 492-509.

Jensen. 1993. The modern industrial revolution , exit , and the failure of internal control systems. Journal

of Finance , The, 48: 831.

Jensen, M., & Meckling, W. 1976. Theory of the firm: Manegerial behaviour, agency costs and ownership structure. Journal of Financial Economics, 3: 55.

Johannisson, B., & Huse, M. 2000. Recruiting outside board members in the small family business : An ideological challenge. Entrepreneurship and Regional Development, 12(4): 353.

Johnson, J., Daily, C., & Ellstrand, A. 1996. Boards of directors : A review and research agenda. Journal

of Management, 22(3).

27

La Porta, R., Lopez, F., Shleifer, A., & Vishny, R. 1997. Legal determinants of external finance. Journal

of Finance , 52(3): 1131.

Labie, M., & Mersland, R. 2009. Corporate Governance Challenges in Microfinance. 2nd International

Workshop on Microfinance Management and Governance: 13. University of Agder, Norway.

Lehn, K., Patro, S., & Zhao, M. 2009. Determinants of the size and structure of corporate boards : 1935 - 2000. Journal of the Financial Management, 38: 747.

Linck, J. S., Netter, J. M., & Yang, T. 2008. The determinants of board structure. Journal of Financial

Economics, 87(2): 308-328.

Lorsch, J., & Maclver, E. 1989. Pawns or Potentates: The Reality of America's corporate Boards. Cambridge: MA: Harvard Business School Press.

Markarian, G., & Parbonetti, A. 2007. Firm complexity and board of director composition. Corporate

Governance: An International Review, 15(6): 1224-1243.

Mersland, R. 2009b. The governance of non-profit micro finance institutions: lessons from history. Journal of Management and Governance: 22.

Mersland, R., Randøy, T., & Strøm, R. Ø. 2011. The impact of international influence on microbanks' performance: A global survey. International Business Review, 20: 13.

Mersland, R., & Strøm, R. Ø. 2008. Performance and Tradeoffs in Microfinance Organizations: Does Ownership Matter. Journal of International Development, 20: 14.

Mersland, R., & Strøm, R. Ø. 2009a. Performance and Governance in Microfinance Institutions. Journal

of Banking and Finance, 33(4): 17.

Milton, H., & Raviv, A. 2008. A Theory of Board Control and Size. The Review of Financial Studies,

21(4): 1797-1832.

Mishra, C., Randøy, T., & Jensen, J. 2001. The effect of Founding family influence on firm value and corporate governance: A study of Norwegian firms. Journal of International Financial Management and

Accounting, 12(3):235-259.

Morck, R., & Yeung, B. 1991. Why investors value multinationality. Journal of Business ( Spokane,,

Wa), 64(2): 165.

Olomi, D. 2008b. Performance of Microfinance Institutions and Impact to Socio-Economic Development: A case of Zanzibar. Dar es Salaam: International Labour Organization.

Osterloh, M., & Frey, B. 2005. Shareholders Should Welcome Employees as Directors. Zurich: Institute for Empirical Research in Economics.

Oxelheim, L., & Randøy, T. 2003. The impact of foreign board membership on Firm value. Journal of

Banking & Finance, 23(12): 27.

28

Pearce, J., & Zahra, S. 1991. The relative power of CEOs and boards of directors : Associations with corporate performance. Strategic Management Journal, 12(2): 135.

Pearce, J., & Zahra, S. 1992. Board composition from a strategic contingency perspective. Journal of

Management Studies, 29(4): 411.

Pfeffer, J. 1973. Size, Composition, and Function of Hospital Boards of Directors: A Study of Organization-Environment Linkage. Administrative Science Quarterly, 18(3): 16.

Pfeffer, J., & Salancik, G. 1978. The External Control of Organizations: A Resource-Dependence

Perspective. New York: Harper & Row.

Raheja, C. 2005. Determinants of board size and composition : A theory of corporate boards. Journal of

Financial and Quantitative Analysis, 40(02): 283.

Randhawa, B., & Gallardo, J. 2003. Microfinance Regulation in Tanzania: Implications for Development and Performance of the Industry. African Region Working Paper Series.

Rhyne, E. 2005. Perspectives from the Council of Microfinance Equity Funds. Enterprise Development

and Microfinance, 16(1): 9.

Riahi-Belkaoui. 1998. The effects of the degree of internationalization on firm performance. Scandinavian International Business Review, 7(3): 315.

Sabana, B. 2006. Governance: Organizing, Developing and Empowering Boards to Oversee MFI Operations. Nairobi: Equity Bank.

Sanders, G., & Carpenter, M. 1998. Internationalization and firm governance : The roles of CEO compensation , top team composition , and board structure. Academy of Management Journal, 41(2): 158.

Saunders, M., Lewis, P., & Thornhill, A. 2003. Research Methods for Business Students (3rd ed.). Sydney: Prentice Hall.

Schreiner, M. 2000. Ways Donors Can Help the Evolution of Sustainable Microfinance Organizations. 23. St. Louis.

Steane, P., & Michael, C. 2001. Nonprofit boards in Australia : A distinctive governance approach. Corporate Governance: An International Review, 9(1): 48.

Steier. 1998. Confounding market and hierarchy in venture capital governance : The Canadian immigrant investor program. Journal of Management Studies, 35(4): 511.

Strøm, Ø., D'Espallier, B., & Mersland, R. 2010. Gender, Performance and Governance in Microfinance Institutions 3rd International Workshop on Microfinance Management and Governance: 36. University of Groningen.

Tallman, S., & Li, J. 1996. Effects of international diversity and product diversity on the performance of multinational firms. Academy of Management Journal 39(1): 179.

29

Terjesen, S., Sealy, R., & Singh, V. 2009. Women directors on corporate boards : A review and research agenda. Corporate Governance: An International Review, 17(3): 320-337.

Wagner III, J. A., Stimpert, J. L., & Fubara, E. 1998. Board composition and organizational performance:Two studies of insider/outsider effects. Journal of Management Studies, 35(5): 655.

Wasserman. 2003. Founder - CEO succession and the paradox of entrepreneurial success. Organization

Science, 14(2): 149.

Wooldridge, J. 2006. Introductory Econometrics: A modern Approach (3 ed.). Ohio: Mason Ohio: Thomson South Western.

Yermack, D. 1996. Higher Valuation of Companies with a small board of directors. Journal of Financial

Economics, 40.

Tables

Table 1: Comparing MFIs from the global rating reports and our East Africa data

Rating Reports – 379 MFIs

(Mersland et al., 2011)

East Africa Data- 63 MFIs

This dataset

Mean Median Mean Median

Regulated MFIs 0.25 0.00 0.39 0.00 Size- Assets (mill USD) 10.50 3.07 9.84 4.70 Age-Years 9.00 7.00 9.69 8.00 International network member 0.33 0.20 0.66 1.00 Board gender diversity-proportion 0.28 0.20 0.25 0.22 Founder CEO 0.38 0.31 0.54 1.00

Table 2: MFIs surveyed

This table reports on the description and number of MFIs surveyed.

Country Unregulated NGO-MFIs Regulated SHFs-MFIs Total

Kenya 14 8 22 Uganda 9 10 19 Tanzania 14 8 22 Total 37 26 63

30

Table 3: Variables and Summary statistics

Variables Definition/Measure Mean S. dev Min Max Obs.

Dependent variables

Board size Logarithm of number of members 7.184 2.237 2 14 331

Board independence Proportion of board members who are not

affiliated by the MFI

0.365 0.256 0 1 325

Board gender diversity Proportion of female board members 0.250 0.226 0 1 320

Independent variables Regulated-SHFs Binary variable equal one if the MFI is a

regulated

0.387 0.487 0 1 328

International influence Binary variable equal one if the MFI is member

of any international network

0.659 0.475 0 1 343

Founder management Binary variable equal one if the CEO of the MFI is a founder

0.543 0.499 0 1 315

Control variables MFI Size (Ln Assets) Logarithm of total assets 15.578 2.517 8.411 24.902 255

MFI Age Number of years since the MFI starts services 9.697 6.960 0 31 311

International initiated Binary variable equal one if the MFI is initiated

internationally

0.271 0.445 0 1 343

Kenya Binary variable equal 1 if MFI is from Kenya 0.344 0.475 0 1 343

Uganda Binary variable equal 1 if MFI is from Uganda 0.300 0.459 0 1 343

Table 4: Correlations of Variables

Bivariate Pearson correlations among board structure and independent MFIs’ variables: Significant correlations at

the 5% are **, at the 1% level are *

Variable B. size Indep. Diversity Regulat. Intern. F. Mgt Age Size

Independence 0.380*

Diversity 0.149* 0.336*

Regulated 0.356* 0.092 -0.045

International 0.256* -0.087 0.173* -0.043

Founder Mgt -0.288 -0.232* 0.045 -0.126** -0.132**

Age 0.514* 0.326* 0.367* 0.454* 0.121 -0.272*

Size 0.434* 0.033 0.039 0.466* 0.058 -0.305* 0.504*

International

initiated

0.054 -0.237 -0.225* -0.072 0.439* -0.139 -0.136** 0.042

31

Table 5 Random Effect Regression of Board structure on MFIs variables

This table shows single measures of board structure variables on MFI characteristics and controls. Asterisks indicate significance at 0.01 (*), 0.05 (**), and 0.10 (***) levels.

Variable Board Structure - Single Estimations

Independent Variables Ln-board size Board independence Board diversity

Regulation 0.029 0.042 0.011 International influence 0.161** -0.057 0.059 Founder Management -0.048 -0.069*** 0.053*** Control variables Assets – Size 0.020 -0.016 -0.014 Experience – Age 0.023* 0.013** 0.015* Kenya -0.154*** 0.018 -0.012 Uganda -0.145 -0.046 -0.034

Year dummies Yes Yes Yes Overall R-square 0.391 0.201 0.203 F / Wald statistics 42.16 21.62 17.77 Observations 235 235 231

Table 6 Seemingly Unrelated Regression model of Board structure on MFIs Variables

This table shows joint measures of board structure variables on MFI characteristics and controls. Asterisks indicate significance at 0.01 (*), 0.05 (**), and 0.10 (***) levels.

Variable Board Structure - Joint Estimations

Board size Board independence Board diversity

Independent Variables

Regulation 0.085** 0.015 -0.081* International influence 0.181* -0.105* 0.041 Founder Management -0.049*** -0.092* 0.093* Control variables Assets – Size 0.019** -0.016** -0.014** Experience – Age 0.016* 0.014* 0.019* Kenya -0.098** 0.073** -0.045 Uganda -0.142* 0.006* -0.011 Year dummies Yes Yes Yes Overall R-square 0.440 0.249 0.246 F / Chi2 statistics 181.23 76.50 75.26 Observations 231 231 231

Correlations of Residuals

Board Size Board independence

Board independence 0.373 Board diversity -0.059 0.333 Breusch-Pagan chi2 (3) 58.498

32

Table 7 Two Stage Linear regressions (SLS) for board structure on MFI characteristics

This table shows SLS regressions of board structure on MFI characteristics. Asterisks indicate significance at 0.01 (*), 0.05 (**), and 0.10 (***) levels.

Board Structure- SLS Regression

Board size Board independence Board Diversity

Independent Variables

Regulation 0.429** -0.538** -0.334** International influence 0.197*** -0.254** -0.238 Founder Management -0.068 -0.009 0.092** Control variables Experience – Age 0.009 0.028* 0.029* Instruments International initiated; Asset size R Square 0.104 F statistics 17.88 5.32 11.92 Observations 161 180 179