The non-profit board : a concise review of the empirical literature

This article gives a concise review of the literature on non-profit boards. Given the lack of theoretical work on the subject, the focus of the article lies on the empirical work. This empirical work seems to have developed on two roads: one where the non-profit board is the dependent variable and one where the non-profit board is an independent variable in relation to other variables such as organisational performance. As this article will show, these empirical results often remain equivocal and consensus is lacking. The authors argue that a better insight in the composition and structure of the non-profit boards can contribute a lot to the understanding of the behaviour of non-profit organisations.


I. Introduction
Non-profit organisations (NPOs) differ in many ways from for-profit corporations and public institutions.The most distinguishing characteristic is the so-called non-distribution constraint, prohibiting the NPO to distribute its earnings to those who control or own the organisation. 1Consequently, these organisations are governed differently, leading to succinct governance structures.One of the most important governance structure in the NPO, the non-profit board, is the subject of the present paper.Since NPOs are becoming a more and more important sector in the economy, it is important that these organisations have adequate governance structures.Both practitioners and academic scholars recognise this need and by consequence, the literature with respect to non-profit boards has grown, especially in the last decade.It is not our intention to provide the reader with a prescriptive list of best governance practices.For that we refer to the more practitioner-oriented literature, which often provides a more normative approach to non-profit governance.The emphasis of this text is more on the empirical work with respect to non-profit boards.This focus is chosen because, in comparison with theoretical work on non-profit boards, empirical literature is more developed.The paper is organised as follows: Paragraph 2 discusses more thoroughly the concept of non-profit boards with a strong focus on principal-agent relations in NPOs.The third paragraph is entirely devoted to the empirical literature on non-profit boards.Empirical work seems to be developed on two different avenues and we will also use this distincti-

II. The non-profit board
One of the factors leading to a divergence in governance mechanisms is the institutional framework used to identify the role of governance.A distinction often used in the corporate governance literature is the distinction between a stakeholder and a shareholder framework.In a shareholder framework, the role of governance and thus of the board is to represent and defend only the interests of the shareholders.In a stakeholder view on governance, the interests of all possible stakeholders2 must be represented.We argue that in the non-profit context, because of the absence of shareholders and the existence of an heterogeneous group of stakeholders, the latter will be a more appropriate framework.A descriptive study of Australian non-profit boards gives empirical support to this argument. 3n the line of this stakeholder framework, we can define non-profit governance as the control on the NPO to ensure that the interests of all its constituencies are served, which is the most important task of the board.Compared to for-profit firms, NPOs have little choice in governance mechanisms.In contrast to corporate governance, a take-over market to align the manager's decisions with the owner's interests is not available and so is the possibility of profit-sharing plans. 4ince so few alternatives are available, Desai and Yetman analyse the effectiveness of legal and reporting rules on US NPOs in providing effective governance. 5By calculating a state-level index of these legal and reporting rules, the authors find that NPOs located in states where stricter legal rules count, more money is devoted to the charitable activities of the NPO and less money goes to fundraising activities and salaries.So, these results suggest that in the absence of better alternatives, the state can provide a framework for effective non-profit governance by imposing strict legal and reporting rules.Nonetheless, the most obvious governance mechanism available for NPOs is their board of directors.Therefore, the literature on non-profit governance focuses almost entirely on the role of the board and/or the relationship between board and executive staff.Saidel argues that the subject of non-profit governance theory should be extended to other governance participants like advisory groups, which are, according to him, supplementary to the non-profit boards. 6Non-profit governance literature however still seems to emphasise the role of the board.
ZögU 30.Jg. 1/2007 Glaeser argues that the main consequence of "these weak governance institutions on nonprofit behaviour" is that the preferences of non-profit workers are comparatively more important in the organisation than those of workers in the for-profit sector. 7Because governance mechanisms are far more limited in the non-profit context, workers here will have more influence on decisions made inside the organisation.Glaeser's vision of weak non-profit governance is feasible, for other authors ascertain that non-profit boards are often self-perpetuating mechanisms that do not always fulfil the role of a good governor. 8aybe the reason for this lack of governance incentives is a free-rider problem caused by the multiplicity of constituents which makes it not worthwhile to participate in the governance of the organisation.Another possible reason could be the lack of professionalism in governance issues and the diversity of the often voluntary members of the board.9Whatever the reason is for this lack of governance incentives, we believe that a good understanding of the governance of non-profit organisations can prove useful in the development of a non-profit theory.Besides, NPOs are not spared from the agency problems that justify the existence of governance mechanisms.As agency theory postulates, interests and objectives of the people who work for the organisation are not necessarily aligned with the objectives of the owners, regardless of the ownership type.As Slivinski notes, there is no basis to presume that non-profit organisations are spared from "moral hazard, opportunism, adverse selection and other problems of asymmetric information". 10he choice in governance mechanisms for the NPO is thus very limited.The most studied subject in the non-profit governance literature remains the non-profit board and even here, a theoretical framework is not yet well developed.In the next two paragraphs we will therefore review the empirical literature with respect to the non-profit board.The following paragraph discusses the empirical work which treats the non-profit board as an independent variable while the third paragraph focuses on studies with the non-profit board as the dependent variable.

III. The non-profit board as an independent variable
In this section we will review literature where the non-profit board is treated as an independent variable.The relevant question is 'Does the non-profit board have a significant influence on the behaviour and performance of the NPO?'.Following the corporate governance literature, a strong-held assumption is that a good non-profit board, both in its internal role as controlling the institution and in its external role of fund-raiser and linkage with the organisational environment, is essential for the good working of the NPO.We will divide this section, following the conceptual framework of Bradshaw, Murray and Non-profit board ZögU 30.Jg. 1/2007 81 Wolpin, in two parts, i.e. the structural characteristics of the board and the board's process characteristics.11

The board's structural characteristics
With the board's structural characteristics, we refer to elements like size, age or composition of the board.As in the corporate governance literature, a lot of empirical writings have examined the relationship between those characteristics and the performance of the organisation and again, like in the corporate governance literature, unanimous conclusions are lacking.
In their seminal article on the separation of ownership and control, Fama and Jensen write "without the take-over threat or the discipline imposed by residual claimants with the right to remove members of the board, non-profit boards composed of internal agents and outside experts chosen by internal agents would provide little assurance against collusion and expropriation of donations.Thus, non-profit boards generally include few if any internal agents as voting members".12With reference to this article, Williamson proposes that NPOs that have a higher proportion of insider members in their board of directors will be less efficient (in resource allocation). 13These propositions are meanwhile known as the Fama-Jensen-Williamson conjecture that states that for a good internal function of the non-profit board, the board should have a limited proportion of insider members.The reasoning behind this idea is that insiders on the board will try to prevent the board from effectively controlling management behaviour and expenses.Callen and Falk empirically test this conjecture for a sample of 72 charities in the health sector.The authors make a regression of various measures of (allocative and technical) efficiency on the board's proportion of insider members.Their results do however not support the conjecture. 14 Dyl, Frant and Stephenson (2000) also examine this conjecture by studying the relationship between board structure and various measures of financial organisational performance for 54 medical research charities.They find that only the representation of the executive director on the board has significant influence on the behaviour of the NPO in that it decreases expenditures on program activities and increases the fund-raising expenditures. 15Size of the board does not seem to have a significant effect on performance, unless the executive director is a board member.When this is the case, board size tempers the decreasing of the program expenditure and the increasing of the fund-raising expenditure.In general, their results support the Fama-Jensen-Williamson conjecture in that is the proportion of insiders which is an important characteristic of the board but fail to support the conjecture since the proportion of insiders does not seem to have a positive influence on the management expenses.Also the study of Brickley, Van Horn and Wedig examines the board membership of the CEO in relation to this CEO's compensation.
ZögU 30.Jg. 1/2007 They find that in non-profit hospitals were the CEO is a voting member of the board, his/her compensation is almost ten percent higher than in hospitals where the CEO does not have voting power in the board. 16So their findings are consistent with the Fama-Jensen-Williamson conjecture that non-profit boards should not have internal managers as voting members.The study of Oster and O'Regan also suggests that having the CEO as a board member pushes the non-profit board towards more fundraising and less monitoring activities.These authors also examine the relationship between the personal characteristics of the board members and performance of the board but do not seem to find statistical evidence of this relationship. 17allen, Klein and Tinkelman concentrate on the board membership of large donors in relation to organisational efficiency.For 123 American NPOs, they find a significant statistical association between the board membership of large donors and efficiency as measured by administrative expenses. 18nother study that focuses on the composition of the non-profit board and its relation to organisational performance is the study by Brown.He examines whether racial heterogeneity and participation of underrepresented social groups on the board has a positive influence on performance.His results only show a very modest, positive relationship between board heterogeneity and organisational performance for 125 human service NPOs. 19rovan examines the influence powerful boards of directors have on the ability to attract resources of 46 human services NPOs.Composition of the board is translated in the number of members that have power, i.e. have strong linkages with society.He finds that board power can indeed have a positive influence on performance of the organisation measured as the fundraising result. 20When the author however uses growth in fundraising income as the dependent variable, the significant result disappears.Siciliano on the other hand focuses on the relationship between occupational diversity of the board and organisational efficiency.Based on data for 240 YMCA organisations, she finds that board members having greater occupational diversity are associated with higher performance on the social and the fundraising level. 21ome of the studies thus seem to find a relationship between board structure and organisational performance, but there is no indication on the causality of the relation. 22Empirical work on the impact of non-profit board composition on organisational effectiveness is obviously still too fragmentary to draw straightforward conclusions. 23

The board's process characteristics
Here we review empirical research on the influence the board's processes and strategies have on organisational behaviour and performance.We interpret the term processes very broadly and include strategic planning, decision making but also board-staff relations.Herman and Tulipana argue that the quality of the relations between the board and the staff is crucial for a good internal functioning of the board.Hence, the authors use this quality to conceptualise the internal role of the board.The board-staff relationship is measured on the dimension of influence, and organisational performance is measured on the basis of member ratings.Their results indicate a weak positive relationship between board processes, conceptualised in the board-staff relationship, and organisational performance. 24radshaw, Murray and Wolpin examine the relationship between various practices of the board and board effectiveness as perceived by the chief executive officer in various NPOs.They find a positive relationship between the perception of board effectiveness and the use of many advocated board practices.However, when they use objective indicators of organisational performance (i.e.size of any incurred deficit and change in the annual budget) they do not find a significant association anymore. 25reen and Griesinger examine the relationship between various tasks and processes of the non-profit board and organisational performance.The authors assume that certain practices lead to board effectiveness and so they use behaviour-based measures of effectiveness.They find a significant positive relationship between performance of the board and effectiveness of the organisation with policy formation, strategic planning, program monitoring, financial planning and control, resource development, board development and dispute resolution as the board activities most strongly correlated with organisational effectiveness. 26Some of the papers thus seem to indicate a positive relation between performance of the board and of the organisation.But although some of the empirical papers discussed here seem to agree on a relation between board effectiveness and organisational effectiveness, there is no indication on the causality of the relation.It may well be the case that effective organisations are able to attract more effective board members than less effective organisations.As Herman and Renz note "Although evidence supporting the relationship between board effectiveness and organisational effectiveness is increasing, in what ways and how boards contribute to organisational effectiveness is still unclear". 27 See Herman and Tulipana (1985).25 See Bradshaw, Murray and Wolpin (1992a).26 See Green and Griesinger (1996).27 Herman and Renz (2000) p. 150.

IV. Non-profit board as the dependent variable
The central question in this section is What influences the (composition, role, performance of the) non-profit board?If we assume as in the previous section that the non-profit board is an important institution that influences non-profit behaviour, it is important to know what determines it and how.

Board's structural characteristics
A lot of the practitioner-oriented literature gives prescriptions for the optimal composition and structure of the non-profit board.The board should not have too many members, should represent all constituencies etc.Most of these prescriptions are a result of the personal experience of the author or are based on anecdotical evidence.It is obvious that the environment plays an important role in the determination of the board structure and composition, especially in NPOs.At this point however, empirical research is rather scarce.Pfeffer, based on data of 57 mostly non-profit hospitals, concludes that when hospitals depend more on local donations and local financial resources, they are more likely to have more members in their board that have strong linkages with the organisation's environment. 28 more recent study of Cornforth and Simpson, based on sample of 500 English and Welsh charities, focuses on the number of board members.Their results show a positive relationship between the size of the non-profit board and the size of the organisation, measured as the income of the NPO.29

The board's process characteristics
Here, focus is often on the power of the non-profit board, or more precisely, on the relationship between the board and its staff.Provan finds a positive relationship between the receipt of information and the ability to influence the decision-making process in the organisation.Hence, he studies the organisational characteristics that have an influence on the receipt of information by three of the most important governance participants in a hospital, i.e. the medical staff, the board and the chief executive officer.The principal conclusion from his research is that the number of beds, as measure of hospital size, is an important determinant of the receipt of information by the board and medical staff.Besides, Provan finds that the CEO receives the most information (and the medical staff the least) and also has the highest influence on decisions.Based on the distribution of power in the organisation, Murray, Bradshaw and Wolpin develop five patterns of board governance: a CEO-dominated board, a powerless board, a chair-dominated board, a power-sharing board and a fragmented power board.The authors develop these five patterns based on in-depth case studies and interviews and note that often the characteristics of the individual board members determined the pattern of governance. 31ornforth also studies the power relation between the non-profit board and its management.Based on in-depth case studies, Cornforth identifies various factors that influence the distribution of power between the non-profit board and management.The most important factors were differences in expertise and skill, board processes and procedure. 32ecently, more attention is given to the relationship between the board and management.The importance of a good board-staff relationship is acknowledged but still more research is needed in this direction.Some authors view the board as a 'rubber stamp' of management.Another view is the so-called 'micro-managing' board where it is too much involved with the day-to-day management of the NPO.McClusky argues that the appropriate roles of staff, board and other volunteers must shift according to seven variables: the size of the budget, board and staff, the number of volunteers, the stage of the NPO's life cycle, the level of trust between the board and the CEO, executive transition, the presence of organisational crisis and environmental factors. 33

V. Problems in empirical studies on governance
Hermalin and Weisbach note that empirical work on governance, and thus also on boards of directors, is complicated by two factors.We already briefly mentioned the first problem concerning the endogeneity of the crucial variables when we discussed the relation between board effectiveness and organisational effectiveness.Organisational performance can be a result of a well-functioning board, but it can also be the reason why the organisation attracts good, capable and dedicated board members.Often the empirical studies are based on cross-sectional data and fail to recognise this problem of endogeneity.Therefore one must interpret their results with caution.The second problem Hermalin and Weisbach discuss is the difference between equilibrium and out-of-equilibrium results.Maybe we can best illustrate this point with the Fama-Jensen-Williamson conjecture.These authors hypothesise that the proportion of insiders in the board will be negatively related with the efficiency of resource allocation.In the interpretation of an out-of-equilibrium result, this means that decreasing the number of insiders on the board will have a positive effect on allocative efficiency of the NPO.In contrast, in the equilibrium interpretation another factor lies on the basis of this relation-ZögU 30.Jg. 1/2007 ship between number of inside board members and allocative efficiency.So also here, caution is recommended. 34n the concrete case of NPOs, one can maybe note a third problem inherent to empirical studies with respect to governance and its relationship to performance.Performance of non-profit organisations happens to be very hard to measure since their objectives are often socially orientated and less easy to quantify. 35By consequence it is even harder to measure impact of governance on the performance.

VI. Conclusion and research issues
This overview of the literature reflects the growing interest of researchers in the nonprofit board.Although empirical results are not always straightforward and sometimes even contradictory, there exists large consensus that the non-profit board is one of the most, if not the most, important governing institutions of the NPO.Hence, the increased scholarly attention is justified, although it is primarily of empirical nature.This empirical research on non-profit boards seems to have developed on two roads: one where the non-profit board is treated as the dependent variable and one where it is treated as an independent variable.Both research tendencies have to deal with methodological problems obstructing direct interpretation of their results.The endogeneity of many variables in the relationship between non-profit boards and other variables can only be accounted for by using panel data instead of cross-sectional data, which is hardly available. 36The difference between an equilibrium and an out-of-equilibrium interpretation also influences the interpretation of the results.That the non-profit board is related to organisational behaviour and by consequence, organisational effectiveness, is without doubt.More research is however needed to identify the exact relationships and causality of these relations.