Where to microfinance?

Woller, Gary M. ; Dunford, Christopher ; Woodworth, Warner 等

Abstract

The microfinance industry is characterized by a "schism," or debate, between two camps that represent broadly different approaches to microfinance: the institutionists and the welfarists. How this debate is resolved has crucial implications for the future of microfinance--its guiding principles, its objectives, its clients, and its impact on the poor and poverty in general. The institutionist approach, with its emphasis on financial self-sufficiency and institutional scale, appears to have gained ascendancy over the welfarist approach, with its emphasis on direct poverty alleviation among the very poor. The institutionists, however, base their arguments on a number of debatable assertions and questionable empirical methodologies. This article critically examines some of these with the intent of placing institutionist claims in their proper perspective and tempering the hegemonic aspirations of some institutionist writers. It concludes by proposing a middle ground between the two approaches in the hope that it will lead to more productive dialogue between the two camps in the future.

Introduction

Like many popular grassroots movements, the microfinance movement is characterized both by widespread agreement on broad objectives and by multiple rifts on key issues. The movement itself is driven by the shared commitment to provide credit for small enterprise formation and growth. It is also bound together by a common rhetoric of concern for the poor. This unity of commitment and rhetoric, however, masks a bewildering variety of philosophical approaches, types of institutions and borrowers, and delivery systems that shelter uneasily together under the big tent called "microfinance."

The movement has come to be divided by two broad approaches, or opposing camps, regarding the best way to help the poor through access to financial services: the institutionist approach and the welfarist approach.(1) Jonathan Morduch (1998d) refers to this division as the microfinance schism. The irony is that while the worldviews of each camp are not inherently incompatible, and in fact there are numerous microfinance institutions (MFIs) that appear to embrace them both, there nonetheless exists a large rift between the two camps that makes communication between them difficult.

The institutionist approach focuses on creating financial institutions to serve clients who either are not served or are underserved by the formal financial system. Emphasis lies on achieving financial self-sufficiency; breadth of outreach (meaning numbers of clients) takes precedence over depth of outreach (meaning the levels of poverty reached); and positive client impacts are assumed. The center of attention is the institution, and institutional success is generally gauged by the institution's progress toward achieving financial self-sufficiency. The best-known examples of the institutionist approach are Bank Rakyat Indonesia (BRI) and Banco Solidario (BancoSol) in Bolivia.

Institutionists argue that a primary objective of microfinance is financial deepening, the creation of a separate system of "sustainable" financial intermediation for the poor. Theirs is a "financial systems" approach to microfinance, in which the future of microfinance is dominated by numerous large-scale, profit-seeking financial institution that provide high quality financial services to large numbers of poor clients. Because of their insistence on financial self-sufficiency, institutionists eschew subsidies of any kind. The institutionist position is articulated in virtually all the literature coming out of the Ohio State University Rural Finance Program, the World Bank and the Consultative Group to Assist the Poorest (CGAP) in the World Bank, and USAID. It is also found in the many writings of Maria Otero (ACCION International) and Elisabeth Rhyne (formerly of USAID). Most published literature in the field of microfinance espouses the institutionist view.

Welfarists, on the other hand, emphasize depth of outreach. Welfarists are quite explicit in their focus on immediately improving the well-being of participants. They are less interested in banking per se than in using financial services as a means to alleviate directly the worst effects of deep poverty among participants and communities, even if some of these services require subsidies. Their objective tends to be self-employment of the poorer of the economically active poor, especially women, whose control of modest increases of income and savings is assumed to empower them to improve the conditions of life for themselves and their children. The center of attention is the "family." Like the institutionists, welfarists have assumed more impact than they actually have been able to document. The most prominent examples of welfarist institutions are the Grameen Bank in Bangladesh and its replicates elsewhere, and FINCA-style village banking programs in Latin America and, more recently, in Africa and Asia.

Obviously, there are fundamental differences between the two camps. These differences, moreover, are much more than merely philosophical debates. How they are resolved has crucial implications for the future of microfinance--its guiding principles, its objectives, its clients, and its impact on the poor and on poverty in general. Our purpose in this paper is first, to trace through these implications; second, to evaluate critically the arguments, assumptions, and methodologies of the institutionist camp; and third, to offer our views on how the two approaches might be reconciled.

Before proceeding, however, we should make our position clear. We are welfarists. Moreover, we have grave concerns about the direction that the institutionists are attempting to push the industry. The institutionists clearly have won the debate to date. They have defined "best practices," and the most prominent donors and even the most prominent welfarist practitioners have embraced that definition. The institutionists have been more articulate and persuasive and certainly more prolific in their writing, and their message has been more in tune with the times, the currently dominant culture of laissezfaire business.

We should also add that we have tremendous sympathy for the institutionist position, and we share the institutionists' vision of financial deepening. But this is not the limit of our vision. We foresee an industry in which the two approaches work in tandem to reach different, but equally deserving, populations of poor clients. We do not eschew profits, but neither do we eschew "subsidies." Nor, finally, do we dispute the institutionists' principled commitment to poverty alleviation.

Our specific concern is that in advocating their position, some prominent institutionists have gone too far--to insist that all MFIs adopt institutionist values and "best practices," to attempt active suppression of the welfarist point of view, and to cause the expansion of "best practices" to become antithetical to the welfarist objective of direct poverty alleviation among the very poor. Thus we believe it is important that the many pronouncements emanating from the institutionist camp be rationally challenged in terms of logic and fact, which we attempt to do in this paper. Our purpose is not to invalidate the institutionist view, but rather to put it into perspective, temper its hegemonic aspirations, and argue for the vision in which both approaches can work simultaneously toward shared or disparate goals. Both approaches are needed--in whatever combination possible.

Two Nations Divided By A Common Language

In thinking about the rift between the institutionist and welfarist camps, we are reminded of the quote by George Bernard Shaw that Great Britain and the United States are "two nations divided by a common language." Although the two camps share a common commitment to microfinance services and a common rhetoric of concern for the poor, many in the industry mistake this unity for a unity of purpose.

The stated ultimate goal of both camps is poverty reduction. Yet the practical objectives each camp has set for itself diverge. Each has defined "poor" differently, and each has articulated different visions of how the poor can be helped by increased access to microfinance services. The practical implications of these differences between the two camps are at least threefold: (1) differences in the population segments served (the not-so-poor true entrepreneurs vs. those who struggle on the margins of survival); (2) differences in the designs (and the reasons for the designs) for service delivery to these populations (lending to individuals vs. small solidarity groups vs. large village banks);(2) and (3) differences in the institutional structures and financing to support these services (social service NGOs vs. community-based credit unions and community banks vs. commercial banks and finance companies).

These differences are legitimate, if and only if the objectives from which they derive are considered equally legitimate. But they are not considered equally legitimate by many persons in each camp. Heightening the potential for conflict is the apparent unity of purpose in the microfinance community, which has fostered a mentality of "one way" for microfinance. Donors have become confused by the veil of unity and the argument that a common set of standards is needed to advance the apparently common agenda. There has developed in the 1990s a struggle to define that "one way" for both microfinance practitioners and donors.

The Institutionist Attack. The conflict between the two camps is driven by the belief that the alternative approach threatens the fulfillment of the movement's broadly shared goal--poverty reduction. Institutionists believe that successful poverty reduction requires massive scale, given both the worldwide prevalence of poverty and the estimated demand for microfinance services. Rough estimates put the total demand for microcredit at $12.5 billion by 2005 and $90 billion by 2025 (or 10 and 30 percent respectively of the world's low-income entrepreneurs, CGAP, 1995b), or alternately at between 100 and 200 million persons (Christen et al., 1995). By comparison, the total microlending portfolio today is estimated at only $2.5 billion (GGAP, 1995b).

But massive scale in turn requires massive financial resources--far beyond the ability of donors to provide. Even if donors had sufficient resources, they are subject to fads, have their own agendas, and are not a reliable long-term source of funds. It is from this "recognition of scarcity" (Gonzalez-Vega, 1993: p. 25) that springs the institutionists' concern for financial self-sufficiency. Moreover, the only way to attract the requisite financial resources is by tapping into private sources of capital.(3) But widespread access to private capital in turn requires that MFIs be well run, operate efficiently, and, most important, be profitable. Finally, to satisfy the world demand for microfinance services, and thus make a dent in world poverty, it is not sufficient that a relatively small number of MFIs tap into private capital. It is necessary to create an entirely new financial system consisting of a large number of privately financed, large-scale financial intermediaries that provide financial services to the poor.

To guide the industry's transition to for-profit status, institutionists have spent much time in an attempt to design a set of "best practices" for industry adoption. Best practices refer to those practices that improve institutional efficiency and effectiveness in areas such as management and management systems, finance and accounting, marketing, service delivery, or product design and development. The identification, standardization, and widespread adoption of "best practices" are believed to be an essential step on the path to industry-wide financial self-sufficiency, capital market access, and maximum outreach to poor clients.

The perceived welfarist threat to the institutionist vision is that its relative uninterest in the issues of scale, "best practices," and financial self-sufficiency--and thus its continuing dependence on donor "subsidies"--threatens to undermine the march toward industry-wide financial self-sufficiency and all that entails. Implied herein is the belief that true concern for the poor requires the kind of scale of microfinance services that only the institutionist approach can deliver. This belief is captured by Christen et al. (1995: pp. viii, 8) who write that "it is scale, not exclusive focus, the determines whether significant outreach to the poorest will occur.... Programs that do not attempt to achieve large scale outreach are simply not making a dent in the global problem."

Institutionists have aggressively evangelized the microfinance community to have their views adopted, and they have enjoyed much success. That prominent institutionists have occupied at times key positions at the World Bank, CGAP, and USAID has greatly aided their cause, as has the impressive body of literature produced by institutionist writers. The influence of institutionist thought is clearly evident in that institutionist terms and concepts (e.g., sustainability, outreach, subsidy, subsidy dependence index, and best practices) have become the lingua franca of the microfinance industry.(4)

The evangelization of institutionist precepts also has taken place within the donor community. Donors are urged to "husband microfinance by creating an environment that rewards success [progress toward financial self-sufficiency] and punishes failure [unsatisfactory progress toward financial self-sufficiency]. To culture strains of [MFIs] that balance sustainability and outreach, donors must lubricate entry and exit" (Schreiner, 1997a: p. 1). In the institutionist worldview, the donor role "should essentially be to underwrite the commercialization of microenterprise finance" (Christen et al.: p. ix).

The Welfarists React. Welfarists distinguish themselves from institutionists primarily by their value-based commitment to serve the very poor. While they acknowledge the benefits and necessity of scale in attacking world poverty, their inclination is to place greater weight on depth of outreach than on breadth. They do not differentiate themselves by any lesser degree of commitment to sound operational and management practices or to institutional efficiency or effectiveness. But whereas they believe that increasing financial self-sufficiency is generally desirable, they are unwilling to take the next step--to accept that financial self-sufficiency is necessary to fulfill their institutional missions.

The perceived threat to welfarists posed by the institutionist approach is multifold. First is the belief that the commercialization of microfinance and the need to satisfy the "selfish" demands of outside investors will inexorably lead to profit motive displacing social mission. According to Renee Chao-Beroff (1997: p. 105):

 There is thus great risk of diverting the newly created professor
 of 'people's banker' or of the 'micro-financing for the poor' from
 its proper objective. The fact is that if priority is given to
 making [MFIs] profitable as quickly as possible, then the poorest
 will automatically be marginalized in favor of populations that are
 supposed to be more creditworthy. Similarly, the rural areas, in
 favor of urban areas, which are more densely populated and provide
 better commercial opportunities.

Second, in a philosophical sense the fear is that the commercialization of microfinance will divert the industry from its "spiritual foundation," which was and is the movement's animating force. The result is a profitable but soulless endeavor. According to Rodey (1997: p. 12), "Spiritual principles linked to sound financial principles must be a central tenant of the microfinance movement so that this noble effort to eradicate poverty does not become simply business as usual, with money at the bottom line. Again, the issue is not only whether we reach the numerical goal, but how that will determine the outcomes."

Summarizing the first and second concerns, the perceived threat is that if the industry embraces the institutionist position, it will have embarked on a potentially errant path that will have profound impact both on the industry itself and on those whom it serves. Thomas Dichter (1996: p. 259) captures the essence of these two concerns when he writes that the overarching emphasis on financial self-sufficiency

 has consequences ... for the soul of many NGOs (compassion vs.
 making a buck) both in terms of outreach to the very poor and in
 terms of impact and effect of recipients.... NGOs who shift into
 sustainable credit programs may be losing their real competitive
 advantage: the capacity to reach the very poorest and engage in a
 variety of activities that help people change, but which cannot
 necessarily be financially supported by recipients of
 assistance.... Financial self-sufficiency will bring in its train
 deep changes in the ways NGOs do work, not to mention who and what
 they are.

Third is the concern that the call for donors to withdraw support from "unsuccessful" programs amounts to the attempted suppression of dissident viewpoints, which, if heeded, will result in a "broad-brush resource allocation on the basis of good institutional performance alone" (Rogaly, 1996: p. 106), regardless of actual program impact. Fourth, the drive to define and codify "best practices" risks the imposition of a blueprint approach to microfinance that will stifle innovation and experimentation in the design of new products and delivery systems for the very poor. (For example, MFIs will hew strictly to "best practices" for fear of losing donor support.)

Genesis And Assumptions Of The Institutionist Approach

To understand the arguments of the institutionists, it is helpful to trace the genesis of their thought. Their position is a direct outgrowth of the experiences of Rural Development Institutions (RDIs) during the 1960s and 1970s and the lessons derived thereof by researchers at the Ohio State University Rural Finance. During the stated time period, development agencies and Third World governments funneled large sums of money to a numerous state-run RDIs that provided agricultural credits to poor farmers. As a general rule, the RDIs offered credit at highly subsidized terms to farmers, targeted credit to specific uses, and did not offer savings services.

From the beginning these RDIs were plagued by a number of problems, including a grant mentality among clients, high overhead and transaction costs, and heavy corruption. In terms of impact, these programs produced only a limited production response among farmers, tended to be co-opted by wealthy interests who valued the loans primarily for their subsidy value, and suffered from abysmal repayment rates. Not surprisingly, therefore, donor money eventually dried up, and the RDIs almost universally failed (see Adams, Graham, and Von Pischke, 1984).

According to institutionist literature, several lessons could be learned from the experience of the RDIs. [See Gonzalez-Vega (1993, 1994) for an in-depth discussion of the "lessons learned."] But above all, "the most severe deficiency of the traditional rural financial organizations ... has been their lack of institutional viability" (Gonzalez-Vega, 1993: pp. 23-34). In hindsight, donors proved both unable and unwilling to provide long-term funding to support the RDIs. This lack of institutional viability in turn led to even lower repayment rates because borrowers had little incentive to repay loans to institutions whose futures were in doubt.

The experiences of the old RDIs and the lessons learned thereof form the basis for the institutionists' approach to microfinance today. Most of their prescriptions should be understood within this light.

In the following sections, we offer some critical observations of the "lessons learned" that inform the institutionist worldview. Some of our criticisms are influenced by Jonathan Morduch's (1998d) earlier critique of institutionist thought.95) For the purpose of brevity, and to avoid treading too much on ground already trod by Morduch, we focus our criticisms on the following "lessons learned" and their subsidiary assertions:

* The experiences of the old RDIs have direct relevance to the microfinance industry today.

* Subsidized institutions inherently are inefficient, unable to innovate or implement new technologies, and unable to achieve significant scale.(6)

* Institutional sustainability requires financial self-sufficiency.

* Institutional "subsidies" are determined by the selfish opportunity cost of capital.

* Financial self-sufficiency is the measure of a "successful" MFI.

i. Financial self-sufficiency is achievable for many, if not most, MFIs.

ii. Profit-seeking MFIs can maintain a commitment to very poor clients while simultaneously earning the high returns on equity (ROE) demanded by their "selfish" investors.

iii. Financial self-sufficiency is a rational objective of many, if not most, MFIs.

iv. Financial self-sufficiency is a means and not an end.

Again, we do not offer the following critique in an attempt to dismiss institutionist views. Rather, we hope that consideration of the issues we raise will encourage a less doctrinaire promulgation of institutionist claims and thereby help to open the door to a greater spirit of accommodation in each camp.

The Relevance Of The Old Rdis To Microfinance Today

Just how relevant were the experiences of the old RDIs to the microfinance industry today? Very relevant, according to Adams and Von Pischke (1992), "both programs involve similar assumptions, both contain similar policies, both tussle with definitional issues, both use the same type of project justification, and, as a result, both are likely to encounter similar problems" (p. 1463). They thus conclude that such similarities portend similar results: "Many of the loans being made to microenterprises will not be repaid, most of these programs are likely to be transitory, and many of the targeted borrowers will be materially assisted in the long run through programs that increase their debt" (p. 1468).

The equating of the current microfinance industry with the old RDIs, however, ignores several fundamental differences between the two. Numerous technical innovations (e.g., group lending and village banks) have significantly reduced the information asymmetries and transaction costs of providing financial services to the poor relative to the earlier experience. In contrast to the old RDIs, the microfinance industry today emphasizes small-scale, short-term lending; the need to charge market or near-market rates of interest; the importance of mobilizing saving among the poor; the illusory nature of tying loans to specific activities; the value of convenience to poor borrowers; low overhead and simplified transaction processes; and social collateral to ensure repayment.

While undoubtedly many MFIs suffer from severe institutional or program deficiencies, the industry today includes a large number of well run NGOs and a handful of well run government programs, all with successful long-term track records of expansion, high-quality client service, and attracting financial support. In well run programs it is not unusual to find repayment rates of 95 percent or better.

Finally, many MFIs today takes seriously their obligation to produce client impact. Of the relatively small number of MFIs that have been evaluated for client impact, "the findings indicate positive impacts across different types of programs, contexts, socioeconomic groups, and gender of clients" at the enterprise, household, and individual levels (Sebstad and Chen, 1996: p. 20). To be sure, the industry has a long way to go in developing social impact measurement systems and in credibly documenting the impact of microfinance, but it is still light years ahead of where it was back in the "bad old days."

Thus on virtually every major count, the microfinance industry today is vastly different than the old RDIs of the 1960s and 1970s. While equating the two might make a handy strawman to buttress institutionist policy prescriptions, it is nonetheless inaccurate.

The Perils of "Subsidization"

Drawing on the experience of the old RDIs, institutionist writers have drawn a number of sweeping conclusions about the perils of "subsidization." In response, a number of things merit mention.

Virtually all the pathbreaking innovations in the microfinance industry have come from "subsidized" MFIs. Perhaps two of the most significant innovations in the industry--group lending and village banking--were developed by mission-driven MFIs (e.g., Grameen Bank, ACCION, and FI NCA), each heavily dependent on donor funding at the time of innovation. Donors have proven willing to date to invest in (or "subsidize") the experimentation and innovation responsible for shifting the industry's production possibility frontier to where it is today.(7)

Implied by the argument that "subsidized" MFIs are inherently inefficient (or less efficient than for-profit institutions) is that the absence of a profit motive fails to create the proper incentives for management. Morduch dispatches this argument by correctly pointing out "maintaining 'hard' budget constraints is the key [to efficiency], not maximizing profits" (p. 12). That and management commitment to running an efficient operation. Over the last several years, large numbers of donor-dependent MFIs have made tremendous strides in improving administrative and operational efficiencies, and the same MFIs are at the forefront of technological innovation today. These clearly are not the RDIs of yesteryear.

The institutionists' sweeping indictment of "subsidization" is so broad in scope that it must also stands as a serious indictment of the entire nonprofit industry (or at least the vast majority of nonprofits who rely on donor support). Moreover, it implicitly idealizes the for-profit industry. The argument that "subsidized" institutions are inefficient and cannot achieve significant scale ignores the large number of well run nonprofits, both inside and outside of the microfinance industry, that have achieved impressive scale (e.g., Grameen Bank, BRAC, FINCA, SEWA, CARE, Save the Children, Christian Children's Fund, Red Cross, United Way, March of Dimes, and Greenpeace). It may be the case that profit-seeking firms tend to surmount more easily the obstacles to scale (although this assertion remains unproven in microfinance), but this is different from the blanket assertion that "subsidization" precludes significant scale.

There exist thousands of well run nonprofit organizations that have thrived and grown, despite heavy reliance on donor funding. At the same time, one need not look far to find poorly run for-profit firms that suffer from serious managerial or operational inefficiencies. For-profit firms cease to exist, and they do so in large numbers, including even commercial banks. In fact, far more for-profit firms fail than succeed.

The point is not to argue that one form of organization is inherently superior to the other, but to point out the obvious fallacy of making such sweeping claims about either. In each sector there are well run and poorly run organizations, efficient and inefficient organizations, large-scale and small-scale organizations, innovative and static organizations, and sustainable and unsustainable organizations. The sweeping assertion that for-profit organizations are always inherently superior to nonprofit organizations in each of the above areas is more a reflection of one's personal bias than it is an objective assessment of each of the sectors.

The fact is that the old RDIs constituted one form of organization that existed at one point in time. One simply cannot extrapolate directly from them to existing MFIs or, by logical extension, to the entire nonprofit industry.

Institutional Sustainability Requires Financial Self-Sufficiency

It is true that donor funds are limited, and it is true that donors can be fickle, faddish, and unreliable. The argument that institutional sustainability requires financial self-sufficiency, however, obscures the context under which donors withdrew their support of the old RDIs. The old RDIs were highly inefficient and ineffective programs that suffered huge financial losses and had questionable or harmful impacts on their intended beneficiaries. It is thus hardly surprising that the donors eventually withdrew their support. It would have been surprising if donors had not withdrawn their support.

There is much semantic confusion surrounding the word "sustainable." In general terms, sustainability implies institutional permanence--it captures the idea that an institution is and will continue to be a "going concern." In line with this idea, Navajas et al. (1998: p. 5) define sustainability as "to reach goals in the short-term without harming your ability to reach goals in the long-term." Similarly, Edgcomb and Cawley (1994: p. 77) define sustainability as the ability of an organization to "sustain the flow of valued benefits and services to its members or clients over time." (Both sets of authors, however, later clarify their remarks to make clear that, in their view, the only way an MFI can become truly "sustainable" is to reach financial self-sufficiency. Edgcomb and Cawley (p. 86), for example, argue that "sustainable institutions can and must [emphasis ours] meet 100 percent autofinancing for their credit operations.")

We propose the following definition of sustainability offered by Brinkerhoff (1991: p. 22): "Sustainability can be defined as the ability of a program to produce outputs that are valued sufficiently by beneficiaries and other stakeholders that the program receives enough resources and inputs to continue production." This definition transforms the debate about sustainability, for it opens the very real possibility that an MFI could be viable in the long-term, despite dependence on donor funding.

This definition also requires that we recast the way we think about donors. It is odd to us that all economic actors are assumed to be rational, with the important exception of donors. In the institutionist literature, donors are portrayed as motivated almost solely by "irrational" impulses: donors are fickle, donors are faddish, and donors are unreliable. The possibility that there exist rational donors who seek to maximize social returns on social investments is rarely, if ever, allowed.

We argue that donors are as rational as any other economic actor is. It is this rationality that led them to abandon the old inefficient and ineffective RDIs. It is true that donors can at times be fickle, faddish, and unreliable (just like other economic actors). But it is by no means certain that rational donors (in particular, governments who "remain committed to poverty alleviation well after international agencies have moved on to the next Big Idea") will abandon microfinance "if subsidized microfinance proves to deliver more bank for buck than other social investments" (Morduch, p. 1998c: p. 44). Again, that so many MFIs and other nonprofits have survived and thrived for so long would appear to belie the rather sweeping assertion that institutional sustainability requires financial self-sufficiency. (Freedom from Hunger, for example, has operated on "subsidies" for fifty-two years now, far longer than the average life span of a for-profit business.)

Institutional "Subsidies And The Opportunity Cost Of Capital

The term "subsidy" is used in the institutionist literature to describe any financial resource received by an MFI at below market prices, which includes all types of donations. We could just as well talk about donations instead of subsidies, but the fact that the two carry different connotations has important implications for the tenor of the debate. The term "subsidy" is a loaded word that carries highly negative connotations. As used, the term implies that any resource received at below market cost is somehow tainted. Thus, like substituting the word debt for credit (another semantic trick in institutionist literature), its effect is to shock the reader or to play into preexisting biases.

We propose an alternative definition of "subsidy." Our definition requires a distinction between a "social" investor and a "selfish" investor.(8) There are two kinds of social investors. The first seeks solely a social rate of return in the form of, for example, higher incomes for the poor, better nutrition, clean water, or lower infant mortality rates. (Most donors fall into this category.) The second seeks both a social and a financial return (e.g., capital gains, interest, and dividends). This investor is willing to accept a "below market" financial return in exchange for a compensatory amount of social return. A selfish investor, on the other hand, seeks solely a financial return. The investor may not be interested in the social mission of the institution, but any interest in the social mission is subordinated to the selfish motives behind the investment.

For the first kind of social investor, a subsidy is an investment in an MFI at an expected social return below the social opportunity cost of capital, which is the expected return from foregone social investments. For the second kind of social investor, a subsidy is an investment at an expected return below the combined opportunity cost of capital, which is the expected return from foregone social and selfish investments. For the selfish investor, a subsidy is an investment at an expected return below the selfish opportunity cost of capital, which is the expected return from foregone selfish investments.(9)

Using the above definition of subsidy, a donor-funded MFI that has achieved significant outreach and impact such that its social benefits exceed those of alternative social investments is not considered subsidized. On the other hand, a donor-funded MFI that has poor outreach and poor impact such that its social benefits are less than those of alternative social investments is considered subsidized. In the first case, rational donors can be expected to continue to support the MFI. In the second case, rational donors can be expected to withdraw their support.

At the same time, a for-profit MFI that yields a below-market ROE for similar risk-adjusted investments is considered subsidized. The exception is the case in which private investors seek social returns in addition to selfish returns, but then by definition these are social investors and not selfish investors. In this case the MFI is not considered subsidized if its social return compensates the investor for forgone selfish returns.

Financial Self-Sufficiency As A Measure Of "Successful" Mfis

Two core assumptions of the institutionist camp are (1) financial self-sufficiency is achievable for many, if not most, MFIs, and (2) profit-seeking MFIs can maintain a commitment to very poor clients while simultaneously earning the market ROE demanded by "selfish" investors. The validity of these two assumptions is key to the institutionist position. If both are true, then the institutionist vision for microfinance would appear compelling. But if either is false, then the institutionist position collapses. Given the stakes, the industry can reasonably demand fairly compelling evidence before embarking down this path. Instead, the institutionist arguments are almost uniformly anecdotal and/or based on sample sets that typically are both small and biased.(10)

The study cited most frequently by institutionist writers as "proof" of the above assumptions is the Christen et al. (1995) study of "successful" MFIs. The eleven MFIs examined were not selected at random, but according to three criteria: breadth of outreach (number of borrowers), depth of outreach (average loan size), and reputation for financial strength. In other words, the MFIs examined in the study were selected because they were big, financially or operationally self-sufficient, and had very poor clients.(11)

After examining the eleven MFIs, the authors reached the following conclusion: "These results show no evidence of a direct tradeoff between outreach, either deep or extensive, and financial viability. The two goals are clearly not in opposition" (p. 27). That is one possible interpretation. Of course, an alternative interpretation of the findings is that the authors hewed closely to their selection criteria.

The authors reach another problematic conclusion from the data set. They write that "among high-performing programs, no clear tradeoff exists between reaching the very poor and reaching large number of people. In fact, mixed programs that serve a range of clients, such as BancoSol and BRI, have successfully reached very poor clients.... In short, is it scale, not exclusive focus, that determines whether significant outreach to the poorest will occur" (p. viii). In essence, what the authors argue is this:

a. BancoSol and BRI have achieved significant scale (a = b);

b. BancoSol and BRI reach very poor clients (a = c); therefore

c. Significant scale is necessary to reach very poor clients (b = c).

The reader will note that this is an example of nonsequitur reasoning. Regardless of the findings, however, the data set itself is totally insufficient to draw any meaningful inference about the industry as a whole.

To their credit, the authors add the caveat that "because three of the five fully self-sufficient institutions are in Indonesia, this assessment cannot state conclusively that full profitability is routinely possible" (p. 27). Unfortunately, this caveat does not stop others from doing precisely that. CGAP Focus Note 2, for example, asserts that the Christen et al. study demonstrates conclusively that "The conventional wisdom is quite wrong. Micro-finance institutions can [emphasis ours] and indeed need to be self-sustaining if they are to achieve their outreach potential providing rapid growth in access to financial services by poor people" (1995a: p. 1).

Is Financial Self-Sufficiency Generally Achievable? The conclusion that MFIs can be financially self-sufficient is an artifact of the sample set chosen. Different sample sets can yield very different conclusions. For example, separate studies of nine Western African MFIs (Webster, 1995) and five South Asian MFIs (Bennett et al., 1996) with reputations of excellence found that most had achieved significant depth of outreach, but that revenues covered only a relatively small percentage of operating expenses (only 30 to 40 percent for the African MFIs). For the South Asian MFIs, the authors conclude that financial self-sufficiency is a very difficult proposition for MFIs working in harsh socio-economic conditions and geographically isolated communities.

More generally, the CGAP Secretariat reports that the "vast bulk" of MFIs "do not see the potential for their specific institution to become financially viable in the foreseeable future, and expect to continue their dependence on donor funds for their operations and survival" (Malhotra, 1997: p. 8).(12) This is a decidedly less optimistic conclusion than the one cited earlier in the CGAP Focus Note. We suspect that a randomly drawn and representative sample of MFIs likely would portray a vastly different picture of the microfinance industry and financial self-sufficiency than that of the relatively small handfull of MFIs touted by institutionist writers.

Financial Self-Sufficiency and Institutional Commitment to the Very Poor

Even if we accept that financial self-sufficiency is generally achievable, what will keep profit-seeking MFIs from straying too far from their mission to serve the very poor? According to Maria Otero (1994), this protection will come in the form of board members who are to ensure that "maximizing returns does not overtake the priority objective of reaching the poor" (p. 98). In other words, "who invests in these institutions and what values they bring as shareholders will either safeguard or compromise the social commitment of the institution" (Otero, 1994: p. 102). But what confidence can we have that boards of directors will routinely safeguard the social commitment of MFIs? We suggest that, in answering this question, the industry consider the following caveats.

A key motivation for transforming MFIs into for-profit financial institutions is because "a financially self-sufficient [MFI] could attract capital from selfish private investors" (Schreiner, 1997b: p. 2). But as we have seen, selfish investors seek a financial rate of return equal to or greater than the risk-adjusted expected return of alternative selfish investment opportunities. Their primary interest in the social commitment of the organization is whether and the extent to which it increases or reduces their ROE. Furthermore, in profit seeking, publicly held institutions, maximizing returns is the priority objective. The social mission of the institution is inevitably a subordinate, albeit important, objective. To make the social mission equal or superior to maximizing returns implies a willingness to trade off selfish returns for social returns, which, according institutionist reasoning, is tantamount to subsidy.

In for-profit institutions, the board's fiduciary duty is to represent the interests of the owners--not those of the clients. When profit and social mission come into conflict (as they inevitably will at times), the board is bound to give greater weight to the interests of the owners. This is not to say that the two interests always will conflict or that the board necessarily must dismiss the interests of the clients if a conflict occurs. However, if the board consistently sides with the clients over owners, it will have failed in its duty as a representative of the owners, and it will have created a situation in which selfish investors involuntarily "subsidize" the social mission of the institution.

A way to avoid this conflict is to ensure that ownership, or a significant portion thereof, remain in the hands of social investors who are willing to trade off selfish returns for social returns. According to institutionist logic, however, this solution is not acceptable, for two reasons. First, this constitutes social investment, but social investment is tantamount to subsidy, and subsidy is not acceptable. Second, as institutionists frequently point out, the world supply of social investment is insufficient to meet the world demand for microfinance services. To satisfy world demand, MFIs must attract large amounts of selfish investment, which in turn creates a de facto change both in institutional mission and in the nature of institutional accountability to investors. It implies, moreover, that policies to increase social returns but that diminish selfish returns constitute a subsidy, and subsidy is not acceptable.

Thus we see that by their rigid definition of subsidy and by their opposition to subsidy in principle, institutionists have boxed themselves into a rhetorical corner. Either the MFI is fully financed by selfish investors at a market ROE, or the MFI is subsidized, and subsidy is not acceptable.

Two avenues of escape, however, lie open. One, the standard institutionist position (e.g., Christen et al.), is to argue that there are no real tradeoffs between the selfish mission and the social mission of profit-seeking MFIs. They are, however, far from proving their point (see below), and they bear a considerable burden of proof. (Most practitioners remain skeptical on this point.) Another is to relax the rigid definition of subsidy and the objection to subsidy in principle and accept the legitimacy of social investment.

Some additional insight on this question can be gleaned from the institutionists' two flagship MFIs, BRI and BancoSol. If we use average loan size as a proxy for depth of outreach, neither institution appears to have achieved significant depth of outreach. The average loan size at both BancoSol and BRI is over $500 (Gonzalez-Vega, 1997; Seibel and Parhusip, 1998), which far exceeds the average loan size of around $100 for MFIs that focus more sharply on poverty alleviation (Morduch, 1998c). Indeed, the study of five Bolivian MFIs by Navajas et al. (1998) found that around 97 percent of BancoSol's borrowers were among the marginally poor (those slightly above or below the poverty line) or among the not-so-poor.

Commercial banks that have entered microfinance so far have fared even worse in terms of depth of outreach. An examination of seventeen commercial banks offering microfinance services found loans sizes ranging from $500 to several thousand dollars (Bayadas et al., 1997) with an average loan size of "not more than $1400." This is hardly an auspicious beginning. In sum, "Few of the programs that cover all costs have proven able to reach the core of poor households. The typical borrower from financially self-sufficient programs ... tend to be among the 'better off' of the poor or are even slightly above the poverty line" (Morduch, 1998c: p. 5).

Institutionist writers counter this criticism by arguing that it is the absolute number of very poor reached who matter, not the relative proportion of very poor clients. According to Navajas et al. (p. 26), "Just because the proportion of clients among the poorest of the poor [is lower] ... does not mean that the [MFIs] served few households in this class. An estimate of breadth of outreach is the absolute number of poorest households reached." The argument is that a large-scale MFI with significant breadth but proportionately little depth of outreach will still reach more very poor clients than a small-scale, poverty-focused MFI. (This is one of the key findings of the Christen et al. study. Note also the a priori assumption that poverty-focused MFIs are necessarily small-scale.)

Even if we grant this argument, BancoSol still has not achieved significant depth of outreach. Of its 30,000 borrowers, only 3 percent, or 900, were among the very poor. For that matter, none of the five financially self-sufficient Bolivian MFIs examined by Navajas et al. achieved significant depth of outreach as measured by absolute numbers of very poor borrowers. In total, the five MFIs reached only around 2600 very poor borrowers.

Navajas et al. rationalize their findings three ways. First, they argue that few very poor borrowers reached will have longer-term access to financial services because "financially sustainable" MFIs will be around for the long-term, while poverty-focused, but financially unsustainable, MFIs will not. (We have already dealt with this argument.) Second, that the MFIs failed to reach significant numbers of the very poor "does not necessarily mean that they failed. [They] have other goals besides depth of outreach. For example, all five keep an eye on their profits" (p. 27). Third, they conclude that "the poorest of the poor may not be creditworthy ... This means that donors and governments, if they have the welfare of the poor in mind, may need to step back and to think about whether public funds meant to help the poor could be spent in a better way. After all, microcredit may not always be the best way to lift the poor out of poverty" (pp. 27, 37).

What can we conclude from the available evidence? It is still too early in the evolution of the industry to state definitively whether financial self-sufficiency is achievable for most MFIs or whether profit-seeking MFIs can achieve significant depth of outreach, although to date the sum of the evidence is not favorable on either count. Time and further investigation will clarify this.

Certainly, however, there is insufficient evidence to support the conclusion that "the poorest of the poor may not be creditworthy." This is a rather sweeping inference based on the experience of five Bolivian MFIs for an industry of over one thousand institutions, about most of who we know little, if anything. Moreover, it begs the following question: "How easily and based on what standard of evidence is the industry prepared to abandon the movement's animating vision of poverty alleviation among the very poor?"

By way of final comment on this topic, Elisabeth Rhyne asserts that MFIs that focus on the very poor "bear the burden of proving that they are as efficient and low cost in operations as technically possible. If not, subsidies support inefficient operations, and concern for the poor, however earnest, can become an excuse to avoid making difficult improvement" (1998: p. 6). Fair enough--we have no problem with this assertion. But turnabout is fair play. We would thus assert in response that MFIs that focus on financial self-sufficiency bear the burden of proving that they are truly reaching the very poor. If not, then they are pushing the microfinance industry to abandon its value-based roots, and concern for the poor, however earnest, can be become simply an excuse to make a buck.

The Rationality of Financial Self-Sufficiency. The institutionist approach takes a financial systems view--that is, it examines important issues in microfinance from the perspective of building a poor-persons' financial system. Thus it tends to extrapolate from the "system" to the individual MFI. In other words, what is good for the system is good for the individual MFI. This reasoning is known as the "fallacy of composition:" what is good for the one is good for the whole, and vice versa.

It may be that the objective of an individual MFI is scale and financial self-sufficiency, but then again it may not. It is entirely possible that an MFI has priority objectives, such as reaching a particular segment of the poor, that do not require full financial self-sufficiency. For a number of reasons, it may also not be in an individual MFI's interests to become fully financially self-sufficient. (For example, full financial self-sufficiency might be seen inconsistent with the MFI's priority objectives.) There is nothing inherently wrong with being small or with using donor funds to extend financial services to the poor, nor does either of these imply that an MFI is unworthy of donor support, particularly if its clients belong to a hard-to-reach population.

To insist that donors withhold or withdraw support from "unsuccessful" MFIs is in many cases tantamount, we suspect, to trying to compel them to behave in an otherwise economically irrational and potentially counter-productive manner. Economists refer to this as introducing "distortions" into the marketplace. What matters is "how subsidies are used" (Bennett, 1996: p. 287). In other words what matters is that the MFI produce improved social welfare. Quoting Jonathan Morduch (1998b: p. 5), "as long as [an MFI] delivers ample social benefits to its clients and can continue to receive sponsorship, [its] subsidies should be judged by their costs and benefits."

This line of argument raises a related issue that is at the heart of the institutionist versus welfarist debate: the need to perform impact assessments of microfinance programs. If, as institutionists claim, profitability is sufficient to demonstrate social impact, then impact assessments are an unnecessary redundancy, and MFIs should "concentrate on evaluating the quality of services and their institutional setting" (Rhyne, 1994: p. 107), which translates usually to the narrow measurement of progress toward financial self-sufficiency.

If, on the other hand, we assume that, as evidence now suggests, the "vast bulk" of MFIs will depend on "subsidies" to one extent or another indefinitely, then the need to document the impact of microfinance moves to the top of the agenda. (Also implied is the need to identify, target, and reach the core poor households. Something, quite frankly, welfarist institutions have not done well enough.) This is particularly true when we consider, as Elisabeth Ryhne (1998: p. 8) points out, that "important voices" outside of microfinance argue "that the very poorest people are not reached by even the most poverty-oriented microcredit programs, and that credit is not an appropriate service for people on the margins of survival." But it also implies, as Jonathan Morduch notes, and as we have implied above, that the industry "take public economics more seriously" and acknowledge that "even when poverty-focused programs do not meet all of their expenses, the benefits of ongoing subsidization may exceed their costs" (Morduch, 1998c: p. 6).

Financial Self-Sufficiency: A Means to an End? Charles Goodhart, a former official at the Bank of England, is given credit for the maxim, also referred to as Goodhart's Law, "If an economic statistic becomes the focus of attention, that statistic is likely to distort." We argue that there is reason to believe that Goodhart's Law applies to microfinance. In particular, we would argue the following. If MFIs and donors give the symbol concept of financial self-sufficiency too great a focus, then a force for change is created. That is, if the symbol becomes all-important, the thought behind it becomes lost, and it is transformed into an end unto itself. The movement to the all-dominating concept that financial self-sufficiency is synonymous with "success" is subtle, and not all involved will agree that it has occurred. Nonetheless, if an MFI finds, for whatever reason, that financial self-sufficiency has become a symbol of "success" (particularly among donors or investors), then the approach to managing the institution will change. (13)

Institutionist writers are quick to argue that such concerns are both ill founded and nonproductive. According to Elisabeth Rhyne (1998: p. 7), for example, "Sustainability is but a means to achieve [outreach].... Sustainability is in no way an end in itself; it is only valued for what it brings to the clients of microfinance. This is a point on which the 'poverty' camp frequently misstates the motives of the 'sustainability' camp. It would do wonders for the state of the debate if the poverty camp more readily acknowledged that the sustainability camp values sustainability only as a tool." While we do not doubt the sincerity of Rhyne's avowal, it is contradicted both in the writings of leading institutionist writers and in the internal logic of their arguments.

According to Navajas et al. (1998), the end of microfinance is "improved social welfare." This implies, then, that the ultimate measure of a "successful" MFI is whether it improves social welfare. The problem with improved social welfare, however, is that it is notoriously difficult and costly to measure. Consequently, some institutionist writers substitute outreach as a proxy for social welfare. This helps some. Breadth of outreach is easy to measure--simply count clients--but other dimensions of outreach, particularly depth of outreach, are more difficult to measure.(14) The typical proxy for an MFI's depth of outreach is average loan size. But this measure is both crude and flawed (e.g., it does not account for the variability in loan size or for median loan size, both of which are superior measures).(15) Rather than address these difficulties, institutionists take yet one more shortcut to estimate social welfare. Their proxy is financial self-sufficiency.

To measure the impact of microfinance on social welfare, one must calculate both social costs and benefits. Measuring the social costs of microlending is easy enough. This is equal to the selfish opportunity cost of capital. The difficulty comes in measuring the social benefits of microfinance. Fortunately for institutionists, microeconomic theory offers what seems to be an easy way around what would otherwise be a daunting measurement problem.

Rational consumers will not purchase a good or service unless they expect a net economic gain as a result (or are at least no worse off than before). If rational consumers pay the full economic cost of microfinance services, then by definition the private economic benefit of microfinance services (the benefit to the client) exceeds the private economic cost (the selfish opportunity cost of capital). Furthermore, if the MFI earns a profit, this implies that the sum of private benefits exceeds the sum of private costs. Absent significant negative externalities, this means that total social benefits exceed total social costs. To sum up, "Profits are necessary for sustainability, and sustainability is sufficient for worthwhileness" (Schreiner, 1997a: p. 5).

Tracing through the logic of this argument yields the following:

a. Financial self-sufficiency equals improved social welfare (a = b);

b. Improved social welfare is the end of microfinance (b = c); therefore

c. Financial self-sufficiency is the end of microfinance (a = c).

A similar conclusion can be reached by observing what criteria institutionist writers use to define "successful" MFIs. A few quotes should make this clear. "Two objectives are paramount for a rural financial institution to be successful: financial self-sufficiency and substantial outreach to the target rural population" (Yaron, 1994: p. 49). "The few [MFIs] that have been judged as successful have achieved that status because the [subsidy dependence index] showed them as either almost financially self-sufficient or just barely self sufficient" (Schreiner, 1997b: p. 4).(12) "The criteria for evaluating the success of such efforts [microfinance in Sub-Saharan Africa] should be on whether the institution achieves financial sustainability (Trape and Benhamou: p. 21). "We adopt the criteria suggested by Yaron to judge success ... self-sustainability" (Chaves and Gonzalez-Vega, 1996: p. 66.). "The new standards of judgement for the performance of [MFIs] have been described in terms of sustainability and outreach" (Navajas et al., 1998: p. 5). Conspicuously absent from the stated criteria in each cited example is "improved social welfare." Instead, institutionist writers assume that financial self-sufficiency and improved social welfare are one and the same.

There is one final logical flaw with the institutionist position. Again it stems from the argument that improved social welfare is the true end of microfinance. It is this: If an MFI produces improved social welfare, it is logically irrelevant whether the MFI is financially self-sufficient.

A possible rebuttal to this last argument is that while donor-dependent MFIs might improve social welfare, large-scale, financially sustainable MFIs, owing to their greater breadth and depth of outreach and their long-term permanence, can improve social welfare more. (Again, the a priori assumptions that only financially self-sufficient MFIs are sustainable and can achieve significant scale). Suffice it to say that this argument is based on a number of dubious assertions and the questionable "findings" of a small handful of "successful" MFIs. It also ignores important counterexamples of subsidy-dependent but significant outreach MFIs (both breadth and depth) that have been going concerns for a while now and show no signs of collapsing (e.g., Grameen Bank, BRAC, BKK, Working Women's Forum, PRADHAN, SEWA, CRECER, ASA, CARD Bank, and PPPCR).

In conclusion, we would urge the industry to consider where this overarching emphasis on financial self-sufficiency is leading. What is most important? Is it to build social enterprises that can last long enough to bring about major improvement in the lives of very large numbers of people? Or is to become certified as totally subsidy-free. We do not pretend to speak for all practitioners, but for many MFIs, the goal is not to become totally subsidy-free. That is neither necessary nor sufficient to achieve their priority objectives.

The reality is that social investment is available. There is a market for social investment for traditional social services. It is called philanthropy or charity. NGOs have more or less thrived on this market for decades now. And now there is a developing social investment market for MFIs--for start-up capital, for technical assistance, and for loans at concessional rates. Should not MFIs tap that market for one-time or occasional infusions of social investment? Social entrepreneurs should lose their business license if they did not!

Concluding Remarks

Institutionist writers portray a dichotomous view of microfinance. In this view, MFIs are both financially self-sufficient and large, or "the alternative to viable organizations are expensive, inviable quasi-fiscal programs that reach only a select few beneficiaries" (Gonzalez-Vega, 1994: pp. 16-17). Many welfarists also fall into the dichotomy trap, in which they envision a single "correct" approach for microfinance. As such, the discussion has gone the way of too many other discussions in development--it has polarized, and it has produced a fruitless debate about who is more truly concerned for the welfare of the poor.

Rather than continue with this nonproductive dichotomous view of microfinance, it would be more helpful to characterize the diversity of microfinance practitioners as lying somewhere along a continuum from traditional business (a purely financial bottom line) at one end to traditional social service (a purely social bottom line) at the other end. In the middle is the emerging phenomenon of the "social enterprise," which manages toward a double bottom line in a search to achieve a productive balance between selfish and social returns. The emergence of social enterprise can be seen in many sectors, but it may be best developed in the microfinance world.

Many of the models of the institutionist approach, like BancoSol, operate as traditional businesses. Any social objectives they may have are assumed by-products of their financial and institutional objectives. They do not measure success by social impact or by depth of outreach. There is nothing wrong with this approach, as long as practitioners are upfront about their objectives, and they do not try to attract social investors who explicitly want to pursue social objectives. These traditional business MFIs appear to address significant market failure to serve the borrowing needs of marginally poor and not-so-poor. Only incidentally, however, do they serve the very poor. Moreover, serving the very poor is not their "priority objective," and it would be helpful to donors and practitioners alike for those to whom this applies to say so.

Likewise, it would be helpful for traditional social service providers to admit that sustainable institution building is not their objective. They and their donors would do well to acknowledge that fact by making plans for leaving a legacy to be proud of when their microfinance projects phase out, sooner or later. They can provide loans at or below market rates to the poor needing special consideration (e.g., refugees and disaster victims) and still do a good job of loan recovery and managing their costs. Again, there is nothing wrong with this approach, as long as social service MFIs develop good strategies for eventually handing off their clients to more sustainable service institutions.

In fact, traditional social service providers can at times serve certain market niches better than sustainability-oriented MFIs. They can do a great deal of good during the "life of the project," provided they do not compete for clients who can better benefit from long-term microfinance services, put the meager assets of the poor at risk, or use their social mission as an excuse to operate inefficient and low impact programs.

Traditional business and traditional social service approaches are familiar polar opposites, the two ends of the microfinance spectrum. What is new and interesting in the microfinance movement is the broad middle ground occupied by the emergent social enterprises specializing in microfinance and related services. This is where the debate over "best practices" for combined impact and sustainability is most productively focused. The debate will improve as the different objectives are articulated and regarded as legitimate by all involved in the debate.

Social enterprises have to be explicit in both their social and financial-institutional objectives. Through appropriate staff incentives for managers and service staff, they need to commit to managing and measuring progress toward both. To date, social enterprises in microfinance have had serious difficulties defining, targeting, and reaching the core poor households, and they have done a very poor job of developing social impact measurement systems, much less actually measuring social impact. All are hard to do, but they have to be done, and MFIs better get started on it in earnest if they are to remain credible as social enterprises. Donors also need to clarify their own objectives and make sure these match up with the objectives of the traditional businesses, social enterprises, and traditional social services in which they invest.

If we had to guess, it would be that the future of microfinance will be characterized by a relatively small number of traditional business MFIs with significant breadth of outreach but limited depth of outreach and a relatively large number of social enterprise MFIs of widely varying sizes, institutional designs, and levels of financial self-sufficiency offering a wide variety of products and services targeted to the more poor. There is no need to make a once-and-for-all choice between competing approaches--a variety of approaches are needed, now and in the future. We would thus agree with Nitin Bhatt (1999: p. 15), who writes that "no one model of microfinance can solve the diverse developmental needs of the poor throughout the world. There is room for different kinds of programs, both subsidized and nonsubsidized, that cater to various segments of low-income communities. Given the need for a diversity of microfinance institutions, institutional plurality is key to prudent microfinance policy."

Finally, for everyone involved in microfinance today, we must know ourselves and be true to ourselves. We need to be more open and honest with each other about our real objectives and our commitment to reach them.

Acknowledgement

We would like to thank Jonathan Morduch and Didier Thys for their insight and helpful suggestions in writing this article.

References

Adams, Dale W. and J. D. Von Pischke. 1992. "Microenterprise Credit Programs: Deja vu." World Development, vol 20(10): pp. 1463-1470.

Adams, Dale W., Douglas H. Graham, and J. D. Von Pischke (eds.). 1984. Undermining Rural Development with Cheap Credit. Boulder: Westview Press.

Barge, Jim. 1985. "Goodhart's Law Strikes Again." The Banker. 135(8): pp. 27-31.

Bayadas, Mayada M., Douglas H. Graham, and Liza Valenzuela. 1997. Commercial Banks in Microfinance: New Actors in the Microfinance World. Occasional Paper No. 2372, Rural Finance Program, Ohio State University.

Bennett, Lynn, Mike Goldberg, and Pamela Hunte. 1996. "Ownership and Sustainability Lessons on Group-Based Financial Services from South Asia." Journal of International Development. 8(2): pp. 271-288.

Bhatt, Nitin. 1999. "Delivering Microfinance in Developing Countries: Controversies and Policy Perspectives." Policy Studies Journal (Forthcoming).

Brinkerhoff, Derick W. 1991. Improving Development Program Performance: Guidelines for Managers. Boulder: Lynne Rienner Publishers.

Chao-Beroff, Renee. 1997. "Developing Financial Services in Disadvantaged Regions: Self-managed Villages Savings and Loan Associations in the Dogon Region of Mali." Microfinance for the Poor? Ed. Hartmut Schneider. Paris: OECD, pp. 87-108.

Chaves, Rodrigo A. and Claudio Gonzalez-Vega. 1996. "The Design of Successful Rural Financial Intermediaries: Evidence from Indonesia." World Development. 24(1): pp. 65-78.

Christen, Robert Peck, Elisabeth Rhyne, Robert Vogel, and Cressida McKean. 1995. "Maximizing the Outreach of Microenterprise Finance: An Analysis of Successful Microfinance Programs." USAID Program and Operations Assessment Report No. 10. Washington, D.C.: U.S. Agency for International Development.

Consultative Group to Assist the Poorest. 1995a. Focus Note No. 2. Washington, D.C.: The World Bank.

Consultative Group to Assist the Poorest. 1995b. Focus Note No. 3. Washington, D.C.: The World Bank.

Dichter, Thomas W. 1996. "Questioning the Future of NGOs in Microfinance." Journal of International Development." 8(2): pp. 259-269.

Edgcomb, Elaine L. and James Cawley. 1994. "The Process of Institutional Development: Assisting Small Enterprise Institutions Become More Effective." In The New World of Microenterprise Finance: Building Healthy Institutions for the Poor. Eds. Maria Otero and Elisabeth Rhyne. West Hartford, CT: Kumarian Press, pp. 76-93.

Gaile, Gary L. and Jennifer Foster. 1996. "Review of Methodological Approaches to the Study of the Impact of Microenterprise Credit Programs." Assessing the Impact of Microenterprise Services. Washington, D.C.: Management Systems International.

Gonzalez-Vega, Claudio. 1993. "From Policies to Technologies, to Organizations: The Evolution of the Ohio State University Vision of Rural Financial Markets." Paper presented at the conference Finance 2000, Financial Markets and Institutions in Developing Countries: Reassessing Perspectives, May 28-27, Washington, D.C.

Gonzalez-Vega, Claudio. 1994. "Stages in Evolution of Thought on Rural Finance: A Vision from The Ohio State University." Occasional Paper No. 2134, Rural Finance Program, Ohio State University.

Gonzalez-Vega, Claudio. 1997. "The Challenge of Growth for Microfinance Organisations: The Case of Banco Solidario." In Microfinance for the Poor? Ed. Hartmut Schneider. Paris: OECD, pp. 129-161.

Gonzalez-Vega, Claudio. 1998. "Microfinance: Broader Achievements and New Challenges." Occasional Paper No. 2518, Rural Finance Program, Ohio State University.

Hulme, David. (1997). "Impact Assessment Methodologies for Microfinance: A Review." Assessing the Impact of Microenterprise Services. Washington, D.C.: Management Systems International.

Hulme, David and Paul Mosley. 1996. Finance Against Poverty, vol. 1. London: Routledge.

Malhotra, Mohini. 1997. "Micro-finance: The New Emerging Market?" CGAP Newsletter No. 3, pp. 7-8. Washington, D.C.: The World Bank.

Morduch, Jonathan. 1998a. "Does Microfinance Really Help the Poor?: New Evidence from Flagship Programs in Bangladesh." Unpublished manuscript. Cambridge: Department of Economics and HIID, Harvard University.

Morduch, Jonathan. 1998b. "The Grameen Bank: A Financial Reckoning." Unpublished manuscript.

Morduch, Jonathan. 1998c. "The Microfinance Promise." Unpublished manuscript.

Morduch, Jonathan. 1998d. "The Microfinance Schism." Development Discussion Paper No. 626. Cambridge: Department of Economics and HIID, Harvard University.

Navajas, Sergio, Mark Schreiner, Richard L. Meyer, Claudio Gonzalez-Vega, and Jorge Rodriguez-Meza. 1998. "Microcredit and the Poorest of the Poor: Theory and Evidence from Bolivia." Paper presented at the Latin American and Caribbean Economic Association Conference, October 22-24, Buenos Aires, Argentina.

Otero, Maria. 1994. "The Evolution of Nongovernmental Organizations Toward Financial Intermediation." In The New World of Microenterprise Finance: Building Healthy Institutions for the Poor. Eds. Maria Otero and Elisabeth Rhyne. West Hartford, CT: Kumarian Press, pp. 94-104.

Rhyne, Elisabeth. 1994. "A New View of Finance Program Evaluation." In The New World of Microenterprise Finance: Building Healthy Institutions for the Poor. Eds. Maria Otero and Elisabeth Rhyne. West Hartford, CT: Kumarian Press, pp. 105-116.

Rhyne, Elisabeth. 1998. "The Yin and Yan of Microfinance: Reaching the Poor and Sustainability." MicroBanking Bulletin. 2(1): pp. 6-8.

Rodey, Barbara J. 1997. The Spiritual Dimensions of Microfinance: Towards a Just Civilization and Sustainable Economy. Paris: European Baha'I Business Forum.

Rogaly, Ben. 1996. "Micro-finance Evangelism, 'Destitute Women,' and the Hard Selling of a New Anti-poverty Formula." Development in Practice. 6(2): pp. 100-112.

Schreiner, Mark. 1997a. "Ways Donors Can Help the Evolution of Sustainable Microfinance Organizations." Occasional Paper No. 2362, Rural Finance Program, Ohio State University.

Schreiner, Mark. 1997b. "How to Measure the Subsidy Received by a Development Finance Institution." Occasional Paper No. 2361, Rural Finance Program, Ohio State University.

Sebstad, Jennefer and Gregory Chen. 1996. "Overview of Studies on the Impact of Microenterprise Credit." Assessing the Impact of Microenterprise Services. Washington, D.C.: Management Systems International.

Seibel, Hans Dieter and Uben Parhusip. 1998. "Microfinance in Indonesia: As Assessment of Microfinance Institutions Banking with the Poor." Occasional Paper No. 2365, Rural Finance Program, Ohio State University.

Trape, Philippe and Jean-Francis Benhamou. 1996. "French Assistance for the Development of Micro-finance Institutions." CGAP Newsletter No. 2. Washington, D.C.: The World Bank, pp. 20-22.

Von Pischke, J. D. and Dale W. Adams. 1980. "Fungibility and the Design and Evaluation of Agricultural Credit Projects." American Journal of Agricultural Economics. 62(4): pp. 719-726.

Webster, Leila. 1995. "The Informal Sector and Micro-finance Institutions in West Africa [Implications for the World Bank]." CGAP Newsletter No. 1, pp. 9-10. Washington, D.C.: The World Bank.

Yaron, Jacob. 1994. "What Makes Rural Finance Institutions Successful?" The World Bank Research Observer. (9)1: pp. 49-70.

Yaron, Jacob. 1997. "How to Assess Performance of Development Finance Institutions." In Microfinance for the Poor? Ed. Hartmut Schneider. Paris: OECD, pp. 63-71.

Notes

(1.) In using the broad categories "institutionist" and "welfarist," we recognize that neither camp is characterized by complete unity of thought and that there exists substantial crossover between the two camps. We have tried to limit our discussion to those issues on which there appears to exist widespread agreement within each camp. The core beliefs of primary institutionist writers can be discerned both from their many publications and by the fact that the same themes appear time and again.

(2.) Most institutionists do not have a problem with group-based lending schemes. In fact, BancoSol does a great deal of it. For them, the issue in dispute has to do with the related aspects of "social intermediation" (Bennett et al., 1996), or the social benefits of group lending. Institutionists tend to look on group lending as a purely financial mechanism whereas welfarists tend to look on it as something more comprehensive.

(3.) Private sources capital consists primarily of depositor savings, commercial debt, equity, and venture capital. Among institutionists there is a strong current who look to venture capital, equity markets, and refinancing facilities to help MFIs grow. An even stronger current, however, looks to savings mobilization as the main source of MFI growth.

(4.) The Microcredit Summit, for example, has made "institutional sustainability" one of the Summit's core themes.

(5.) Specifically, Morduch critiques the following eight institutionist claims: (1) Raising interest rates does not substantially diminish demand for loans. (2) Financially sustainable programs can achieve greater scale than subsidized program. Thus, they can make a bigger dent in poverty. (3) Financial sustainability is critical for institutions as it is the route to being able to access capital from commercial financial markets rather than donors. (4) Since sustainable programs do not require outside funding, consideration of costs and benefits is irrelevant. There are no costs borne by governments or aid agencies--there are only benefits. Sustainable programs are thus superior to subsidized programs. (5) Subsidized credit programs are inefficient and ultimately bound to fail. (6) Subsidized credit more often ends up in the hands on nonpoor households. (7) Microfinance has been and should continue to be a movement with minimal government involvement. (8) Mobilizing savings is not likely to make sense for subsidized credit programs.

(6.) By way of example, Schreiner (1997a: p. 2) claims that "without profits, an [MFI] will shrink and die." Similarly, Gonzalez-Vega (1998: p. 7) argues that the "successful implementation of the new technologies will only occur, in turn, if the structure of organizational incentives promotes sustainability [financial self-sufficiency]." It is important to note the tone of such statements. The authors do not suggest general tendencies (which are easier to defend) but make absolute assertions (which are harder to defend).

(7.) Hulme and Mosley (1996: p. 158) go further to argue that "the case that formal sector for-profit institutions could take a lead role in providing financial services to low income households finds little support from ... the wider empirical literature.... Private companies are simply not prepared to provide the venture capital for experimental services to low-income borrowers" in order to "extend financial services deeper down the socio-economic pyramid."

(8.) One might charge that our use of the term "selfish" investor constitutes a semantic counter-trick. We did not, however, invent this term. We borrowed it from institutionist literature (Schreiner, 1997b).

(9.) Yaron's (1997) Subsidy Dependence Index (SDI) is one way to measure the social worth of an MFI. The SDI calculates the percentage by which an MFI would need to raise interest rates in order to cover its economic costs of capital. From a social investment perspective, the relevant cost would be the investor's (donor's) social opportunity cost of capital. As typically used, however, the relevant opportunity cost is measured from the MFI's perspective, which is the cost of raising funds from private sources.

(10.) large number of impact studies have been published demonstrating the impact of microfinance (see Sebstad and Chen, 1996). Nonetheless, the findings of many of these studies are questioned for a variety of methodological failings, such the inability to control for the fungibility of funds, the lack of appropriate control groups, and selection bias (Gaile and Foster, 1996; Hulme, 1997; Morduch, 1998a; Von Pischke and Adams, 1980). It is ironic that while welfarists are being held to ever-increasing methodological standards for demonstrating program impact, institutionists routinely make broad inferences based on very small and typically highly biased sample sets.

(11.) Six of the eleven MFIs examined clustered in range of $200 to $400 for average outstanding loan size, five of the eleven were financially self-sufficient, and ten of the eleven were operationally self-sufficient.

(12.) (1998d) reports that "some prognosticators at [CGAP] ... estimate that 3-5 percent of programs worldwide are currently financially sustainable--and another 7-10 percent will cross the hurdle within ten years."

(13.) paragraph paraphrases the original language by Barge (1985: p. 28).

(14.) writers identify six dimensions to outreach: (1) depth of outreach, (2) breadth of outreach, (3) quality of outreach, (4) cost of outreach, (5) length of outreach, and (6) variety of outreach (Navajas, et al., 1998: pp. 6-11).

(15.) fairness to institutionists, it was the welfarists who raised average loan size as an indicator of poverty. The former were resistant to using any indicator at all.

Biographical Sketches

Gary M. Woller

Romney Institute of Public Management

Marriott School of Management

Brigham Young University

Christopher Dunford

Freedom from Hunger

Warner Woodworth

Marriott School of Management

Brigham Young University