Ownership and Technical Efficiency of Microfinance Institutions:

Empirical Evidence from Latin America

April 26, 2011

Roselia Servin[1] ·Robert Lensink[2]·Marrit van den Berg[3]

Abstract

By using stochastic frontier analysis, this paper examines technical efficiency of different types of microfinance institutions in Latin America. More specifically, the paper tests whether differences in technical efficiency can be explained by differences in ownership. We focus on Non-Governmental Organizations, Cooperatives and Credit Unions, Non-Bank Financial Intermediaries and Banks. Our dataset contains 1,681 observations from a panel of 315 MFIs operating in 18 Latin American countries. We differentiate between intra-firm and inter-firm efficiency. Our results show that Non-Governmental Organizations and Cooperatives/Credit Unions have much lower inter-firm and intra-firm technical efficiencies than Non-Bank Financial Intermediaries and Banks, which indicates the importance of ownership type for technical efficiency.

Keywords: Microfinance · intra-firm · inter-firm · catch-up · technology · Stochastic Frontier · ownership type. Latin America.

Introduction

Microfinance Institutions (MFIs) provide financial services to poor people who have no access to commercial funding.Until recently, many people had an extremely positive view about the potential role of microfinance. An increase in access to microfinance was considered to be an important instrument for reducing poverty in developing countries (Mersland, 2000). Many people even saw the development of microfinance, by providing financial services, like loans, deposits, insurance and organizational help to poor households, as one of the main innovations in the last 25 years(Meeusen and Broeck 1977, Montgomery 1996, Qayyum 2006).

However, recent developments challenge the extremely positive view on microfinance and question whether microcredit can contribute to a serious reduction in world-wide poverty in the short or long run. Especially after the stories about loan-shark-style MFIs who have driven borrowers to suicide in the Indian state of Andhra Pradesh, the rosy view about microfinance started to vanish. Some authors argue that these negative developments are a result of governance failures of MFIs, partly related to the recent trend of commercializing microfinance, and the corresponding expansion of services that MFIs offer. Whereas once they focused only on microcredit, in the form of small loans to the poor, today they have shifted from microcredit to microfinance, including savings and insurance. The funding situation of MFIs has also started to change rapidly. Whereas once they relied mainly on private and public donors and aid organizations, an increasing number of MFIs turn to the capital market for funding. The strive for commercialization also induced, and partly required, a change in ownership of MFIs: several MFIs started to change their ownership structure by changing from a Non-Governmental Organization (NGO) into a share-holder company.

A better understanding of governance issues is needed to explain why some MFIs are successful where others fail, and to preclude Andra Pradesh like developments. Unfortunately, the literature on microfinance governance is stillvery scarce and often anecdotal(Lebie and Mersland 2011). Some papers study the relevance of certain governance characteristics of MFIs(see for instance Hartarska(2005) and Lebie (2010)). Other papers more specifically focus on the role of the type of ownership of MFIs (see e.g. Mersland and Strøm, 2008). The ownership structure is an important element of the governance of an MFI. Several authors argue that nonprofit MFIs should transfer into shareholder owned firms (SHFs) (seeMersland and Strøm(2009)for references), e.g. since SHFs can be regulated by banking authorities, and benefit from superior corporate governance because they are privately owned. Mersland and Strøm (2008), however, contradict this hypothesis. Their study suggests that SHFs and NGOs perform similarly in terms of both social and financial aims.

In line with Mersland and Strøm (2008), our study examines the relevance of various ownership structures for the performance of MFIs. However, unlike them, following Lansink et al., (2001)we focus on intra-type and inter-type technical efficiency. We will derive and compare technical efficiency for four groups of MFIs in Latin America: NGOs, Cooperatives/Credit Unions, Non-Bank Financial Intermediaries and Banks. The ownership type of MFIs differs considerably from that of traditional commercial banks. In general, MFIs are significantly smaller in size, limit their services towards poor households and often provide small collateral-free group loans(Haq et al. 2010){, 2010 #9;Haq, 2010 #9}. In addition, most MFIs have dual objectives, a financial and a social, and their source of income is not only deposits, but also donations(Gutierrez-Nieto et al. 2009){Gutierrez-Nieto, 2009 #81;Gutierrez-Nieto, 2009 #81}. Furthermore, only a limited number of MFIs are regulated and allowed to mobilize savings(C-GAP 2003, Mersland 2009).

Our study applies stochastic frontier analyses (SFA) to estimate technical efficiency of the four groups of MFIs. There are several other MFI efficiency studies available. However, these studies focus on cost efficiency(Gregoire and Tuya 2006, Hartarska et al. 2006, Hassan and Tufte 2001, Hermes et al. 2008). We are the first who analyze the impact of the type of ownership on technical efficiency of MFIs. We have several reasons why we estimate technical efficiency, and ignore cost and/or profit efficiency. First, for most MFIs good price information is lacking, which makes profit and cost functions difficult if not impossible to estimate.Second, and most importantly, there are theoretical and methodological objections to focus on profit efficiency and/or cost efficiency for MFIs since these concepts assume that MFIs are maximizing profits and are price-takers in input markets and, in the case of profit efficiency, output markets. MFIs, however, have multiple objectives involving both helping the poor and financial sustainability and therefore do not necessarily maximize profits, and hence do not aim to become profit efficient. Moreover, most MFIs have at least some sovereignty in setting interest rates and can affect the costs of their capital –the share and nature of subsidies, through lobbying. Technical efficiency involves achieving the maximum output given inputs and seems therefore a more relevant concept for MFIs.

Another main difference between our study and the existing MFI efficiency studies is that all previous efficiency studiesestimated a single frontier for all MFIs types (NGO, Cooperative/Credit Union, NBFI and Bank). Since it is highly likely that the technology differs per ownership type, the common frontier assumption may lead to biased efficiency estimates(Bos and Schmiedel 2007){Bos, 2007 #402;Bos, 2007 #385}. We will explicitly allow for differences in technology and test the appropriateness of this assumption.

Our study focuses on Latin America because this region contains a rich variety of MFIs in terms of ownership type, which is needed to conduct our analyses. There are many small non-profit MFIs in Latin America, financially supported by international funders (Gutiérrez-Nieto et al. 2007). However, Miller (2003)show that some of the most experienced, developed, and profitable MFIs around the world can also be found in Latin America.On average, MFIs from Latin America have more assets, are more leveraged, and make use of an increasingly growing share of commercial funds compared to institutions from other regions.

Latin America is also interesting to study since especially for this continent many MFIs are under pressure to transform their organizational structure from an unregulated non-profit institution to a regulated share-holder institution(Mersland and Øystein Strøm 2009, Nimal 2004). In Latin America, competition in microfinance has triggered MFIs to transform into a share-holding company, for which it becomes important to cover lending costs with income generated from the outstanding loan portfolio and to reduce these costs as much as possible Rhyne (2006). We focus on one region since we want to have a relatively homogeneous sample of MFIs, so that performance differences can be attributed to differences in ownership types, and not caused by severe regional disparities.

The remainder of this paper is structured as follows. Section 2 surveys the literature on MFI ownership, and derives hypothesis regarding the link between ownership type and technical efficiency. Section 3 discusses the methodology. Sections 4 and 5 describe the empirical model and the data, respectively. The estimation results are presented in Section 6. Section 7 gives a conclusion.

2. Ownership structure and MFI Technological Efficiency[4]

The main function of micro finance governance is to control self-interested managers to solve possible agency problems. However, corporate governance is one of the weakest features of MFIs(CSFI 2008, Hartarska 2005, Mersland and Øystein Strøm 2009). The philanthropic status of many MFIs reduced demands for accountability (CSFI 2009)so that MFIs were able to attract funding at an increasing rate and to keep on growing without adjusting their governance system. When sufficient oversight is lacking, managers are likely to enrich themselves or pursue other self-interests at the MFI’s expense. In addition, corporate governance in MFIs is particularly difficult due to their dual mission, i.e. to be financially sustainable while reaching out to the poor. The dual mission of MFIs, and the lack of external control, provides managers from MFIs some managerial discretion which affect outcomes and efficiency. However, as we will argue below, this managerial discretion will not be the same for all types of MFIs.

The ownership type of MFIs is very important for the governance system within the MFI, and ultimately MFIs performance. Within the microfinance industry, a variety of ownership structures exists, such as Banks, Non-Bank Financial Institutions (NBFI), Credit Unions and Cooperatives and Non-Governmental Organizations (NGO). Banks and NBFIs are shareholder firms, which distribute excess profits to their shareholders. NBFIs, unlike banks, are limited by law in the range of services they can offer; some cannot provide savings accounts. Credit unions and Cooperatives are non-profit organizations that are owned and controlled by their members. Unlike NGOs, they are allowed to distribute profits to their members. NGOs are non-profit, non-governmental organizations, characterized by a non-distribution constraint.

MFIs of all ownership types have social and financial motives. However, the relative weight of the two objectives differ per MFI type. NBFIs and banks are shareholder firms, with clearly defined financial objectives, whereas NGOs and Cooperatives or Credit Unions put much more weight on the social objectives.Since the main objectives are not the same for all ownership types, it is highly likely that the technology will differ per organization structure for the technology determines the optimal relationship between inputs and outputs. Organizations serving poorer clients, such as NGOs and Cooperatives or Credit Unions, have higher average costs than NBFIs and Banks, because small loans are costlier to provide(Cull et al. 2009). Moreover, MFIs with a clear social orientation often combine loan provision with training, which is a labor intensive activity that does not directly contribute to output in terms of loans provided. Finally, a focus on poverty orientation may result in targeting of people in more remote areas and visiting these people, who would otherwise be at their own communities. These differences in orientation make it highly likely that the technology differs for the different groups of MFIs: an appropriate technology for a NBFI or Bank type of MFI may not be the best technology for a NGO and/or a Cooperative or Credit Union. This will lead us to our first hypothesis:

Hypothesis one: MFIs characterized by different ownership types will use different technologies.

The difference in technology that is used by a particular MFI may be a deliberate choice since the MFI focuses on different objectives. A MFI focusing primarily on social objectives may therefore decide not to adapt the best technology available in terms of increasing the output (for instance number of loans) given the available inputs. Rather, this type of MFI may opt for a technology that is not best in providing loans but provides better opportunities to offer trainings. This may, for instance, be reflected in the type and skills level of the loan managers that are hired. A loan manager that is good at finding new clients and/or improving repayment levels may not necessarily be the best loan manager offering trainings. However, a difference in inter-MFI technology may also be caused by constraints in adopting innovations, such as information technology, or constraints in improving input qualities on account of managerial capability, experience or education (Lansink et al. 2001). Obviously, MFIs that are faced with lower inter-MFI technology due to constraints that force them to adapt less efficient technologies should adapt better technologies if these constraints are relaxed.

Regarding our classification of MFI types, it is to be expected that NGOs and Credit Unions or Cooperatives are more constraint in funds since they do not raise funds via capital markets, and have, given their focus on social goals, less opportunities to attract funds from private investors. The combination of a stronger focus on social goals and possibly a more severe constraint in terms of funding possibilities leads us to our second hypothesis.

Hypothesis two: NGOs and Cooperatives or Credit Unions will be faced with lower inter-firm technologies than NBFIs and Banks

The production possibilities frontier of share-holder NBFI and Banks is higher relative to non-shareholder NGO and Cooperative or Credit Union as the former make use a more optimal combination of inputs and outputs in the production process leads them to achieve higher inter-firm efficiency over time. Then, the differences in inter-firm efficiency between ownership types of MFIs steam from efficiency differences and as well as from differences in technology. For instance, NGOs are generally providing credits trough the group lending methodology whereas Banks and NBFIs focus a lot on individual lending. As a result, this differences in technology requires different skill of personnel working in such methodology, different capital requirements and leads to different operating costs per loan or borrowers. All this characteristics of shareholder MFIs (Banks and NBFI) make them to be close to the best production possibility frontier with respect to their peers, the NGO and Cooperative and Credit Union which rely in dissimilar technology that is more suitable to meet the social needs of this ownership types of MFIs.

The ownership type of MFIs is also important since managerial discretion differs across organizational types. Managers in an for-profit MFI (NBFI or Bank) may try to pursue policies in their own interest, and to gain private benefits, possibly at the expense of the overall performance of the MFI. Managers may deliberately decide not to use the existing technology in the most optimal way in order to increase his/her own private benefits. The possibility to do this will increase the more managerial discretion a manager has.

Managerial discretion depends on several issues which are relevant for explaining why different ownership types may offer different possibilities to gain private benefits leading to intra-firm technological inefficiencies. An important mechanism concerns the governance within the MFI. Since NGOs and Credit Unions or Cooperatives are nonprofit organizations, their governance is not tied to ownership. Board members of nonprofit organizations, unlike NBFIs and Banks, normally do not have a financial stake in the organization, and often lack financial knowledge and experience with risk management, which probably will reduce active control of the managers. Moreover, NGOs, unlike Credit Unions and Cooperatives, NBFI and Banks, are not allowed to distribute profits to any stakeholders. Since excess profits need to be returned to the organization, the non-distribution constraint will provide considerable organizational slack, which will give NGO executives considerable managerial discretion(Glaeser 2003). Regulation is another important mechanism that may limit discretionary power of managers. Since NGOs and Cooperatives or Credit Unions do not offer savings products, they are normally not regulated by central banking authorities. Banks and NBFI, on the other hand, are usually heavily regulated by the monetary authorities, which will reduce their discretionary power. Finally, NBFI and Banks have more clearly defined financial objectives, while the objectives of NGOs and Credit Unions and Cooperatives are much more unclear. If objectives are unclear and/or are more of a dual character, it becomes more difficult to develop an appropriate governance system and incentive system that will induce managers to behave such that MFI performance will be optimized.

Table 1 ranks the organization types of MFIs according to the extent of managerial discretion. The table suggests that NGOs have the most discretion and Banks the least.

Table 1. Managerial Discretion Typology of MFIs