The Determinants of Micro Finance Institutions’ Decision to Receive a Social Rating: An Institutional and Resource-Dependence Perspective

Abstract

We examine the determinants of MFIs’ social rating decisions according to institutional and resource dependency pressure factors. We exploit a unique data set containing 221 MFIs resulting in 767-year observations, obtained from both mixed market and social rating agencies (Planet, MicroFinanza, and Microrate rating agencies). Our data was collected for both rated and non-rated MFIs between the years 2006 and 2018, in six major less-developed regions of the world. Our study contributes to the CSR literature in microfinance as the decision to be rated acts like a CSR signal. Our research shows that the tighter the rule of law the less inclined MFIs are to seek a rating. In addition, we borrow from the resource-dependence perspective to demonstrate that proxies for resource (in)dependence such as for-profit status, mature stage, and subsidization are negatively associated with the obtention of a social rating. Overall, in the less-developed regions of the world, when MFIs operate in a strongly regulated environment and have a strong position (financial and mission-wise), with respect to investors, they are less willing to obtain a social rating.

1. Introduction

Microfinance is defined as the supply of savings, credit, insurance, and payment services to relatively poor people [1]. Originally, microfinance institutions (MFIs) offered financial services only to low-income clients, but they have since broadened their scope to include anyone usually excluded by mainstream financial services [2]. They are recognized for bridging the gap between formal financial institutions and the rural poor by making individual microcredit loans directly to villagers, micro entrepreneurs, impoverished women, and families. MFIs are not fringe elements of the finance industry anymore. The microfinance business model has proven profitable and is robustly expanding worldwide. This model is gaining ground in the financial and banking systems, with outstanding growth, at least in developing countries; both in the number of entities and in the number of clients [3].

However, a recent trend towards commercialization in the industry has led to mission-drift according to pioneers in the industry [4,5,6]. The emergence of crises and scandals from different countries across the world over the past few years has polarized the opinion about the real impact of microfinance [5]. Nowadays, a valid question has been raised regarding whether microfinance institutions (MFIs) actually help bring people out of poverty. This is, to a large extent, due to MFIs charging excessively high lending rates and practicing heavy-handed collection methods [7]).

In order for MFIs to make a significant and long-term contribution to improving financial inclusiveness, it is important to know more about factors that may help these institutions achieve their financial and social goals [8]. Following concerns about predatory lending and erupting scandals in the industry, the rise of socially responsible investing in the MFI sector has induced a demand for criteria in order to assess the Corporate Social Responsibility (CSR) behavior of these institutions. Social ratings may be one such device that can help MFIs fulfill their mission. In the last decade or so there has been a surge in rating agencies’ activities in the microfinance space [9]. Traditionally, socially responsible investors rely heavily on ESG scores devised by sustainability rating agencies, which provide legitimacy to these types of investment decisions [10]. The purpose of rating reports is to present independent information that stakeholders, such as lenders, donors, owners, or managers, can use to make informed decisions [7].

Social ratings in the microfinance industry act as a proxy for CSR. Several theoretical and empirical studies have examined the institutional determinants of corporate social responsibility in corporations [11,12,13,14,15,16,17,18,19]. These studies provide evidence that the institutional environment creates pressure for corporations to act in a socially responsible manner. There is evidence in the banking sector that certain aspects of the institutional environment affect social responsibility disclosure [20,21,22].

Social ratings are obtained on a voluntary basis and thus are the result of a strategic decision by MFIs. What are the determinants of this decision to be rated? Several studies have delt with the determinants of MFIs’ rating decisions [3,7,9,23,24,25,26,27,28]. Tchakoute-Tchuigoua [26,27,28] indicates that the rating decision is significantly associated with the institution’s age and social performance. The author further shows that MFIs only ask for ratings insofar as they are capable of supporting full or partial rating fees. Hatarska et al. [25] conclude that the choice of being rated is significantly influenced by the MFI’s age, its strong focus on credit activity, and its outreach, as represented by the number of loans granted. Beisland et al. [24] show that social ratings of MFIs are significantly related to financial performance, greater outreach especially in rural areas, well-defined social objectives, staff commitment, service quality, and enhanced customer service.

On the other hand, the extant literature has not yet analyzed how the constraints imposed by the institutional environment and the MFIs’ dependence on financial resources may impact their decision to obtain a social rating.

The institutional environment plays a pivotal role in the MFI sector today. Generally, regulations serve to manage and prevent the risks that MFIs face. Institutional theory broadly states that the behavior of firms is governed by its institutional environment, with corporations subjected to more or less formal rules ranging from coercive political regulation to constraints such as the normative influence of professional groups [11,12,15,20,22,29,30]. Institutional Theory as popularized by DiMaggio and Powell [31] ascertains that institutions create a variety of coercive, normative, or mimetic pressures on firms to adopt particular structures or practices that establish their legitimacy. Institutionalists have, for instance, suggested that the governance system of non-profit organizations is less efficient than that of regulated MFIs [32].

On the other hand, MFIs are characterized by a high level of dependency on subsidies and donations [33,34]. Because donors and investors want to know how well their money is managed, this has given more impetus to the rating business, which, for instance, is widely used as a control mechanism for Non-Governmental Organizations (NGOs) [3].

The funding landscape of MFIs has also mutated towards increased financial integration and sustainability. Today, these institutions are classified as for-profit and not-for-profit MFIs. The variation in ownership forms is, in itself, an interesting aspect of the microfinance industry. An essential characteristic that separates microfinance from most other financial industries is the coexistence of donors and professional investors within the same industry [23].

Verbruggen et al. [35] use Resource Dependence Theory (RDT) and the concept of coercive isomorphism [36] to explain not-for-profit organizations’ compliance with accounting and reporting standards and conclude that organizations that rely on governmental resources and financial loans are more strongly inclined to comply with IRS standards. According to RDT, three main factors determine the dependency of a firm on scarce resources. These factors are the resources’ value, the availability of that resource, and the competition between firms to control that particular resource. This issue is particularly salient in less-developed parts of the world.

Gálvez Rodriguez et al. [37] conclude that not-for-profit organizations that receive funding and those with better financial standing disclose more information. Chih et al. [21] raise the question of whether institutional variables affect responsible behavior in the financial sector and provide strong evidence that CSR is affected by both financial and institutional variables.

Nowadays, MFIs attract various types of investors: social investors willing to support the poverty-alleviation goals of these MFIs and conventional investors diversifying their portfolios by investing in social enterprises such as MFIs. Credit rating measurements as well as efficiency and sustainability ratios have gained in popularity [38]. These have become a means for MFIs to satisfy the informational demands of these investors, and thus to access external financing more easily. Investigating the MFIs’ environment and resource dependence’s impact on their social rating decision might, for instance, be useful for gauging the efficacy of policies such as interest rate subsidization.

Thus, in this article, we assess the determinants of the MFIs’ decision to receive a social rating. Specifically, we consider whether institutional coercive pressures [11,12,15,20,22,29] and resource dependency [39] drive that decision, while accounting for the uniqueness of the microfinance sector’s institutional forms (e.g., profit status); with a focus on MFIs located in the less-developed regions of the world. Hence, we contribute to the literature on the determinants of MFIs’ social rating in the microfinance sector by bringing into focus the role played by resource dependence and institutional factors in MFIs’ rating obtention choice.

Our empirical analysis exploits a unique data set of 221 MFIs resulting in 767-year observations, obtained from both mixed market and social rating agencies (Planet, MicroFinanza, and Microrate rating agencies). Our data was collected for both rated and non-rated MFIs between the years 2006 and 2018, in six major less-developed regions of the world.

Because the choice of a social rating is a binary choice, we implement a Probit model [40,41] and conclude that, in line with Institutional and Resource Dependence Theories, MFIs’ decision to receive a rating is determined by the rule of law and other variables that more or less constrain MFIs’ access to financial resources.

The remainder of the paper is organized as follows. Section 2 provides an overview of the background and literature. Section 3 presents the research methodology, data, and description of variables. Section 4 features the empirical results. Section 5 presents a discussion of the results, and Section 6 concludes and acknowledges the study’s limitations.

2. Theoretical Foundations and Hypothesis Building

2.1. Institutional Theory and MFIs’ Social Rating

We use DiMaggio and Powell [31] institutional framework approach to investigate MFIs’ incentives to seek a social rating. According to these authors, in order to survive, firms adapt to their institutional environment through three isomorphic institutional processes (coercive, mimetic, and normative). Coercive isomorphism results from political influence, in other words, these are formal and informal pressures exerted by other organizations, which MFIs are dependent upon. Mimetic isomorphism results from responses to uncertainty. In times of uncertainty, MFIs may look to imitate other organizations perceived as legitimate and successful role models. Normative isomorphism is associated with professionalization. Formal education and the cognitive legitimation of academics, as well as the growth and development of professional networks through which organizational models spread, are two aspects of professionalization considered important sources of normative isomorphism [36].

MFIs that adjust their practices, such as, for instance, choosing to adopt the IFRS framework in drafting financial statements, do so in response to the three types of response to institutional pressures: (1) from the organizations on which they depend, (2) from the uncertainty pushing them to imitate social actors viewed as successful and legitimate, and (3) from the professionalization of their field of activity.

Institutional Theory emphasizes the role played by institutional structures in imposing order on an incomplete world. In that framework, organizational change is motivated by a search for formal legitimacy or the stakeholders’ need to conform to expectations in the environment [42]. Prior to the 1970s, most organizational analysis focused on the internal workings of organizations [36]. After the 1970s, scientific research began to focus on both internal and external organizations. Authors who propounded new institutionalism included scholars such as DiMaggio and Powell [31], Meyer and Rowan [43] and Pfeffer and Salancik [44].

The application of Institutional Theory to understand CSR-related phenomena is a rather recent trend. Only in the mid-2000s did a literature emerge which broadened the array of conceptual tools used in CSR research. In its very definition, Institutional Theory appears to be right at the center of what CSR is all about [30]. When faced with the pressure to conform to evolving societal expectations, many corporations demonstrate their CSR engagement by introducing new initiatives and new ways of communicating with stakeholders.

According to Campbell [12], institutional analysis is important because institutionalists understand that institutions beyond the market are often necessary to ensure that corporations are responsive to the interest of social actors besides themselves. Campbell [12] proposes that “Corporations will be more likely to act in socially responsible ways if there is a system of well-organized and effective industrial self-regulation in place to ensure such behavior, particularly if it is based upon the perceived threat of State intervention or broader industrial crisis, and if the State provides support for this form of industrial governance” [12] (p. 956).

Chih et al. [21] set out to undertake a thorough, point-by-point examination of the theory postulated by Campbell [12], based on a sample heavily skewed towards developed countries. They analyzed the conditions under which corporations may or may not act in socially responsible ways. They empirically concluded that financial firms in countries with stronger levels of legal enforcement tend to engage more in CSR activities.

Microfinance has often been criticized on the basis that it has become a socially acceptable mechanism for extracting wealth and resources from poor people with many lenders charging exorbitant interest rates that sometimes reach up to 200% per annum, as in the case of Banco Compartamos [45]. In spite of the CSR issues faced by the industry, most MFIs claim to adopt an approach where they follow a double bottom line [39], focusing on both social and financial performance. These MFIs are orienting their business practices not only to improve financial performance or market appreciation, but also to comply with more socially responsible practices [46].

At the same time, socially responsible investors in the MFI space rely heavily on social rating agencies because they do not have the capacity to assess these institutions’ sustainability on their own. Social rating agencies collect information from the public as well as directly from companies to provide social scores. Rating Micro Credit Companies’ (MCCs) operations can, in theory, bring CSR legitimacy and transparency [37]. Among the possible benefits derived from establishing legitimacy are: (1) lowering regulatory burden and (2) avoiding the stigma associated with a reputation for environmental irresponsibility.

Thus, on the one hand, one could consider the obtention of a social rating as a complement to pressures brought about by the institutional environment. That is, ‘good’ organizations will choose to obtain a rating to further solidify in investors’ and other fund providers’ minds the notion that they (MFIs) are aligned with CSR expectations [47,48]. Obtaining a rating is thus expected in the financial landscape.

On the other hand, the obtention of a social rating may function as a substitute for institutional pressures, especially when the institutional environment is not conducive to producing good CSR behavior, and there MFIs need to find ways to signal their good behavior. In that case, a social rating acts as a proxy signal for good CSR behavior.

Here, we examine the impact of the rule of law on the MFI industry’s decision to obtain a social rating in less-developed countries. We use the strongest reported measures of legal protection since it is through law and legal institutions that states seek to order the behavior of individuals and organizations so economic and social policies are converted into outcomes.

In line with Institutional Theory [12], and coercive isomorphism, i.e., pressure from other organizations, we postulate that a strong institutional environment is a viable route to put CSR at the heart of the relationship between the MFI business and society. In developed countries, the obtention of a rating is considered a necessary step in the lifecycle of a financial institution in an already strong CSR environment. In that case, very high levels of peer and legal pressure have created a social norm around mimicking others’ CSR corporate policies. Thus, ratings act as a complement to an already elevated level of application of the rule of law.

However, in less-developed countries, the choice of obtaining a social rating is more appealing for MFIs when they are not receiving enough pressure to conform to CSR standards from the legal environment. In this case, social ratings are substitutes for a lack of institutional pressure. Consequently, for less-developed countries, our hypothesis is that the obtention of a rating plays the role of a CSR signal that substitutes for a lack of law enforcement and regulatory pressures. Thus, in these countries where CSR behavior is less the norm, i.e., less coercive isomorphism than in developed nations, there is more of a need for MFIs to distinguish themselves by using ratings. We hypothesize that:

Hypothesis 1 (H1).

MFIs operating in countries with a stronger rule of law will be less likely to ask for a social rating.

It is worth noting that this hypothesis is not in opposition with the results presented by Chih et al. [21] that financial firms in countries with a higher level of legal enforcement engage more in CSR behavior. This is because their result is based on a sample heavily skewed towards developed countries.

2.2. Resource Dependence and MFI Social Rating

MFIs are characterized by their reliance on government funding, private donations, and fees together with commercial sources of funding (Debt and Equity). This dependence on outside resources makes them vulnerable to changes in the flow of resources on top of institutional pressures.

Both the Institutional and Resource Dependence Theories agree that organizations endeavor to obtain stability and legitimacy [49]. Compliance is necessary to satisfy stakeholders demanding increased responsibility and value-added transparency. The financial needs of MFIs define their dependency. The alignment of MFIs with their funders can be theorized as coercive isomorphism, which means that the pressure exerted by one organization on another comes from the similarity in these organizations’ processes [35].

According to Fehr and Hishigsuren [50], internal financing in MFIs takes the form of savings, either forced or voluntary. On the other hand, external financing can be subdivided into (1) equity financing (quasi-equity, retained earnings, socially responsible equity), (2) debt financing (commercial loans, guarantee funds, bonds, securitization and interbank lending, and commercial equity) and (3) grant financing.

Funding of MFIs’ activities originates primarily from outright donor grants, government subsidies, and, often, debt capital, including debt with non-market terms favorable to MFIs. Each of these funding sources supports the MFIs’ poverty reduction strategies. However, especially in less-developed countries, these traditional sources of financing may not be sufficient to allow MFIs to provide maximum services. As such, these MFIs might be more inclined to approach new classes of potential investors.

Some MFIs may have access to market-based forms of debt financing. However, market-priced financial capital is usually only accessible to for-profit MFIs. When MFIs become sufficiently profitable, they can tap into private capital markets for investment funds. Organizations that rely heavily on few resource providers are likely to experience stronger constraints from their environment. Because socially responsible funders are sensitive to CSR signals (i.e., social ratings), we hypothesize that there is a relationship between the MFIs’ decision to obtain a social rating and resource dependence. One may expect that MFIs in need of external funding will be more likely to ask for social ratings. Hence, we postulate:

Hypothesis 2 (H2).

MFIs that are more dependent on external sources of financing for their lending activities will choose to obtain a social rating.

3. Research Design

3.1. Data and Sample

We construct our sample by combining different databases, i.e., the Microfinance Information Exchange (MIX) for financial and social performance indicators, Micro Finance Rating Agencies reports, and World Bank Word Governance Indicators (WGIs). The microfinance rating agencies’ reports were exploited in order to identify those institutions that decided to obtain a rating.

Initially, we collected 399 rating reports gathered from three main rating agencies between the period 2006–2020: Microfinanza, Planet rating, and Microrate rating agencies. We then selected MFIs that had reports and were in the MIX database. In order to ensure data reliability, unrated MFIs were chosen among the more transparent MFIs (MFIs with at least a three-diamond level of disclosure rating). As shown in

Table 1. Distribution of MFIs according to Region. This table presents the distribution of rating reports across regions. The data comes from two sources: first, the websites of rating agencies specialized in microfinance social ratings (Microrate, Microfinanza, and Planet rating) and second, from the Microfinance Information eXchange database.

Region	No of Observations	%
Sub-Saharan Africa	84	10.95
East Asia and the Pacific	91	11.86
Eastern Europe and Central Asia	99	12.91
Latin America and The Caribbean	368	47.98
Middle East and North Africa	69	9.00
South Asia	56	7.30
Total	767	100.00

and

Table 2. MFIs’ Distribution according to Fiscal Year. This table presents the distribution of rating reports across fiscal years. The data comes from two sources: first, the websites of rating agencies specialized in microfinance social ratings (Microrate, Microfinanza, and Planet rating) and second, from the Microfinance Information eXchange database.

	Whole Sample		Rated MFIs		Non-rated MFIs
Fiscal Year	No of Observations	%	No of Observations	%	No of Observations	%
2006	63	8.21	2	0.74	61	12.35
2007	64	8.34	3	1.10	61	12.35
2008	72	9.39	15	5.49	57	11.54
2009	81	10.56	25	9.16	56	11.34
2010	96	12.52	47	17.22	49	9.92
2011	96	12.52	44	16.12	52	10.53
2012	76	9.91	25	9.16	51	10.32
2013	65	8.47	13	4.76	52	10.53
2014	48	6.26	24	8.79	24	4.86
2015	45	5.87	26	9.52	19	3.85
2016	28	3.65	16	5.86	12	2.43
2017	23	3.00	23	8.42
2018	10	1.30	10	3.66
Total	767	100	273	100	494	100.00

, our final sample includes 767 MFIs-year observations for 221 MFIs between the period 2006–2018. We have 160 rated MFIs for a total of 273 (MFIs-year) observations and 61 MFIs that did not choose to be socially rated over this period, for a total of 494 (MFIs-year) observations. All these institutions are located in six major less-developed regions of the world.

3.2. Variables

3.2.1. Dependent Variable

The choice to obtain a social rating: our study makes use of social rating reports available with MicroRate, Planet Rating, or MicroFinanza rating agencies between the years 2006–2018. The decision to obtain a social rating constitutes what is known as a discrete choice model. Based on these reports, we attribute the dummy value of 1 to those MFIs that have decided to go for a social rating and the value of 0 to those who did not.

3.2.2. Main Explanatory Variables

Maturity: the life cycle theory of microfinance institutions denotes that various MFIs pass through stages of growth which are eventually characterized by their relationship with their funding type. According to Fehr and Hishigsuren [50], in the early stages of an MFI’s life, the major source of finance is primarily from donors who mostly anticipate social returns. As the enterprise progresses through the next stages, commercial loans become increasingly relevant for mature NGOs. Thus, the capital structure evolves with the changes in stages of the institutional life cycle. Our study considers several levels of maturity, 1 = New, 2 = Young, and 3 = Mature. According to the life cycle theory, a New MFI ranges between 0–4 years, the range is between 4–8 years for ‘Young’, whilst a lifetime in excess of 8 years is dubbed ‘Mature’.

Subsidized debt: captured by the difference between financial expenses on borrowing and the country lending market rate. It is defined as a dummy that takes the value 1 when the difference is negative, which reflects subsidization, and 0 otherwise.

Profit status: different types of organizational forms operate in the market. There are cooperatives, banks, non-bank financial institutions (NBFIs), and NGOs. NGOs are the main microfinance institutional forms categorized below. As an NGO transforms into a regulated institution, it starts looking for commercial sources of funding from financial markets and commercial banks in order to fund its rapid growth. Not-for-profit MFIs have smaller access to external funding sources and have less ability to raise external financing. We will thus attribute a binary variable: 1 when an MFI is profit-oriented, and 0 otherwise [28].

Country rule of law index: law enforceability captures the extent to which agents have confidence in and abide by the rules of society and, in particular, the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence. A better regulatory environment, or respect for the rule of law, leads to fairer processes for selecting and replacing governments and less abuse of public office for private gain. The rule of law variable was collected from World Bank Word Governance Indicators [51]. The index runs between values of −2.5 to +2.5 with higher index values being considered better.

3.2.3. Control Variables

Financial performance and risk: following previous studies that investigate the determinants of ratings [25,27,28], we control for MFI profitability and the riskiness of the loan portfolio, mainly because a social rating is solicited and has a cost that only profitable organizations can bear. We, therefore, measured profitability by two variables, namely the profitability (return on assets) and the real yield on the loan portfolio. We measure the riskiness of the loan portfolio by the portfolio at risk at 30 days.

Outreach to the poor: as a result of commercialization trends in MFIs, the poor are getting more vulnerable [34,52]. We consider that MFIs that take their CSR role seriously are those institutions that serve the poor and other financially excluded categories of people like women [53,54,55]. We thus control for (1) depth of outreach, measured by the average loan balance per borrower/GNI per capita. It is important to note that, here, the greater the magnitude of this variable is, the less the MFI serves the poor. On the other hand, we also control for (2) breadth of outreach, which is measured here as the percentage of female borrowers.

Other control variables include an asset side variable that captures the degree of an MFI’s asset structure measured by the loan portfolio to asset ratio and a liability side indicator measured by deposit to asset ratio.

A country’s overall economic condition is captured using the annual GDP growth rate of the said country.

3.3. The Empirical Model

Our objective is to analyze the decision to seek a social rating based on a sample of 767 MFI-year observations between the period 2006–2018. Our outcome variable captures the decision taken by an MFI to seek a rating or not and is measured by a dummy (discrete) variable. To answer this question, we thus apply a Probit model that is best suited to fit maximum likelihood models with dichotomous dependent variables [40,41].

Our data structure looks like an unbalanced panel data. As shown in Table 1 and Table 2, we are working with a sample made up of 221 MFIs for a total of 767-year observations over 13 years, i.e., an average of 3 observations per MFI over the period considered. For the rated MFIs, we found an average of almost 2 observations per MFI. Most of the rated MFIs that had requested the rating only did it once, which sounds completely logical. In that case, we do not necessarily have repeated observations for each cross-section over time. Thus, our sample more resembles one with independently pooled cross-section data. Rather than estimating a panel Probit model, we, therefore, pooled the data and controlled for year-fixed effects [40] The estimated model is as follows:

Social rating decision_ijt = α + β X_ijt + γ MFI controls_it + λ Country control_jt + δ_t + ε_it

(1)

where i indexes MFIs, t indexes years, and j the country. Social rating decision_ijt is the binary dependent variable taking the value 1 if the MFI solicited the rating in a given year, and 0 otherwise. X_ijt is a vector that capture our main explanatory variables: Maturity, Subsidized debt, Profit status, and Country rule of law index. MFI controls_it is the vector of control variables, which include: Financial performance (return on asset, real yield on loan portfolio) and Risk (portfolio at risk at 30 days); Outreach to the poor (Depth of outreach and Breadth of outreach); asset and liability side variables. Country control_jt captures the country’s rate of economic growth that we measure by GDP growth. The variable δ_t is the year-fixed effect and ε_it is the idiosyncratic error.

4. Results

4.1. Descriptive Statistics

The descriptive statistics and regressions use winsorized observations in order to minimize the influence of outliers.

Table 3. Descriptive Statistics. This table reports on the summary statistics of the variables included in the estimations. The data comes from two sources: first, the websites of rating agencies specialized in microfinance social ratings (Microrate, Microfinanza, and Planet rating) and second, from the Microfinance Information eXchange database.

Variable	Definition	Whole Sample				Rated MFIs			Non-Rated MFIs
		Number of Observations	Mean	Standard Deviation	Number of Observations	Mean	Standard Deviation	Number of Observations	Mean	Standard Deviation
Socially rated	Dummy taking the value 1 if the MFI is rated, and 0 if the MFI is unrated.	767	0.356	0.479	273	1	0	494	0	0
Young	4 to 8 years.	765	0.067	0.250	271	0.066	0.249	494	0.067	0.250
Mature	More than 8 years.	765	0.924	0.265	271	0.919	0.274	494	0.927	0.260
Subsidized debt	Difference between financial expenses on borrowing and the country lending market rate. Dummy taking the value 1 if the difference is negative (subsidization, and 0 otherwise).	767	0.897	0.304	273	0.850	0.358	494	0.923	0.267
Profit status	Dummy taking the value 1 if the MFI is a commercial-oriented MFI, and 0 if the MFI is a social-oriented (pro-social) MFI.	757	0.680	0.467	270	0.648	0.478	487	0.698	0.460
Country rule of law index	The rule of law index captures perceptions of the extent to which agents have confidence in and abide by the rules of society, in particular: the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence, better regulatory environment or that respect for the rule of law leads to fairer processes for selecting and replacing governments and less abuse of public office for private gain [51]. The index runs between −2.5 and +2.5 with higher indexes being considered better.	755	−0.607	0.391	261	−0.579	0.401	494	−0.622	0.385
Profitability: ROA)	Net operating income divided by average total assets.	723	0.031	0.044	255	0.031	0.050	468	0.031	0.041
Yield on loan portfolio	(Yield on Gross Portfolio (nominal)—Inflation Rate)/(1 + Inflation Rate).	729	0.242	0.124	260	0.289	0.138	469	0.215	0.108
Riskiness of the loan portfolio: Portfolio at risk at 30 days	Outstanding loan balance due above 30 days divided by gross loan portfolio.	709	0.040	0.043	255	0.042	0.040	454	0.039	0.045
Depth of outreach	Average loan balance per borrower/GNI per capita.	738	0.529	0.569	265	0.349	0.433	473	0.630	0.611
Breadth of outreach	Number of female borrowers divided by the total number of active borrowers.	660	0.668	0.222	234	0.662	0.224	426	0.671	0.221
Gross loan portfolio	Measure of the diversification of MFI asset: Gross loan portfolio/total assets.	736	0.811	0.095	264	0.816	0.092	472	0.809	0.096
Deposits to Assets ratio	Total deposits/assets.	723	0.221	0.283	258	0.155	0.245	465	0.258	0.296
GDP Growth	Annual growth rate of real gross domestic product.	675	0.049	0.031	246	0.041	0.033	429	0.054	0.029

shows that about 36% of all MFIs in the sample choose to be rated, 92% of them are mature, 90% get subsidized debt, and 68% are for-profit organizations.

In terms of social performance, MFIs have about 67% of female borrowers. This proportion is evenly distributed among rated and non-rated MFIs in our sample. The average loan balance per borrower represents on average 53% of gross national income per capita in our sample.

The correlation table (

Table 4. Correlation matrix. This table reports on the correlations between all the variables included in the estimations. The data comes from two sources: first, the websites of rating agencies specialized in microfinance social ratings (Microrate, Microfinanza, and Planet rating) and second, from the Microfinance Information eXchange database.

	Socially Rated	Young	Mature	Subsidized Debt	Profit Status	Country Rule of Law Index	Profitability (ROA)	Yield on Loan Portfolio	Portfolio at Risk at 30 Days	Depth of Outreach	Breadth of Outreach	Gross Loan Portfolio	Deposit to Asset	GDP Growth
Socially rated	1
Young	−0.001	1
Mature	−0.015	−0.933	1
Subsidized debt	−0.115	−0.116	0.130	1
Profit status	−0.051	0.079	−0.098	0.075	1
Country rule of law index	0.052	−0.027	0.003	−0.134	−0.275	1
Profitability (ROA)	−0.007	−0.048	0.090	0.092	−0.082	−0.008	1
Yield on loan portfolio	0.285	0.059	−0.071	−0.161	−0.065	0.132	0.051	1
Portfolio at risk at 30 days	0.034	−0.057	0.073	−0.122	−0.029	0.114	−0.422	−0.015	1
Depth of outreach	−0.237	−0.077	0.074	0.007	0.141	−0.356	−0.059	−0.420	−0.015	1
Breadth of outreach	−0.019	0.088	−0.114	−0.113	−0.095	0.043	0.041	0.274	−0.184	−0.401	1
Gross loan portfolio	0.033	0.050	−0.005	0.084	0.146	−0.022	0.184	−0.136	−0.158	−0.068	−0.028	1
Deposit to asset	−0.174	−0.144	0.144	−0.037	0.177	−0.125	−0.144	−0.309	0.152	0.497	−0.273	−0.246	1
GDP growth	−0.187	0.083	−0.092	0.052	0.031	−0.067	0.122	−0.114	−0.241	0.100	0.074	−0.004	0.074	1

) shows some relatively low correlations between all variables. Some higher absolute levels of correlations are between ROA and the riskiness of the loan portfolio, −42% Depth of outreach and Yield on loan portfolio at −42%, and Depth of outreach and Breadth of outreach at −40%.

4.2. Main Results

We implement several versions of the Probit pooled regression in order to determine whether institutional and resource-dependence factors impact the social rating decisions of MFIs.

Table 5. Baseline model—Institutional determinants of the social rating decision. The table shows pooled Probit estimations of the effect of MFI characteristics on the social rating decision. The data comes from two sources: first, the websites of rating agencies specialized in microfinance social ratings (Microrate, Microfinanza, and Planet rating) and second, from the Microfinance Information eXchange database.

Regression	(1)	(2)	(3)	(4)
Constant	0.1696	1.5469	1.1338	3.0403 ***
	(0.9061)	(1.0204)	(0.9152)	(1.0339)
Young	−0.8373	−1.2896 *	−0.8650	−1.2764 **
	(0.6524)	(0.7380)	(0.5518)	(0.5893)
Mature	−0.6306	−1.2108 *	−1.0580 **	−1.7228 ***
	(0.6131)	(0.6657)	(0.5134)	(0.5220)
Subsidized debt	−0.6596 ***	−0.7656 ***	−0.4701 **	−0.6726 ***
	(0.1944)	(0.2182)	(0.1974)	(0.2201)
Profit status	−0.1105	−0.3180 **	−0.1001	−0.2575
	(0.1344)	(0.1556)	(0.1466)	(0.1653)
Country rule of law index	−0.3278 **	−0.0468	−0.4200 **	−0.1022
	(0.1640)	(0.1793)	(0.1765)	(0.1882)
Profitability (ROA)	0.3033	0.8042	0.7167	1.2240
	(1.4257)	(1.5193)	(1.5629)	(1.6957)
Yield on gross loan portfolio	2.6528 ***	0.9459	2.5293 ***	0.2809
	(0.5833)	(0.7175)	(0.6166)	(0.7670)
Risk of the gross loan portfolio	0.1897	0.2318	0.6645	0.2658
	(1.6125)	(1.6402)	(1.7851)	(1.9047)
Depth of outreach	−0.7311 ***	−0.7496 ***	−0.9257 ***	−1.0051 ***
	(0.1948)	(0.2192)	(0.2121)	(0.2389)
Breadth of outreach	−1.1325 ***	−0.2364	−1.1746 ***	−0.2157
	(0.3333)	(0.4166)	(0.3683)	(0.4651)
Gross loan portfolio	1.4154 **	0.9059	1.0003	0.5410
	(0.6964)	(0.7315)	(0.7636)	(0.8107)
Deposits to asset	0.1544	−0.1710	−0.1579	−0.5869 *
	(0.2818)	(0.3062)	(0.3061)	(0.3422)
GDP Growth	−5.4255 ***	−4.2863 **	0.7665	1.6762
	(2.0189)	(2.0155)	(2.5171)	(2.4863)
Year-fixed effects	no	No	yes	yes
Regions-fixed effects	no	Yes	no	yes
Pseudo R2	0.1480	0.2026	0.2356	0.2973
Log pseudo-likelihood	−297.89	−278.8	−238.45	−219.21
Observations	543	543	476	476

Robust standard errors are in parentheses and are adjusted for heteroscedasticity; *** p < 0.01, ** p < 0.05, * p < 0.1.

features four representative models. Models 1 to 4 include our key independent variables, taking aspects of institutional coercive pressure and Resource Dependence Theory into account. Model 1 is the baseline. Model 2 includes regional dummies to account for regional specificities. Model 3 includes year dummies, and model 4 includes both regional and year dummies. Table 4 replicates these same specifications, but focuses on marginal effects.

Regarding our hypothesis H1, Table 5 indicates that an MFI’s decision to get a rating is impacted by the rule of law in their particular country. This result agrees with the literature that shows how the existence of a common legal environment affects many aspects of an organization’s behavior and structure. It is because the legal environment requires setting up organizational controls in order to honor legal commitments.

Hypothesis H1 is confirmed when looking at Models 1 and 3. The coefficients on the Country rule of law index variable have the expected (negative) sign indicating that as the rule of law tightens, a lower proportion of MFIs tend to seek a rating. Furthermore, these coefficients are significant at the 95% confidence level. Nevertheless, the results for Models 2 and 4 are not conclusive, as the coefficients have the correct sign, but are not statistically significant. One possible reason for this latter result is that, in these regressions, regional dummies are absorbing the effect of the Country rule of law index variable, as this measure keenly depends on the region of the world under consideration.

When moving to hypothesis H2, we get a validation of our resource dependence hypothesis along certain dimensions. When resource dependency is captured by Subsidized debt, we get a strong result across the board. The coefficient on this variable is significant at the 99% level and has a negative sign in all four regressions (except for Model 3, in which the significance level is at 95%). This is as expected because the easier the borrowing terms are for these MFIs, the less inclined they are to seek a social rating. It can also be argued that subsidized MFIs are contractually subjected to more scrutiny by lenders, and thus do not need to be rated as much.

Models 1 and 3 show that the Maturity and the Profit status of MFIs do not seem to have explanatory power, except in the case of mature MFIs (Models 2, 3, and 4). In that case, the coefficient on that variable is negative and significant at least at the 95% level (Models 3 and 4). This latter finding aligns with RDT as mature MFIs do not necessarily need to obtain a rating to satisfy their funding needs.

Models 2 and 4, globally, show the same results as the other two models, except that now Profit status matters (Model 2) in the sense that not-for-profit MFIs, which have harder access to financial markets, will seek to obtain a rating more often than for-profits. Not-for-profit MFIs generally depend on donors, governments, and other NGOs to gain more credibility with private and public funders, which provide them with funds that are usually considered to be scarce in the market [31].

The variable Young (maturity) also becomes significant at the 95% level in Model 4, and again, relative to the New (maturity) MFI-omitted category, we find that these older institutions will not resort to seeking a social rating, as they do not suffer from lack of funding access as much as newer institutions. In other words, the younger the MFIs, the more likely they will choose to get a social rating.

In conclusion, our strongest result in support of hypothesis H2 is for the Subsidized debt dimension (in all regressions). This indicates that when there is a direct proxy for financial constraints (credit policy) and thus resource dependency, a relationship is revealed between MFIs benefiting from these subsidies and their decision to obtain a rating.

It is worth noting that, regarding our control variables, we also get some interesting findings. Focusing on MFIs’ social performance variable, we find that the Depth of outreach variable is significant at the 99% level and has a negative coefficient. This result indicates that when institutions decrease the size of their loans (lending to poorer people) they do seek to be rated. This makes intuitive sense if we consider that getting a rating is used by MFIs to convey information to investors that would not necessarily be available to them by other means.

We also find that the Breadth of outreach (% of female borrowers) is significant at the 99% level in Models 1 and 3 and has a negative coefficient. The higher the proportion of female borrowers, the less MFIs seek to obtain a rating. This particular finding may be explained by the fact that in contrast with the poverty-alleviation mission, which needs to be clearly signaled to investors, female financial inclusiveness might be easier to document outside of ratings. Another possible explanation is that the proportion of female borrowing does not account for the size of personal loans relative to their income. They may be richer or poorer borrowers on average relative to males. We do not know the depth of outreach for females in our sample. The size of loans may be smaller on average for females, as documented by Brana [53], which, as we show here, tends to have a negative effect on the ratings decision.

Regarding secondary control variables, we find that the probability to get a rating increases with the real yield on the gross loan portfolio and with the proportion of loans in the total assets. The former finding points to the fact that obtaining a social rating has a cost and that it is MFIs that have sufficiently high earnings which can do that. The latter finding can be explained by the fact that when MFIs are performing their mission of reaching high volumes of loans, they seek more funding and thus want to obtain a rating. This result is in line with our hypothesis H2 and does not contradict the fact that subsidized institutions (via the interest rates route) choose less often to be rated. Receiving a subsidy lowers the incentive to seek a rating. Moreover, subsidized MFIs are already under close scrutiny by subsidy providers. But at the same time, when their overall loan portfolio size grows, MFIs applying an asset-liability matching logic end up with a larger dependence on external financing.

The search for capital ultimately has to satisfy the informational needs of investors. Our hypotheses H1 and H2 inform us that to attract a significant amount of private investment, the microfinance sector in less-developed countries works to reduce the information asymmetry their potential investors are subjected to. This means that MFIs must submit themselves to the most credible and widely accepted audits, ratings, and supervision available [56].

4.3. Marginal Effects

Table 6. Marginal effects—Institutional determinants of the social rating decision. The table shows marginal effects for our pooled Probit estimations of the impact of MFI characteristics on the social rating decision. The data comes from two sources: first, the websites of rating agencies specialized in microfinance social ratings (Microrate, Microfinanza, and Planet rating) and second, from the Microfinance Information eXchange database.

Regression	(1)	(2)	(3)	(4)
Young	−0.2992	−0.4499 *	−0.3053	−0.4349 **
	(0.2330)	(0.2558)	(0.1939)	(0.1989)
Mature	−0.2253	−0.4224 *	−0.3734 **	−0.5869 ***
	(0.2189)	(0.2309)	(0.1795)	(0.1752)
Subsidized debt	−0.2357 ***	−0.2671 ***	−0.1659 **	−0.2291 ***
	(0.0693)	(0.0753)	(0.0694)	(0.0743)
Profit status	−0.0395	−0.1109 **	−0.0353	−0.0877
	(0.0480)	(0.0541)	(0.0516)	(0.0563)
Country rule of law index	−0.1171 **	−0.0163	−0.1482 **	−0.0348
	(0.0583)	(0.0625)	(0.0618)	(0.0641)
Profitability (ROA)	0.1084	0.2806	0.2529	0.4170
	(0.5093)	(0.5296)	(0.5517)	(0.5771)
Yield on gross loan portfolio	0.9479 ***	0.3300	0.8925 ***	0.0957
	(0.2092)	(0.2523)	(0.2140)	(0.2614)
Risk of the gross loan portfolio	0.0678	0.0809	0.2345	0.0906
	(0.5762)	(0.5721)	(0.6302)	(0.6489)
Depth of outreach	−0.2612 ***	−0.2615 ***	−0.3267 ***	−0.3424 ***
	(0.0682)	(0.0742)	(0.0729)	(0.0781)
Breadth of outreach	−0.4047 ***	−0.0825	−0.4145 ***	−0.0735
	(0.1188)	(0.1453)	(0.1291)	(0.1582)
Gross loan portfolio	0.5057 **	0.3160	0.3530	0.1843
	(0.2485)	(0.2559)	(0.2674)	(0.2759)
Deposits to asset	0.0552	−0.0597	−0.0557	−0.2000 *
	(0.1007)	(0.1068)	(0.1081)	(0.1169)
GDP Growth	−1.9385 ***	−1.4954 **	0.2705	0.5711
	(0.7195)	(0.7036)	(0.8891)	(0.8491)
Observations	543	543	476	476

Robust standard errors are in parentheses and are adjusted for heteroscedasticity; *** p < 0.01, ** p < 0.05, * p < 0.1.

showcases the marginal effects of our four Probit regressions. Marginal effects measure the change in outcome as a function of the change in the treatment (or independent variable of interest) holding all other variables in the model constant. These marginal effects reflect the economic importance of the interaction between changes in explanatory factors and the decision of obtaining a rating.

The results shown in Table 6 replicate nearly exactly what was found in Table 5 in terms of significance levels and expected signs of these relationships. An exception is the maturity variable Young, for which the significance levels change slightly in the four models. Focusing on the interpretable findings: being a subsidized MFI reduces the probability of getting a rating by between 17% (Model 3) and 27% (Model 4). An increase of the Country rule of law index by 1 unit (representing a 20% increase in the index) decreases the probability of getting a rating by between 12% (Model 1) and 15% (Model 3). Being a for-profit MFI reduces the decision of getting a rating by 11% (Model 2). Being more mature decreases the probability of making the decision to be rated by about 44% on average (Young and Mature categories).

Regarding the control variables of interest (social performance), we find that as the Depth of outreach increases by 1%, the probability of choosing to be rated decreases on average by about 30% across the four regressions. As the proportion of female borrowers increases by 1%, the probability of getting rated decreases by 0.4% on average (Models 1 and 3).

5. Discussion

Microfinance institutions engage in a wide variety of peripheral activities to provide the financial support necessary for the continued pursuit of their societal missions. Whether through attracting private contributions, obtaining government grants and contracts, or involvement in commercial activities, these organizations must dedicate substantial resources and attention toward revenue acquisition. Each revenue strategy has its appeal and current niches of opportunity but also carries constraints and pressures [57]. These pressures will necessarily compel an organization to act in a particular way, one of which is by going in for a social rating in order to enable funders the ability to determine how they are performing socially.

According to Fernando [58], many people consider NGOs as the most suitable institutions to reach the poor and provide microcredit. Research has shown that most MFIs that are highly commercialized do not perform their role in providing for the very poor since they are profit-seeking to the detriment of poorer customers. Much of this ethical debate started with the Compartamos’ stock offering in 2007, which, in part, was made possible by lending rates in excess of 100% [4].

In this context, we ask how does the MFI’s environment shape their choice to obtain a social rating? Especially in the context of less-developed regions of the world. We examine here whether both coercive pressures and the MFI’s dependence on resources affect this decision. In that respect, we identified two main hypotheses: H1 about the effects of a country’s rule of law, and H2 about the effect of resource dependency on that decision. Globally, our empirical analysis lends good support for both our hypotheses H1 and H2.

We found that government regulations, and, more generally, the rule of law, play a role in determining the decision of an MFI to get a social rating. The stronger the regulations, the less MFIs will be interested in getting rated in less-developed countries. When MFIs are regulated by banking authorities such as COBAC in Central Africa, the existence of such regulatory authorities ensures a legal framework within which these institutions are supposed to act.

Our study also contributes to the literature by analyzing the impact of resource dependence on social rating disclosure. Our results indicate that MFIs in less-developed countries that are subsidized, for-profit, and/or mature will not seek to get a social rating, as compared to others that are more resource-dependent. Many NGOs rely heavily on external funding sources, but as they mature, they have easier access to capital markets and thus are able to thrive.

The conduct of a microfinance institution is bound by the pressure to ensure legitimacy and financial or material support [35]. Young MFIs that are generally not profitable are largely dependent on external sources for survival. They dedicate substantial resources and attention toward revenue acquisition. Each revenue strategy has its appeal and current niches of opportunity but also carries constraints and pressures that may impinge on the autonomy of the organization [57].

The institutions on which they depend for funding do exercise a lot of power over them. One funders’ requirement is to have access to social ratings so that they can know how well the organization they are investing in is fulfilling its social mission.

To sum up, our study has theoretical and practical implications. Theoretically, our study reveals some of the determinants (institutional and resource-based) of MFIs’ choices to seek a social rating. In particular, we have discovered here that in less-developed countries with stronger rules of law, MFIs that are more subsidized, more mature (for-profit) organizations, providing more female-oriented loans will not seek to be rated as often as others. Our study thus highlights the role that funding constraints play in raising the MFIs’ need for legitimacy.

This knowledge may have important policy and managerial implications regarding how microfinance institutions can continue to serve their noble function to facilitate financial inclusion [8].

Should there be a policy goal to increase the ratings coverage of MFIs in these countries? Given our results, we found that this might not be necessary for a relatively strong and regulated environment when there is pressure to abide by laws that enforce social missions. There are also alternate monitoring mechanisms included in subsidy agreements that fulfill a similar purpose as social ratings. These direct interest rate subsidies are often part of the quick-fix toolbox of governments and central banks.

Of course, government policies must be accompanied by a clear path to sustainable finance, and, in particular, donor agencies should maintain this focus in their policy dialogues and identify subsidies that are fiscally sustainable. However, if MFIs located in less-developed countries want to expand their funding by tapping into a new class of socially responsible investors, our results indicate that they benefit from obtaining these ratings. In that respect, our findings may help policy-makers and regulators design incentives for MFIs to be socially rated and at the same time avoid mission-drift [1,4,5,6].

6. Conclusions

In this article, we have examined the institutional determinants of MFIs’ social rating decisions in a sample of six major less-developed regions. Our study contributes to the CSR literature in microfinance by documenting the important role that institutional factors play in their choice of obtaining a rating.

Our first finding is that a country’s rule of law does determine MFIs’ decision to go for a rating. Generally, regulatory norms exert pressure on corporations by constantly inspecting and providing sanctions. For less-developed countries, this translates into the notion that social ratings function as a substitute for the lack of a strong rule of law. That is, MFIs signal their good CSR behavior when located in an environment where institutions are not “forced” to abide by these behaviors. In line with Institutional Theory [12,31], our research shows that the tighter the rule of law the less inclined MFIs are to seek a rating.

On the other hand, Resource Dependence Theory [35] also applies to microfinance institutions, as many of these are bound by the demands to ensure legitimacy in order to get financial or material support. RDT helps us demonstrate that proxies for resource (in)dependence such as for-profit status, mature stage, and subsidization are negatively associated with the obtention of a social rating. This implies that in less-developed countries, when MFIs are in a strong position with respect to investors, and they need new funds, they are less likely to want a social rating.

Our study is relevant today because in the domain of microfinance the institutional framework has been neglected when looking at the determinants of social rating in microfinance. So far, a lot of attention has been focused on the determinants of rating scores [24], while ignoring what actually prompts these institutions to decide to be rated. The study emphasizes the fact that social rating practices are more likely to be adopted as a result of several environmental factors.

Our study focused only on coercive isomorphism [36] and left out other institutional determinants of social rating decisions such as mimetic pressures and normative pressures as a result of a lack of available data concerning those specific pressures. This is a limitation that can be addressed in future research by undertaking to compile such data.

In conclusion, our study suggests that the environment not only creates pressure for MFIs to decide to obtain a social rating but that the latter can be analyzed as a means to safeguard the flow of resources to these organizations. When microfinance institutions in less-developed regions are convinced that funds are linked with or depend on social rating disclosure, they will make the necessary efforts to ensure that they eventually comply with CSR standards.