cost-efficiency and outreach of microfinance institutions: trade-offs and the role of ownership

FIELD REPORT

COST-EFFICIENCY AND OUTREACH OFMICROFINANCE INSTITUTIONS: TRADE-OFFS AND THE ROLE OF OWNERSHIP

GASHAW TADESSE ABATE1*, CARLO BORZAGA2 and KINDIE GETNET3

1University of Trento, Graduate School of Social Sciences and European Research Institute onCooperatives and Social Enterprises (EURICSE), Trento, Italy

2University of Trento, Department of Economics and European Research Institute on Cooperativesand Social Enterprises (EURICSE), Trento, Italy

3International Water Management Institute (IWMI), Addis Ababa, Ethiopia

Abstract: The focus on achieving financial efficiency by microfinance institutions in recent yearsraises a natural concern on their social outcome, outreach to the poor. Using a stochastic frontierapproach on sample microfinance providers in Ethiopia, this paper analysed the effect of anincreasingly important efficiency requirement on the traditional social mission of microfinance. Italso addressed whether the way ownership is organised and practiced affects the cost of microfinancedelivery. The result indicates a trade-off between the outreach to the poor and cost-efficiency,suggesting the difficulty in trying to achieve the two goals simultaneously. Financial cooperativesare better in their cost containment compared with specialised microfinance institutions owned byshareholders. Copyright © 2013 John Wiley & Sons, Ltd.

Keywords: microfinance; financial cooperatives; cost-efficiency; outreach; trade-off; Ethiopia

1 INTRODUCTION

Providing financial services tailored to the needs of small borrowers is a high costbusiness, as it requires considerable monitoring and enforcement costs. Because of thecosts it entails, conventional banks in most developing countries often systematicallyexclude small borrowers from accessing their financial services. Recent studies indicatethat only 24 per cent of adults in Sub-Saharan Africa and 14 per cent of adults in Ethiopia

*Correspondence to: Gashaw Tadesse Abate, EURICSE, via S. Giovanni 36–38122 Trento (TN), Italy.E-mail: [email protected]

Copyright © 2013 John Wiley & Sons, Ltd.

Journal of International DevelopmentJ. Int. Dev. 26, 923–932 (2014)Published online 21 November 2013 in Wiley Online Library(wileyonlinelibrary.com) DOI: 10.1002/jid.2981

use formal financial institutions (Amha & Peck, 2010; Demirgüç-Kunt and Klapper, 2012).Microfinance emerged as institutional innovation to lend small borrowers throughovercoming prevailing costs of market contracts in credit markets of these low-incomecommunities. Although their innovative loan terms and lending practices enable them tosecure unusual high repayment rates in lending the poor, translating high repayment rates intoprofit and perpetuating financial services to the poor on cost covering basis have been aremaining challenge for most microfinance providers. For instance, during 2000, despitethe 94 per cent repayment rates, the majority of the microfinance institutions in Ethiopia werenot financially self-sufficient (Amha, 2000).Recently, however, the growing commercialization and competition coupled with

withdrawal of subsidies standout the need for financial sustainability and efficiency inthe microfinance industry. These developments, in turn, result in a shift of focus fromoutreach per se to outreach and financial self-sufficiency. Although it can have long-termimperatives in sustaining financial services to the poor, striving to achieve financial self-sufficiency can have a short-term implication on the traditional social mission ofmicrofinance institutions—outreach to the poor—and ways of doing business.Theoretically, achieving financial viability together with serving the poor can be either

conflicting or complementary. On the one side, the two bottom lines can be in harmonyif the impositions of financial sustainability requirements improve efficiency in resourceallocations and attract commercial funds (including voluntary deposits) that can be usedto expand the outreach (Rosengard, 2004; Frank, 2008). On the other side, the pursuit offinancial sustainability can crowed out small size loans that are demanded by the poor,as they are costly to service (Hulme & Mosley, 1996; Weiss & Montgomery, 2005).Despite the growing concerns on the impositions of pursuing financial efficiency on

serving the poor, systematic empirical analysis on the outreach-financial sustainabilitytrade-off is limited, and the evidence emerged from the existing few works is mixed.Studies by McIntosh et al. (2005) and Hermes et al. (2011) found tension between servingthe poor and achieving efficiency. Their results indicated that unbanked wealthier clientsbenefit from the strife for financial sustainability. A global analysis of micro lenders byCull et al. (2007) also corroborates the presence of potential trade-off between the outreachand the financial performance. In contrast, another recent study by Quayes (2012)documented a complementary relationship between serving the poor and improvedfinancial performance in microfinance.Building on prior works, this paper aims to understand whether and to what extent the

focus on financial performance affects the outreach of microfinance providers in theEthiopian context. The information would be useful for policy making purpose, asmicrofinance institutions in Ethiopia are strongly promoted to expand financial servicesto the unbanked poor. The study also compares the cost-efficiency (containment levels)between specialised microfinance institutions that are shareholder firms and financialcooperatives, the two dominate microfinance providers in Ethiopia, with the purpose ofunderstanding the effect of ownership form on costs of microfinance delivery.

2 THE MICROFINANCE LANDSCAPE IN ETHIOPIA

In Ethiopia, conventional financial institutions are not only unwilling but also lack thecapacity to serve the needs of the poor (Amha, 2007). Financial services to the poor arelargely delivered by the microfinance industry, which is mainly made up of services

924 G .T. Abate et al.

Copyright © 2013 John Wiley & Sons, Ltd. J. Int. Dev. 26, 923–932 (2014)DOI: 10.1002/jid

rendered by financial cooperatives, non-governmental organisations (NGOs), non-bankmicrofinance institutions (NBFIs) and informal lenders. Financial cooperatives are theforerunners in delivering financial services for the poor and are notable both in lendingsmall uncollateralized loans, saving mobilisation and in inculcating the importance offinancial services in the society at large (Degefe & Nega, 2000).Currently, as shown in Figure 1, there are about 42 saving and credit unions and over

7000 primary saving and credit cooperatives providing microfinance services (i.e. saving,loan and insurance) for about one million members in the country (Federal CooperativeAgency, 2012). Similar to most credit cooperatives elsewhere, financial cooperatives inEthiopia are organised by individuals working or living in the same localities. They mainlyuse standard bilateral lending contracts between the cooperative and a member borrower.Liability for repaying the loan rests with the individual borrower and the co-signer, whois also a member of the same cooperative.Besides the role played by financial cooperatives, the development of microfinance in

Ethiopia also counts on efforts made by the international NGOs and government creditprogrammes that integrate credit services in their development and relief schemes. Still,today, there are NGOs and government relief programmes such as Productive SafetyNet1 and Other Food Security Programs that are playing considerable roles in terms ofstimulating the microfinance market. The involvements of the government and NGOs incredit delivery have been encouraging in terms of stimulating credit demand and povertyreduction (Gilligan et al., 2008). Nevertheless, poor financial discipline and distortedresource allocation by NGOs and government credit programmes have been equally

1The productive Safety Net is a programme that transfers cash to the food insecure population in chronically foodinsecure districts, and it aims at bridging the food gap and preventing asset depletion. The programme operates asa safety net, targeting cash transfers to poor households in two ways—through public works and direct support.

Figure 1. Number of microfinance and volume of loan outstanding by ownership form

Cost-Efficiency and Outreach of Microfinance Institutions 925


substantial. Interest rate subsidies, debt write-off and equating loans with humanitarianassistances by the NGOs were among the distortions that have indoctrinated a bad creditculture—a culture of entitlement—that undermines the development of micro-creditmarkets in Ethiopia (Degefe & Nega, 2000; Amha, 2007).Following the economic reform implemented in Ethiopia since 1991, some of the NGOs

and government pilot credit programmes engaged in financial intermediation transformedinto formal (specialised or non-bank) microfinance institutions. The transformation wasmade mainly to reverse the bad credit culture instituted by NGOs and state creditprogrammes. This was performed through establishing microfinance institutions thatadhere to the market mechanism while serving the poor. Hence, the NBFIs that evolvedfrom NGOs and government credit programmes are one among the major microfinanceplayers in Ethiopia. In addition to the NBFIs that evolved from prior NGOs andgovernment credit programmes, the industry also witnessed new start-ups of investor-owned microfinance providers of the same nature. As of 2011, a total of 30 NBFIs inEthiopia reported reaching over 2.3 million clients, with total loans outstanding of Birr6.5 billion2 (Figure 1).

3 DATASET AND EMPIRICAL APPROACH

We used primary data collected from 107 microfinance providers in Ethiopia (30 NBFIsand 77 financial cooperatives). The data were collected between April and June 2012.The cost-efficiency of microfinance institutions is measured using stochastic cost frontierapproach. Cost-efficiency is measured in terms of how close a microfinance cost lies tothe efficient cost frontier for a given technology. The efficient frontier is determined bytwo conditions: minimum use of inputs (technical efficiency) and optimal mix of inputs(allocative efficiency), Battese and Coelli (1995) and Kumbhakar and Lovell (2000). Weused the one-step stochastic cost frontier approach proposed by Battese and Coelli(1995), which estimates the cost frontier and inefficiency correlates simultaneously(Equation (1)). The specification of the cost function estimated is given as follows:

ln TCið Þ ¼ αþ β1 ln Salaryið Þ þ β2 ln IntExpið Þ þ β3 ln LLPið Þ þ β4 ln GLPið Þþβ5 ln LLR_GLPið Þ þ β6 ln Depreciationið Þ þ ui þ vi

(1)

where TCi is the sum of interest and operating expense of microfinance institution i; Salaryis the price of a unit of labour for the period, which is the average salary per unit of labourper annum; IntExp is the interest expense faced by a microfinance per unit of borrowingand deposit held; LLP is the loan loss provision expense for the period; GLP is the grossloan portfolio; LLR_GLP is the loan loss provision over gross loan portfolio, whichmeasures an microfinance risk taking strategies; and Depreciation is the financial lossesas a result of obsolescence of physical capital.After estimating the cost-efficiency level of each microfinance institutions, the one-step

approach analysed the correlates of cost-inefficiency consecutively. The inefficiencycomponent of the error term denoted by μi in Equation (2) is specified as a function of aset of outreach and microfinance specific explanatory variables in order to understandthe trade-off between outreach to the poor and cost-efficiency. We also introduced an

2Birr is Ethiopian currency unit. In June 2012, Birr 1 was officially exchanged for US$0.0557.



organisational dummy to understand the effect of ownership form (NBFIs vs financialcooperatives) on cost containment. The specification of the inefficiency function estimatedis given as follows:

μi ¼ αþ δ1LoanSizeþ δ2Womanþ δ3OrgFormþ δ4Ageþ δ5Size (2)

where μi represents the inefficiency component of an MFI i, as defined earlier. LoanSize isone of the generally accepted measures of outreach omnipresent in the microfinanceliterature. It is the ratio of the total loan outstanding and the total number of activeborrowers. The lower is the ratio, the greater is the depth of outreach and vice versa.Women is another accepted measure of outreach, which measures the proportion offemale active borrowers. Higher percentage of women borrowers indicates greaterdepth of outreach. OrgForm denotes the organisational form or ownership dummy(i.e. 1 if the microfinance is financial cooperatives and 0 otherwise). The propositionhere is that the inefficiency of microfinance may depend on the form of ownership, assome ownership forms are more effective in reducing or internalising costs of marketcontracts. Besides outreach variables and organisational dummy, we introduced Ageand Size, as they are also the major drivers of operational expense in microfinanceprovisions (Gonzalez, 2007). Age controls for the effect of experience and learningon cost-efficiency. Size is measured in the total assets of microfinance and accountsfor scales of operation.

4 RESULTS AND DISCUSSION

4.1 Outreach and Cost-Efficiency: Is There a Trade-off?

As shown in Panel B of Table 1, the study results show the presence of trade-off betweencost-efficiency and outreach to the poor (measured by loan size and proportion of womenborrowers). The estimated coefficient for average loan size is negative, even aftercontrolling for organisational form, experience and scale of operation. This signifies thatmicrofinance institutions with higher average loan size are more cost-efficient thanmicrofinance institutions with lower average loan size. The results from the specificationin columns (2) and (4) for the proportion of women borrowers in the loan portfolio havepositive coefficients, indicating that microfinance providers catering more to womenborrowers are less cost-efficient.Besides the evidences that emerge from inefficiency correlates, simple Ordinary Least

Squares (OLS) regression results (unreported) that correlate estimated cost-efficiencyscores with outreach indicators (i.e. average loan size and proportion of women borrowers)also substantiate the tension between serving the poor and cost-efficiency. It is found thatthere is a positive relationship between cost-efficiency scores and average loan size and anegative relation with the proportion of women borrowers. This result further marked thatserving the poor and more to women borrowers is not in harmony with the pursuit ofachieving financial efficiency. Over all, the results that emerged from the analysis correspondto the findings of prior studies by Cull et al. (2007) and Hermes et al. (2011) and areconsistent with the general theoretical predictions that claim the presence of cost differentialbetween serving the poor and the less poor or unbanked wealthier clients (Armendáriz deAghion & Szafarz, 2009). The difference in costliness of lending small size loans, relative



Table

1.Maxim

umlik

elihoodestim

ationof

theparametersforstochastic

cost-efficiency

frontierandcorrelates

ofinefficiency

Costfunctio

nDependent

variable:totalcostin

Birr

Panel

A:inputandoutput

variables

(1)

(2)

(3)

(4)

(5)

ln(Salary(per

personnel))

0.272(3.51)***

0.323(4.12)***

0.248(3.06)***

0.223(3.16)***

0.201(2.86)***

ln(Interestexpenses

(per

unitof

deposit))

0.236(5.28)***

0.220(5.25)***

0.236(5.27)***

0.272(6.37)***

0.278(6.65)***

ln(G

ross

loan

portfolio

)0.835(18.79)***

0.760(19.72)***

0.793(19.18)***

0.839(21.61)***

0.856(23.66)***

ln(Loanloss

reserveover

grossloan

portfolio

)�0

.006

(0.23)

�0.011

(0.40)

�0.002

(0.07)

0.001(0.05)

0.004(0.18)

ln(Loanloss

provisionexpenses)

0.046(2.76)***

0.070(4.63)***

0.055(3.15)***

0.050(3.12)***

0.044(3.05)***

ln(D

epreciation)

0.022(1.00)

0.029(1.33)

0.023(1.02)

0.007(0.36)

0.020(1.13)

Constant

�2.365

(3.61)***

�1.967

(3.07)***

�1.621

(2.31)**

�1.438

(2.37)**

�1.507

(2.62)***

Panel

B:cost-inefficiency

correlates

Average

loan

size

�0.001

(1.67)*

�0.003

(2.86)***

�0.004

(2.25)**

%of

wom

enborrow

ers

1.896(2.44)**

4.802(1.68)*

1.129(0.30)

Financial

cooperatives

�1.034

(1.91)*

�0.865

(0.87)

�4.770

(2.09)**

Age

oftheinstitu

tion

�0.018

(0.13)

Sizeof

theinstitu

tion(intotalassets)

�2.099

(2.33)**

Constant

�0.237

(0.74)

�1.702

(2.84)***

�0.082

(0.22)

0.360(0.20)

8.791(2.24)**

Num

berof

observation

107

107

107

107

107

WaldChi-squared

test

3059.29

3047.94

1972.25

2697.22

3038.35

Prob>chi-squaredtest

0.000

0.000

0.000

0.000

0.000

Log

likelihoodfunctio

n�7

8.511

�76.976

�78.446

�70.672

�67.614

Source:

Authors’calculations

basedon

prim

arydata

collected

betweenAprilandJune

2012.

*Significant

at10%.

**Significant

at5%

.***S

ignificant

at1%

.



to larger loans, can be even more significant in low-income communities. For instance, incountries such as Ethiopia, such a trade-off between cost-efficiency and outreach to thepoor can easily arise because of additional costs linked with the difficulty of accessingthe poor rural clients and monitoring and following up efforts required from microfinanceinstitutions to deal with less educated borrowers.

4.2 Does Ownership Affect the Level of Microfinance Cost-efficiency?

Another issue investigated in this paper is the difference in cost-efficiency amongmicrofinance institutions by ownership form. The question is whether or not the wayownership is organised and practiced has implication on levels of cost-efficiency inmicrofinance. According to the agency theory, organisations that are owned by agents withpecuniary incentives are more able to reduce agency costs (Jensen & Meckling, 1976).Moreover, when it comes to microfinance, organisations that are owned by their customersare more able to reduce costs of market contracts (Hansmann, 1996). Although pecuniaryincentives can be at a play for both ownership forms considered in this study, if thisconjectures are true, financial cooperatives should be cost-efficient compared withspecialised microfinance institutions, as they possess a better position to overcome costsof market contracts.Consistent with theoretical predictions, the predicted cost-efficiency levels show that on

average, financial cooperatives are more cost-efficient than NBFIs. The mean efficiencyscore of financial cooperatives and NBFIs is 66 and 56 per cent, respectively. This denotesthat what costs Birr 1.0 for financial cooperatives costs Birr 1.178 for NBFIs to producesimilar outputs. In other words, financial cooperatives are 17.8 per cent more efficient thanNBFIs, and the efficiency gap is statistically significant. Concurrently, as shown inFigure 2, about 41 per cent of financial cooperatives have an efficiency score closer tothe efficient cost frontier. On the other hand, more than 35 per cent of the NBFIs operateat higher costs compared with the best practicing microfinance providers in the sample.Besides the average cost-efficiency scores, the financial cooperatives dummy included

as the correlates of inefficiency in the cost-efficiency estimation indicates relative cost

Figure 2. Frequency distribution of cost-efficiency scores by ownership form



control among financial cooperatives. As shown in Panel B of Table 1, the estimatedcoefficient for financial cooperatives dummy is negative, implying that financialcooperatives are more cost-efficient. This can be due to the fact that financial cooperativesbenefit from better information and cheaper enforcement mechanisms available to them,as they are embedded to small communities, and their owners are providers of both thedemand for and supply of loanable funds. In contrast, specialised microfinanceinstitutions heavily rely on external sources of capital and serve wider client bases,and the extent of knowing each other, which serves as an enforcement mechanism, isrelatively imperfect.The relative cost-efficiency by financial cooperatives is consistent with prior research

works performed by Haq et al. (2010); Mersland and Strøm, 2008; Lafourcade et al.(2005), and Lapenu and Zeller (2001). For instance, Haq et al. (2010) found that cooperativemicrofinance institutions are the most cost-efficient and productive (serve large proportion ofrural population) compared with NBFIs.

5 CONCLUSIONS

This paper analysed the implication of imposing financial sustainability endeavour on thetraditional social mission of microfinance outreach to the poor. The results show that servingthe poor clients and achieving financial sustainability (measured by levels of cost-efficiency)are difficult objectives to be achieved simultaneously. It is found that providing small sizeloans and catering more to women borrowers, which imply greater depth of outreach, arepositively and negatively linked with the level of cost-efficiency, respectively. The resultsspecifically show that microfinance providers that are closer to the best practicing costfrontier are those with higher average loan sizes and lower proportion of women borrowers.The results also indicate the presence of a wider cost-efficiency gap between financial

cooperatives and specialised microfinance institutions. The cost-efficiency gap can bedue to the inherent ability of financial cooperatives to effectively utilise social collateralsas contract enforcement mechanisms and dispense with information costs, compared withspecialised microfinance institutions. On the other side, the commitment to expandoutreach through increasing branches, reliance on non-commercial funds and lack ofpecuniary incentives in some of multi-stakeholder-owned NBFIs may have also resultedin efficiency gap. Although institutional diversification and resulting competitions canbenefit clients by lowering costs and improving services, financial cooperatives shouldbetter enable the microfinance industry to deliver improved financial services at lowercosts compared with the specialised NBFIs, as such financial cooperatives are found tobe relatively cost-efficient.

ACKNOWLEDGEMENTS

We thank the European Research Institute on Cooperatives and Social Enterprise(EURICSE) and the International Food Policy Research Institute (IFPRI)–analytical supportproject to Agricultural Transformation Agency (ATA) of Ethiopia for their financial andlogistic supports during data collection. We are specifically grateful for Gianluca Salvatori,Riccardo Bodini and Shahidur Rashid. We are also grateful to an anonymous reviewer forcomments and suggestions received during the review process of the paper.



REFERENCES

Amha W. 2000. Review of microfinance industry in Ethiopia: regulatory framework and performance.Occasional paper No. 2, Association of Ethiopian Microfinance Institutions, Addis Ababa.

Amha W. 2007. Managing growth of microfinance institutions (MFIs): balancing sustainability andreaching large number of clients in Ethiopia. AEMFI, Addis Ababa.

Amha W, Peck D. 2010. Agricultural finance potential in Ethiopia: constraints and opportunities forenhancing the system. AEMFI, Addis Ababa.

Armendáriz de Aghion B, Szafarz A. 2009. On mission drift in microfinance institutions. WorkingPaper 09–015, Solvay Brussels School of Economics and Management: Centre EmileBernheim, Brussels.

Battese GE, Coelli TJ. 1995. A model for technical inefficiency effects in a stochastic frontierproduction function for panel data. Empirical Economics 20: 325–332.

Cull R, Demigu¨c-kunt, A, Morduch J. 2007. Financial performance and outreach: a global analysisof leading micro banks. The Economic Journal 117: 107–133.

Degefe B, Nega B. 2000. Annual report on the Ethiopian economy: vol. I. Ethiopian EconomicAssociation, Addis Ababa.

Demirgüç-Kunt A, Klapper L. 2012. Financial inclusion in Africa: an overview. Policy ResearchWorking Paper 6088, The World Bank Development Research Group, Washington DC.

Federal Cooperative Agency. 2012. Annual publication of the Federal Cooperative Agency ofEthiopia. Addis Ababa.

Frank C. 2008. Stemming the Tide of Mission Drift: Microfinance Transformations and the DoubleBottom-Line: Focus Note. Women’s World Banking: New York, NY.

Gilligan DO, Hoddinott J, Taffesse AS. 2008. The impact of Ethiopia’s Productive Safety NetProgramme and its linkages. IFPRI Discussion Paper 00839, Washington DC.

Gonzalez A. 2007. Efficiency drivers of microfinance institutions (MFIs): operating expenses and itsdrivers. Discussion paper No. 2: Microfinance Information Exchange, Washington DC.

Hansmann H. 1996. The Ownership of Enterprise. The Belknap Press of Harvard University Press:Cambridge, MA.

Haq M, Skully M, Pathan S. 2010. Efficiency of microfinance institution: a data envelop analysis.Asia Pacific Financial Markets 17(1): 63–97.

Hermes, N, Lensink, R, Meesters, A. 2011. Outreach and efficiency of microfinance institutions.World Development 39(6): 938–948.

Hulme D, Mosley P. 1996. Finance Against Poverty. Routledge: New York, NY.Jensen MC, Meckling WH. 1976. Theory of the firm: managerial behaviour, agency costs and

ownership structure. Journal of Financial Economics 3(4): 305–360.Kumbhakar SC, Lovell CAK. 2000. Stochastic Frontier Analysis. Cambridge University Press:

New York, NY.Lafourcade A, Isern J, Mwangi P, Brown M. 2005. Overview of the Outreach and Financial

Performance of Microfinance Institutions in Africa. MIX: Washington DC.Lapenu C, Zeller M. 2001. Distribution, growth, and performance of microfinance institutions in

Africa, Asia and Latin America. FCND Discussion Paper No. 114, Washington DC.McIntosh C, de Janvry A, Sadoulet E. 2005. How rising competition among microfinance institutions

affects incumbent lenders. The Economic Journal 115: 987–1004.Mersland R, Strøm RØ. 2008. Performance and trade-offs in microfinance organization: does

ownership matter? Journal of International Development 20(5): 598–612.Quayes S. 2012. Depth of outreach and financial sustainability of microfinance institutions, Applied

Economics 44(26): 3421–3433.



Rosengard JK. 2004. Banking on social entrepreneurship: the commercialization of microfinance.Mondes en Développement 126(2): 25–36.

Weiss J, Montgomery H. 2005. Great expectations: microfinance and poverty reduction in Asia andLatin America. Oxford Development Studies 33(3–4): 391–416.



cost-efficiency and outreach of microfinance institutions: trade-offs and the role of ownership

Documents