Microcredit and microsavings for small enterprises in Pakistan
Lead Research Organisation:
University of Oxford
Department Name: Economics
Abstract
RESEARCH QUESTIONS AND IMPACT: We will address two main questions: (i) How do ROSCA-like products compare to standard loan contracts for microenterprises? (ii) Can banks overcome contract enforcement problems in offering ROSCA-like financial products?
These questions are directly motivated by recent research on microfinance, which suggests that savings products may be more useful for poor communities than loan products. To understand how savings products can be improved for the poor, we need to understand several important issues: Why don't the poor save more? How does household behaviour change when the household is a debtor, rather than a creditor? Can better microfinance products be developed to facilitate saving? These are the issues that underlie our research questions in this project. We anticipate that our research results will therefore have direct relevance to both policymakers and microfinance providers. If our product proves viable, it could provide small enterprises with a mechanism to increase investment and employment - and, in turn, to drive growth.
METHODS AND DATA: We will use a Randomised Controlled Trial. We will use a 'cluster-randomized' design, treating in some locations but not in others. In treated locations, we will (i) randomly assign firms to different positions in a rotating credit structure (see the 'Case for Support' for detail on the product), and (ii) randomly vary the interest rate. In total, we anticipate assigning 800 microenterprises to the new product, 800 as controls in the treated locations, and 800 as controls in untreated locations. We will stratify clusters by average cluster characteristics (specifically, the total number of clients at the branch, the number of female clients at the branch, length of branch operation, and average loan size made at the branch). Within treated clusters, we will stratify microenterprises by gender, sector of operation and baseline profits. A baseline and three follow-up surveys of the targeted microenterprises will be conducted at three-month intervals in control and treated locations, to form a four-wave panel. These surveys will include behavioural games, designed to explore the effect of time preferences, risk preferences, social preferences and peer pressure on take-up of the intervention and repayment/saving. We will also collect high frequency data using short phone surveys, and we will augment this sample with data generated by the bank internally.
DISSEMINATION: Most importantly, we will disseminate by close engagement with our partner bank, NRSP Microfinance Bank, who have expressed direct interest in our results. We will also run a workshop in Lahore at the end of our research to disseminate results to the local academic community and policy community. Finally, we will use CSAE's extensive dissemination network to share results with stakeholders in other contexts (for example, through the CSAE Briefing Paper series, designed to share policy-relevant results).
These questions are directly motivated by recent research on microfinance, which suggests that savings products may be more useful for poor communities than loan products. To understand how savings products can be improved for the poor, we need to understand several important issues: Why don't the poor save more? How does household behaviour change when the household is a debtor, rather than a creditor? Can better microfinance products be developed to facilitate saving? These are the issues that underlie our research questions in this project. We anticipate that our research results will therefore have direct relevance to both policymakers and microfinance providers. If our product proves viable, it could provide small enterprises with a mechanism to increase investment and employment - and, in turn, to drive growth.
METHODS AND DATA: We will use a Randomised Controlled Trial. We will use a 'cluster-randomized' design, treating in some locations but not in others. In treated locations, we will (i) randomly assign firms to different positions in a rotating credit structure (see the 'Case for Support' for detail on the product), and (ii) randomly vary the interest rate. In total, we anticipate assigning 800 microenterprises to the new product, 800 as controls in the treated locations, and 800 as controls in untreated locations. We will stratify clusters by average cluster characteristics (specifically, the total number of clients at the branch, the number of female clients at the branch, length of branch operation, and average loan size made at the branch). Within treated clusters, we will stratify microenterprises by gender, sector of operation and baseline profits. A baseline and three follow-up surveys of the targeted microenterprises will be conducted at three-month intervals in control and treated locations, to form a four-wave panel. These surveys will include behavioural games, designed to explore the effect of time preferences, risk preferences, social preferences and peer pressure on take-up of the intervention and repayment/saving. We will also collect high frequency data using short phone surveys, and we will augment this sample with data generated by the bank internally.
DISSEMINATION: Most importantly, we will disseminate by close engagement with our partner bank, NRSP Microfinance Bank, who have expressed direct interest in our results. We will also run a workshop in Lahore at the end of our research to disseminate results to the local academic community and policy community. Finally, we will use CSAE's extensive dissemination network to share results with stakeholders in other contexts (for example, through the CSAE Briefing Paper series, designed to share policy-relevant results).
Planned Impact
We will engage with NRSP Microfinance Bank extensively during the design, implementation and evaluation phases of the products. We are currently running a pilot study, which will inform the design and implementation of the project. We will organize a workshop to discuss the findings from the pilot with NRSP Microfinance Bank's representatives, as well as with researchers from the Lahore School of Economics, and jointly plan the next steps. We will present the results from the product evaluation in a technical report for the NRSP Microfinance Bank and further discuss with the bank the viability of implementing the product on a wider scale.
We will share the results of our study to the wider financial and policymaking community in Pakistan, through workshops organised by the Lahore School of Economics. We will also publish a CSAE Briefing Paper to share our findings with the broader policy community. Finally, we will present our findings at academic conferences and publish them in economic journals, with the aim of reaching a wide audience within academia.
We will monitor and evaluate the impact of our research in several ways. First, we intend to have a close ongoing relationship with NRSP Microfinance Bank, so we will have direct engagement in any future product roll-out based on our trial results. Second, we will monitor directly the academic impact of our research, through downloads and citations of our research papers. Third, we will keep track of the kinds of products being offered by large microfinance institutions worldwide. This is reasonably straightforward, given that large microfinance institutions make their product offerings public.
Finally, we note that our research provides valuable new opportunities to build capacity in both Pakistan and the UK. Staff and students at the Lahore School of Economics will have the chance, by participating in the design, implementation and evaluation of the new products, to gain experience in conducting a large field project. Through their participation in research activities and dissemination of the results, they will enter in contact with academics and policy makers in the Pakistan, in the UK and abroad. UK-based members of the research team will have the opportunity to gain understanding of the Pakistani context, and to enrich their academic and policy network in the country. In order to facilitate knowledge exchange between the two countries, we will plan visits by researchers from Pakistan to the UK, and by UK-based researchers in Pakistan. These visits will be organised so as to maximise interaction during the design and implementation phase of the project, but also to enable researchers to communicate their findings to a broad audience in the two countries.
We will share the results of our study to the wider financial and policymaking community in Pakistan, through workshops organised by the Lahore School of Economics. We will also publish a CSAE Briefing Paper to share our findings with the broader policy community. Finally, we will present our findings at academic conferences and publish them in economic journals, with the aim of reaching a wide audience within academia.
We will monitor and evaluate the impact of our research in several ways. First, we intend to have a close ongoing relationship with NRSP Microfinance Bank, so we will have direct engagement in any future product roll-out based on our trial results. Second, we will monitor directly the academic impact of our research, through downloads and citations of our research papers. Third, we will keep track of the kinds of products being offered by large microfinance institutions worldwide. This is reasonably straightforward, given that large microfinance institutions make their product offerings public.
Finally, we note that our research provides valuable new opportunities to build capacity in both Pakistan and the UK. Staff and students at the Lahore School of Economics will have the chance, by participating in the design, implementation and evaluation of the new products, to gain experience in conducting a large field project. Through their participation in research activities and dissemination of the results, they will enter in contact with academics and policy makers in the Pakistan, in the UK and abroad. UK-based members of the research team will have the opportunity to gain understanding of the Pakistani context, and to enrich their academic and policy network in the country. In order to facilitate knowledge exchange between the two countries, we will plan visits by researchers from Pakistan to the UK, and by UK-based researchers in Pakistan. These visits will be organised so as to maximise interaction during the design and implementation phase of the project, but also to enable researchers to communicate their findings to a broad audience in the two countries.
Organisations
| Description | 1. Working with our partner organisation (the National Rural Support Programme), we were able to implement a novel microfinance product, based on the rotating structure of a ROSCA (known in Pakistan as a 'committee'). Our results show a substantial demand for both saving and credit contracts: when offered both a credit and a saving contract, 46 percent of participants who take up at least one offer, accept both types of contracts. This is a new empirical result concerning demand for microfinance products; it builds upon an earlier result from our research team on a much smaller scale (funded through a separate project in 2013), showing simultaneous demand for credit and for savings products in daily microfinance products. 2. In offering our product, we were able to test the demand for: soft commitment in the form of reminders; hard commitment in the form of penalties for missing an instalment; and flexibility in the schedule of instalments. People set goals, make plans and resolutions, but often fail to keep them. A number of explanations have been proposed for this widely observed pattern: time-inconsistent preferences, limited attention and inability to resist others' demands are the most cited ones. Individuals can use a number of tools to solve commitment issues. All these tools serve the purpose of restricting one's future consumption choices. The literature distinguishes between two classes of commitment devices: soft commitments devices, which impose primarily psychological consequences for not meeting one's goals; and hard commitment devices, which involve economic incentives. We observe no demand for any of the behavioural features we added to the product, like flexibility and reminders. In particular, individuals appear to actively dislike commitment features. Finally, we find no transformative impact of microfinance on business growth or household welfare. To the best of our knowledge, this is the first time that demand for all most common types of commitment devices is tested within the same experimental design. The project thus contributes both to policy, through the design and evaluation of a novel financial product, and to the academic debate on individuals' demand for commitment. 3. We developed and implemented a new incentivised measure of female empowerment. As part of the baseline survey, we asked all participants to select one of three possible gifts to be received at endline. The gifts were made to order in leather. They were chosen to be similar in value, style, and purpose. But they were aimed at a specific gender or age group: i.e., a female wallet, a male wallet, and a pencil case. In addition, the gifts were to be inscribed with the initials of the intended recipient. We use this (in a way described in detail in the academic paper) as an incentivised measure of female empowerment; i.e. a measure of a respondent's willingness to overrule her husband's choice. We find a small but significant positive effect of our treatment on this incentivised measure. This is a novel kind of result for research on microfinance. It is also a novel form of incentivised measure that could be implemented in a wide variety of other studies in other empirical contexts. |
| Exploitation Route | We see three ways in which our findings might be taken forward. First, we shared preliminary results from the project with our partner institution, the National Rural Support Program (NRSP), and with other microfinance institutions and policy makers during a conference held in Lahore in April 2017. We plan to further engage the policy and financial community in similar events in the future, to discuss the lessons learnt through the project and avenues for future research. Second, this project has prompted further experimentation in the microfinance space. In particular, building on this project, one of the investigators (Simon Quinn) has developed complementary research with another Pakistani microfinance institution (Akhuwat) to study demand for equity-based microfinance; this is a quite different form of microfinance, but one that may also show substantial demand among Pakistani microenterprises. Finally, all of our data will be made available in anonymised form to the wider academic community, for secondary data and replication analysis. |
| Sectors | Communities and Social Services/Policy,Financial Services, and Management Consultancy |