Savings groups and financial intermediation among the poor
Informal savings and borrowing institutions are how to intermediate between savers and borrowers within the developing world. But if these associations attract mostly savers or mostly borrowers, or are concentrated in one occupation, they'll not function also as they ought to . This column uses survey results from Malawi to suggest that commitment savers and borrowers mix in such associations, but occupations have attended stay together . this might make them susceptible to shocks like a nasty harvest.
According to International Bank for Reconstruction and Development estimates, in 2014, two billion adults worldwide didn't have an account with a bank, other financial organization , or mobile money provider (Demirgüç-Kunt et al. 2015). Microfinance has been praised as how to expand formal access to credit to poor households in order that they will make productive investments and lift themselves out of poverty. In 2016, 132 million people were microfinance borrowers (Microfinance Barometer 2017).
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
But evidence from randomised controlled trials shows that access to microcredit produces little or no impact on investment or poverty (Banerjee et al. 2015), perhaps partially thanks to the rigidity of ordinary microfinance contracts (Field et al. 2013). In contrast, access to savings accounts appears to be beneficial for a minimum of some poor individuals (Dupas and Robinson 2011, Karlan et al. 2015). This includes accounts with 'commitment' features that prevent withdrawals within the short term or enforce regular deposits (Ashraf et al. 2006, Brune et al. 2011). Yet it's often difficult to seek out sustainable ways of offering formal savings accounts to the poor, particularly accounts that carry favourable interest rates.
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
Savings and lending groups
One possible solution is to encourage peer-to-peer savings and lending institutions, during which some members can save et al. can borrow by requesting a loan out of saved funds.
On the savings side, peer pressure and therefore the reminders provided by regular group meetings may allow some members to commit themselves to saving without access to a proper account (Gugerty 2007, Kast et al. 2012, Breza and Chandrasekhar 2015).
On the lending side, communities could also be ready to draw on their shared information and social capital to permit repayment to be more flexible than standard microfinance.
Moreover, if this type of monetary intermediation between savers and borrowers takes place, savers can potentially earn interest on deposits.
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
NGOs and other organisations have promoted informal savings and borrowing institutions, referred to as Village Savings and Loan Associations (VSLAs), Self-Help Groups (SHGs) and various other acronyms. These groups now have quite 100 million members worldwide (Greaney et al. 2016). Impact evaluations have shown positive effects from access to VSLAs on household food security (Beaman et al. 2014, Ksoll et al. 2016). this is often possibly linked to increased agricultural investments. Researchers have also shown that creating groups buy their own training might screen credit risk (Greaney et al 2016), and the way there's a trade-off between including poorer individuals and reducing groups' capacity to lend (Burlando and Canidio 2017).
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
Little is understood , though, about whether the groups really promote financial intermediation between savers and borrowers. indebtedness in repaying loans or broader social frictions may stop savers (including commitment savers) and borrowers from grouping together. If this happens, groups may focus almost entirely on saving or almost entirely on borrowing. Similar constraints may stop members of various occupations from grouping together. this is able to again limit the chances for financial intermediation, and it could also leave the groups susceptible to covariate shocks like bad harvests.
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
A broader concern with promoting financial inclusion using these groups, instead of through formal banking, is that funds cannot flow across groups in several communities within the same way that funds can flow across bank branches. Therefore it's important to know how members sort across groups, a minimum of within communities.
Surveying VSLA groups in Malawi
In a recent study, we use a completely unique dataset that we collected from 150 VSLA groups in northern Malawi in 2013 to undertake to answer these questions (Cassidy and Fafchamps 2018). We took a census of all 3,800 members from the groups that formed as a part of a cluster-randomised controlled trial between 2009 and 2011, as evaluated by Ksoll et al. (2016). We also identified around one-fifth of the members as having been a part of the sample for the first evaluation. We are therefore ready to link them back to the evaluation data, including detailed measures of their time preferences and their baseline borrowing and savings behaviour before the introduction of VSLAs. We surveyed the groups between two and 4 years after they formed, so our data are uniquely suited to studying the long-run equilibrium sorting of members across groups, and therefore the long-term functioning of the groups more generally.
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
How members sorted across groups
We use a way called dyadic multivariate analysis to review how members sorted across groups (Fafchamps and Gubert 2007). Since most villages had quite one group – with some villages having up to nine – individuals in theory could have chosen between several groups. We estimate what determines the probability of two individuals from an equivalent village being members of the samegroup, conditional on being members of some group.
Based on the arguments above, the variables we specialise in for every pair of people are:
whether the individuals both seem to possess a requirement for commitment savings – proxied by a measure of present-bias – or whether one does while the opposite one does not; and
whether the individuals have an equivalent occupation or different occupations
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
1. Potential commitment savers match with potential borrowers.
Individuals who appear present-biased are far more likely to sort into groups with individuals who aren't present-biased. All else equal, members who have an equivalent time preferences are 16.6 percentage points less likely to be members of an equivalent group, compared to a baseline probability of 29.6%.This is promising – it suggests that present-biased individuals, who likely join VSLAs out of a requirement for commitment savings, will sort into groups with time-consistent individuals, who likely join VSLAs as how to borrow (we provide more details of this argument in our paper). this is often a crucial way during which savings and loan groups may enhance efficiency and lift the welfare of both commitment savers and borrowers.
2. Individuals from an equivalent occupation stay together .
In contrast, we discover little evidence that individuals engaged in agriculture sort into groups with those engaged in non-farm activities. this is able to allow farmers to save lots of harvest income across the year, whilst enabling those engaged during a small business to possess taken short-term loans for investment. Instead, we discover strong positive assorting on occupation. All else equal, members who have an equivalent occupation are 18.1 percentage points more likely to be members of an equivalent group. this might flow from to high costs of screening, monitoring and enforcing loans, or transacting with individuals from a special occupation.
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
It is important to notice the restrictions of our analysis. We include an upscale set of controls, including village fixed effects. Figure 1 shows that whether members are both from a female-headed or both from a male-headed household may be a particularly predictive control. It increases the probability of being within the same group by 17.6 percentage points. Moreover, by using data on time preferences and occupation from the panel dataset collected before the introduction of VSLAs, we are ready to rule out reverse causation – i.e. that folks who joined an equivalent VSLA group find yourself looking as if they need different time preferences, or similar occupations, for a few reason. But we cannot randomise individual preferences and occupations, then these variables could still be correlated with something that we don't observe. Given this, our evidence is strongly suggestive, but we cannot rule out that some unobserved variable correlated with time preferences or occupation could drive our results.
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
Future research
Our results are promising, insofar as they suggest that VSLAs may have facilitated financial intermediation between commitment savers and borrowers. On the opposite hand, we don't see signs of intermediation between members from different occupations. If true, this exposes informal financial institutions to covariate risk.
linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus linkswithus
Our research isn't designed to check what would happen if NGOs prefer to encourage members to sort into groups explicitly on the idea of proxies for saving and borrowing needs or attempt to encourage members of various occupations to group together. Future research could examine whether assigning members along these lines would enhance saving and borrowing, or whether it might reduce the power of groups to enforce loan repayment (Karlan 2007).
More research is additionally needed to know the sustainability of those groups following negative shocks, like a nasty harvest. a method to protect against this is able to be to integrate VSLAs into a bigger depository financial institution or federation. Alternatively, VSLAs could access funds from the formal sector, as in India under the commercial bank for Agriculture and Rural Development’s SHG Bank Linkage Program. it might even be useful to possess research into whether this has been successful.
