This is the first post of my series of reflections from my development seminar taught by Erica Field and Xiao Yu Wang. I think it’s good to start this by a quote from Banerjee (2001):
Development economists are, perhaps by necessity, optimistic people. One does not become a development economist if one believes the world’s poorest are doing as well as they possibly could.
To me this is the glamor of development economics — it gives us a sense of fulfillment, a feeling of “yes we can help to make the world a better place” (though whether we can really do that is not clear).
In our class today Xiao Yu presented a simple model about borrowing in developing countries (Banerjee 2001) and Erica summarized empirical evidence for how to encourage people to save more.
In many developing countries, people get credit through informal means (family, friends, saving groups etc), but this mechanism is highly insufficient. Informal credit is limited to a small community where people have similar income sources and are subject to negative shocks. So when something bad happens, there might not be enough money available for borrowing. Formal financial institutions (e.g. banks) often find it hard to support their operations in remote and poor neighborhoods because of high transaction costs, petty deposits which they can hardly use to make investments, and high monitoring costs for loan repayment.
Informal credit is more efficient for the lenders in such settings, because lenders can select “safer bet” from both indirect and direct screening. Indirect screening includes eliciting risk levels of projects through the borrowers’ willingness to pay, threatening to cut off future credit given default, and interpersonal linkages to reduce probability of default. Borrowers are also screened directly because lenders know a lot of information about people in the same community and can require collateral before giving out a loan.
Banerjee’s model provides us with a good starting point to think about how credit markets work in developing countries, but many of the assumptions do not seem realistic. For example, default cost as a fraction of wealth is somewhat absurd. And monitoring costs tend to be variable rather than fixed. But one key assumption of this model is perfect information, so there is little room for us to introduce heterogeneity among borrowers.
My takeaway from Erica’s presentation is more disperse: She’s got some really nice questions.
First, under what circumstances does credit reduce poverty, and when does it potentially exacerbate it? It may sound counter intuitive to say that credit might push the poor to the worse, but debt traps are commonly seen in the developing world.
Second, what is the empirical evidence on providing banking services to the poor?
Third, what are the mechanisms of ways to increase savings among the poor?
Empirical evidence from Dupas and Robinson (2013) has shown, through a randomized control trial in Kenya, that
people prefer to “tie their hands” earlier (to have a piggy bank, in this case) so that they can commit to a fixed purpose of saving (in this case, health). They have also demonstrated that social ties and pressure from groups can encourage saving.
One interesting point Erica mentioned was that savings products in LDCs are harder to come by than loans. Banks don’t offer loans because it’s less profitable than loans; transaction costs are higher; hard to charge negative interest or penalties (usually b/c policy reasons); regulations against non-profits and not allowed to hold deposits (especially true for Southeast Asia).
Banerjee, Abhijit, “Contracting Constraints, Credit Markets, and Economic Development,” MIT Working Paper 2001.
Dupas,P. and Jonathan Robinson, “Why don’t the poor save More?” American Economic Review.