Oct 18, 2008

What’s So Hard About Saving?

by Tim Ogden

We’re live-blogging the Innovations for Poverty Action/Financial Access Initiative Microfinance Conference.

One of the many confounding questions of microfinance is why people don’t save more. The first blush assumption that the poor cannot save because they don’t have the money has been dis-proven by data. Surveys of household consumption among those living on less than $2 a day show that they spend a significant portion of their income on non-essential items (see “The Economic Lives of the Poor” by Esther Duflo and Abhijeet Bannerjee). In many cases where savings products are available they are widely used, even when they are objectively terrible (negative interest rates, inconvenient).

The obvious answer is that people don’t save because they don’t have access to savings products. This is where things get strange – anyone who has a business or a loan has a savings product available to them. They can either invest in the business or pay off their loan early. But neither happens regularly.

Two presentations were given that attempt to address why this doesn’t happen in different ways. First, Sendhil Mullainathan reported on a project in Bangladesh and the Philippines with women fruit vendors. These vendors borrow and repay money on a daily basis, at an average interest rate of 5 percent per day. They use the money to buy fruit in the morning at a central market which they resell in their own neighborhood at a small profit. These women, by skipping one purchased cup of tea a day (and saving the money instead) could double their income within 30 days by reducing the amount they borrow each day.

With such high returns from savings, why don’t they save? One possibility is that they don’t understand the math – and you could fix this by providing financial literacy education. To test this idea, half of the women were given a one time grant equal to the amount they borrow each day, essentially freeing them entirely from their debt burden. Mullainathan then tracked the women for up to six months. Half of the women who got these grants were enrolled in financial literacy training that specifically focused on the benefits of avoiding high-interest-rate debt and how their gains from saving would compound. Another possible explanation is that the women value today’s spending far higher than gains in the future. The project tested this possibility simply by watching what the women did – if they in fact put “too high” of a value on today’s cash they would return to high levels of indebtedness almost immediately.

In practice what Mullainathan found was that neither was true. All of the women returned to high rates of indebtedness whether they received the financial literacy training or not (which it must be acknowledge might be because the financial literacy training was poorly delivered). But they didn’t return to indebtedness very quickly – it took quite a while (on the order of months) to end up back where they started. This indicates that they weren’t “blowing” the money on current consumption. What seems to be the case is that the women experienced a series of what the economists call “shocks” – events that required them to spend in the short term, such as buying medicine for a sick child. Over time, these shocks accumulated and ate up all their savings.

A project presented by Pascaline Dupas of UCLA looked at a different set of explanations for low savings rates – specifically whether the cost of a savings account or the actual business returns from saving are lower than expected. The project provided “free” savings accounts to a randomly selected group of microentrepreneurs (though there were fees related to withdrawals) and then followed both their saving behavior and their business outcomes. They found that just over half of the people who accepted the free account used it – and that those that did showed major gains in their businesses and in their personal consumption. Somewhat surprisingly, all of the positive impact was among women participants. Men who participated did not benefit more than men who did not. The results indicate that the returns from savings are high which adds to the confounding question of why almost half of people didn’t use the accounts.

While both presentations were based on preliminary data, and much analysis remains to be done there are some clear implications. First, a lot more work needs to be done to figure out how to improve actual savings rates and improve savings products. Second, the relative focus on credit versus savings among microfinance organizations should quite possibly be flipped.