Ask farmers from low-income countries why they don’t purchase critical inputs, such as improved seeds and fertilizer, and they will most likely tell you that they “lack the funds” to do so. While this may be a catch-all excuse, the answer is all the more surprising given the high marginal returns that these investments usually entail.
One solution to this liquidity problem is to encourage formal savings. Low-income households do save informally in more expensive and risky ways (holding cash at home, purchasing livestock, etc.) but they find it hard to save at a financial institution. There are many reasons for this behavior. Absent the more recent technology-based solutions, such as mobile banking or banking correspondents, transaction costs can be high given the sometimes substantial distances to branches and the costly and unreliable transport. In addition, individuals may lack knowledge about the benefits of formal savings and may not be familiar with account-opening procedures. Banks don’t typically make much money with savings products targeted to low-income products unless they are loaded with hidden fees and commissions, in which case potential customers may be better off not saving at all. (One of these days I’ll blog about an audit study we are doing in Mexico to understand the quality of information and financial products offered to low-income households.)
To understand why households do not save formally in “no-frills” accounts, in a recent paper with Lasse Brune, Jessica Goldberg and Dean Yang, we randomized offers of account-opening and deposit assistance for savings accounts in a nearby bank. Treated farmers were randomly assigned to one of two types of savings interventions. The first group was offered an “ordinary” bank account with standard features. The second group was offered the ordinary account as well as a “commitment” savings account that allowed account holders to request that funds be frozen until a specified date (e.g., immediately prior to the planting season, so that funds could be preserved for farm input purchases). Other farmers were assigned to a control group that was surveyed but not offered assistance with opening either type of savings account.
The shocking result is that while the commitment treatment had large positive effects on agricultural investment of magnitude similar to the amount saved in both accounts, most of the money was kept in the “ordinary” bank account, not the commitment account (see Figure 1 showing the average balances in each account in the commitment and ordinary treatment groups). We conclude that the commitment accounts per se did not therefore help farmers solve their self-control problems, giving them the discipline to maintain their balances until the next planting season, because the balance in the commitment account was only a small fraction of the overall increase in agricultural investment. This increase in inputs led to more land under cultivation, and higher crop output and household expenditures in the months immediately after harvest. By contrast, ordinary treatment effects were always smaller in magnitude and never statistically significantly different from zero.
So if the commitment account does not solve the self-control problem, then what problem is it solving? Here’s where we do not have a clear answer. It could be that commitment accounts may have helped farmers to refrain from sharing with others in their social network. We did collect anecdotal evidence that suggested that individuals were selective as to who they decided to help. In a way, the existence of the account could have provided an excuse to turn down requests for assistance from the social network by claiming that savings were inaccessible. However, we do not find a reduction in net transfers to other households among the commitment treatment group.
It could also be that the effects operate through some other psychological channel such as mental accounting, for example. The idea here is that by opening a second account that they label in their mind as an “investment account,” farmers are somehow able to set aside funds for the next agricultural season although the funds are not physically locked away.
In any event, because most of the savings were kept at home and thus were flexible, the welfare of commitment account holders is likely to have improved as a result of the intervention.
To conclude, although we provide evidence against the self-control hypothesis, more research is needed to uncover the precise mechanism that is driving the results.
(if you can not see the figure, click here)
Join the Conversation