Matthew Olckers Economist by day; graffiti nerd by night.

'Mobile-izing Saving with Automatic Contributions' Blumenstock, Callen & Ghani (2017)

Saving is tough. No matter how hard you try, there always seems to be nothing leftover at the end of the month. You need to trick yourself by hiding money away before it hits your bank account. Saving needs to be automatic, but which incentives are needed to start automatic contributions in the first place, and which mechanisms explain sustained monthly saving? Joshua Blumenstock, Michael Callen, and Tarek Ghani use a comprehensive field experiment to answer these questions. Their analysis highlights the power of defaults.

Before going into the details of the experiment, I would like to highlight the context of the study. The experiment was conducted with the employees of Roshan, Afghanistan’s largest mobile phone operator. Blumenstock, Callen, and Ghani (2017) start the paper by sharing feedback from focus groups they ran before the start of the experiment. It is refreshing to hear the stories behind the data. Here is an excerpt:

While saving money to buy a house or a car seems out of reach for most, saving money in case of death or illness is an unquestioned necessity. They often store their money in a metal box at home (a traditional method), or with a trustworthy (often richer) relative. They tell stories of themselves or people they know going hungry or reducing food quality after a shock of some kind, and describe the humiliation of young men unable to marry for lack of money. Along with the differences, however, we also find similarities between working Afghans and their Western counterparts. Almost all focus group members report struggling with bills and running out of cash before the end of the month.

Financial services in Afghanistan is underdeveloped, and locals distrust formal financial institutions. This is not surprising given the decades of conflict that Afghans have endured. Mobile money is a promising alternative, since there are very few bank branches. With mobile money, any shop vendor can operate as a bank branch. Roshan employees are paid with mobile money, and the experiment offered a separate mobile money account called M-Pasandaz into which automatic contributions could be made. Treatment groups differed in default contribution rates, and matching incentives. Either zero, or five percent of the salary was automatically contributed to M-Pasandaz, and employees received either zero, 25, or 50 percent as a bonus matched to the amount saved, but not withdrawn, after six months. Importantly, it was super easy for employees to change their contribution rate. They merely had to phone or visit their HR office.

The results are striking. The default rate of five percent increased saving by 40 percent, and employees needed a 50 percent matching incentive to reach the same level of saving as those employees who started with the five percent default. Interestingly, the researchers also manage to pin down the mechanism for this strong default effect. They show that present-bias (the “agh, I will do that tomorrow” bias) is what drives the effect of defaults.

People procrastinate – even through some action may have large future benefits, if you incur the cost today, then you are likely to delay to tomorrow. Clearly, saving suffers from procrastination. However, the same effect that prevents you from saving voluntarily helps you to continue saving at a default. Changing your contribution rate (with a phone-call to HR) has a cost today of time and hassle, but the benefit is only seen at your future paycheck. If this behavior is repeated, you can see how you will end up saving more if you are assigned a default rate.

The results of the experiment provide conclusive evidence that defaults matter a great deal – even in a developing country context. Employers, and governments: take note!