Saving is hard for most people (unless you’re Chinese or German it seems). A typical approach to encourage savings might be to emphasize all the good reasons to save (children’s education, healthcare needs, retirement, etc.) and set savings goals for these reasons. So I was interested to come across a post on the Squared Away financial literacy blog  alerting me to a new paper  in the Journal of Marketing Research by Dilip Soman and Min Zhao, two marketing professors from Toronto. They challenge the idea that getting people to think about all the reasons to save is the best way to encourage savings, and instead provide experimental evidence that a single savings goal can lead to higher savings than having multiple savings goals.
They motivate their work in terms of psychology and decision-making processes. They note that prior work has suggested that pursuing a goal occurs in two distinct stages: first an initial deliberative mindset stage in which individuals are uncertain about their goals and consider the tradeoffs between goals, and then an implementation mindset in which having established the goal they wish to pursue, individuals figure out when, where and how to achieve it.
They argue that if you tell people “it is really important to save, for reasons such as retirement, children’s education and future health needs”, this can get people stuck in the deliberative mindset, contemplating which of these is more important and the trade-offs among each- whereas just telling them “save for your kid’s education” allows them to move straight to the implementation mindset, and think about how to do this savings.
This paper is unusual in that it contains four experiments (one field and three lab), each on different populations. It has to be the weirdest set of subjects I’ve seen - the field experiment is on 83 households who agreed to participate in a financial literacy program in a small town in India; while the three lab experiments are done on respectively 194 participants of an executive skills training program in Canada, 134 members of a market research panel in Hong Kong; and 149 members of a market research panel in India. There is no discussion of the external validity of these non-randomly chosen subject pools, nor of whether one can directly compare results across the studies – the presumption is that since the same types of behavior tend to get observed in the different experiments, this is somewhat universal behavior.
Moreover, the results are frustrating to read for an economist, since the norm in marketing appears to be to just discuss the p-values from two-way ANOVA specifications, making it difficult to always see sample sizes and treatment effect sizes.
Nevertheless, the paper is a fascinating read, particularly for the way it generates a result with a field experiment, and then uses a series of lab experiments to explore and rule out various mechanisms that might lead to the results in the field experiment, and to examine how the results may vary as different parameters change.
In the field experiment: the 83 households were divided into 5 groups: a control group, and then a 2 by 2 set-up of treatments. The first treatment was single vs multiple goals. In the single goal condition, households were told to save more as it would help finance their children’s education. In the multiple goal condition, they were provided with two additional savings goals: to save more for future healthcare needs, and for a nest-egg for retirement. Crossed with this was the second treatment of envelope or no-envelope. The envelope was used to give households a way to earmark or designate a mental account for savings.
The authors then survey the households every two weeks for six months, and use as the dependent variable total savings as a percent of total income over this six month period. Here are the results:
Source: Figure 1 in Soman and Zhao.
So in the control group, households only save 3.5 percent of their income, which increases to three times this level when they are given a single goal and an envelope to help them earmark savings – a remarkable increase in savings rates.
A typical field experiment would stop here, with an interesting result, but difficulty in distinguishing what the underlying mechanism for this result could be among several hypotheses. In particular, the authors theory is that the reason for the superiority of a single goal is that having multiple goals gets people thinking about trade-offs and that this defers them moving into an implementation mindset. A competing hypothesis is that perhaps getting people to think about more reasons for savings just overwhelms them with how difficult it is to meet all their needs, potentially discouraging them. The authors use a series of lab experiments to probe into the mechanisms underlying their main result.
To rule out the difficulty of goal accomplishment being harder, they manipulate “mindset” by giving specific instructions in the lab setting that get some people to think about the “how” vs the “why” of savings – getting them to think not just of the savings goals, but about how they would go about implementing them – such as what to invest in, how often to contribute, etc. They find that when people are given this prompting to consider implementation, the advantage of the single goal over multiple goals disappears.
In another of the lab experiments, the authors manipulate how easy it is to accomplish the savings goals, by changing the amount of money people have after meeting basic expenses. They show the advantage of the single goal is greatest when the conditions for savings are difficult – i.e. for people with less disposable income. Finally, they also show that if the multiple goals can be framed as all supporting the same goal, then it can work as well as a single goal.
There are many reasons to be cautious about the generalizability of these results given the small samples and strange subject pools, but this works well as a proof of concept. It suggests that for poor people especially, bombarding them with the many reasons to save might be counterproductive, and fewer goals might lead to more savings. Given the large number of new initiatives in place by the Gates Foundation and others to encourage savings among the poor, this seems like a paper worth building on.
For further reading: the article is in a special issue of the Journal of Marketing Research on consumer financial decision making, with a number of other interesting papers as well. Content is currently all available online .