Financial Concerns, Cognitive Abilities, and Economic Decisions

Study Context

A growing body of literature in psychology and economics is paying attention to the role of mental bandwidth and cognitive function in economic decisions (Enke, 2024). Mental bandwidth has been described as the brain’s ability to perform the basic functions that underlie higher-order behavior and decision-making (Schilbach et al., 2016). Broadly, these are cognitive functions such as attention, memory, and executive control, among others. Researchers theorize that bandwidth can be depleted, allocated, and capped as a function of one’s environment or life circumstances (Mullainathan and Shafir, 2013). One such circumstance is extreme poverty. A life in poverty is associated with worries, stress, and anxiety from the uncertainty of being able to afford essential needs: minimal amounts of food, basic hygiene products, and medicine. The research team refers to the overarching group of psychological hardships related to poverty as “financial concerns.” The burden imposed by financial concerns on individuals’ mental bandwidth can arguably affect their choices and behavior in economically meaningful decisions. Whether these psychological effects exist, their direction, and economic relevance remain open questions in the literature and have direct applications to social policy design. In this project, the research team studies the effects of alleviating financial concerns among the very poor on real-stakes economic decisions.

Study Design

The research team conducts a lab-in-the-field experiment with low-income unemployed adults from the largest marginalized communities in Nairobi, Kenya. The experimental design exogenously varies the timing of a future, unconditional, and high-stakes transfer to study participants. This design isolates financial concerns from consumption and wealth effects. They keep wealth the same across participants by adjusting future payments to the local interest rate, and they keep consumption the same at the baseline, pre-intervention level by collecting data before any payment is made.

Lab activities happen over two consecutive days. At the end of the first day, we inform participants of their experimental assignment. Those in the money-sooner (MC) group are told they will receive an unconditional transfer of KES 2,000 (approximately USD 15) at the end of the second day. Those in the money-later (ML) group are told they will receive an unconditional transfer of KES 2200 (USD 16.35, 9% larger) but 31 days later. On the second day, the research team measures their performance on a broad set of choice and labor tasks.

The transfer amount is meaningful. It is equivalent to a week’s worth of household expenses. The certainty of this incoming transfer is important. Fifty percent of the sample say their savings can only last them two weeks before they need to sell assets to pay for basic needs. In sum, the underlying assumption in the design is that participants in the money-sooner group experience a more significant decrease in immediate financial concerns relative to those in the money-later group. Since all measurements are taken before participants receive any transfer and the transfers are adjusted for time-discounting, changes are interpreted in performance across the two groups as the purely psychological effects of one knowing her immediate financial needs will be met.

Results and Policy Lessons

The results answer two primary questions: (a) Do financial concerns impact whether individuals follow canonical revealed preference theory axioms? (b) Do financial concerns impact productivity?

Answer to (a). They find that the more financially concerned group violates rationality axioms less often than the more financially concerned group: They are less consistent in choosing food bundles and health insurance plans. These results have economic significance. They are real-stake choices with repercussions to participants’ lives outside the lab. In both these exercises, the differences across groups only appear after we allow participants to review and change their initial choices. The two groups are statistically equal in their first choices. When revisions are allowed, the money-later group changes its selections towards rationality. 

Answer to (b). The research team uses six piece-rate labor tasks with varying demands on participants’ cognitive abilities and effort exertion to examine the differences in productivity. In a task that requires purely mechanical effort and no cognitive ability, the money-sooner group’s average performance is 21% lower than its counterpart. They find precisely estimated null treatment effects in all tasks requiring medium to high cognitive faculties. 

Altogether, these results shed light on a complementary relationship between cognitive abilities and effort. The research team capture this relationship with a mathematical model developed in the paper. The findings speak directly to the design and implementation of cash-transfer programs in low-income communities, particularly with respect to the size and frequency of payments.