How do psychological and economic factors shape behavior when resources are constrained?
Psychology and Economics of Poverty Convening
On May 17th 2019, CEGA’s second annual Psychology and Economics of Poverty Convening brought together psychologists and economists to share research on how poverty affects cognition, decision-making, individual bias, and health in low- and middle-income countries. The day started with lightning talks presented by current PhD students and proceeded into four sessions on areas of intersection between the disciplines. Presentations spurred interesting discussions about how psychological scales and tools can be used and validated in behavioral economics and across cultural contexts; the relationship between poverty and cognitive bandwidth; and how best to conceptualize and measure dimensions of scarcity. Read on for brief session summaries, some key takeaways, and reflections from the day.
How do beliefs shape economic behavior?
In development economics, the prevailing paradigm for understanding and addressing the root causes of poverty focuses on external constraints. After all, external factors — such as low wages, bad credit markets, or limited access to education — determine individual preferences under standard economic theory, which assumes people will always do what is best for themselves given the information they have. At the convening, researchers added to a new and growing body of literature that departs from this framework. Their presentations explored and modeled the internal, psychological factors that affect preferences, including how people form beliefs — whether about themselves, others, or even technology. Key takeaways from this session include:
· Having lower access to high-return technologies (e.g. medicine, education, healthcare) reduces belief in their effectiveness — independent of learning. According to standard economic theory, people base their beliefs on information alone. Presenter Vinayak Alladi (UC San Diego) shared results that suggest otherwise. In a lab setting, those who had limited access to a good or service tended to avoid disappointment by tempering their beliefs about how effective it is. These findings were corroborated with a field experiment: when public school students in Los Angeles were offered an SAT prep package for free with randomly assigned probabilities of receiving it (25 percent vs. 75 percent), those in the group with lower probability of access to the SAT prep package were less optimistic about how much the package would improve their scores. What Alladi terms a “sour grapes” effect is found to be more prevalent among the poor. (Alladi 2018: paper, presentation)
· An intervention designed to increase ‘aspirational hope’ among microentrepreneurs improved their psychological outcomes but had limited effects on economic outcomes one year later. Presenter Bruce Wydick (UC San Francisco) provided results from a randomized experiment designed to generate greater “hope” for microenterprises among women in Oaxaca, Mexico. Wishful hope (“I hope that”) may motivate fatalism and complacency, whereas aspirational hope (“I hope to”) motivates proactivity. To increase the latter, study participants went through a 12-month intensive program targeting “three essential elements of hope from the positive psychology literature: aspirations, agency, and pathways”. Results one year later are mixed: the intervention has strong effects on psychological outcomes (including agency, happiness, and future-orientation); impacts on economic outcomes (sales, profits, savings) are more modest, but larger for lower-income, less initially-hopeful women. (Wydick et al. 2018: presentation)
· Beliefs about the mental capacity of charity recipients affects whether donors choose “paternalistic” or “agentic” aid. Presenter Juliana Schroeder (UC Berkeley Haas School of Business) shared findings from an experiment in which subjects were given the choice of what type of aid to give — whether to the American Red Cross (“paternalistic” aid that restricts the recipients’ freedom to choose) or GiveDirectly (“agentic” aid, e.g. no-strings-attached cash transfers given directly to recipients). People chose more agentic giving when recipients were described as “enterprising entrepreneurs” (implying higher mental capacity) and gave paternalistically when informed recipients were “unemployed” (implying lower mental capacity). In another experiment, the authors also found that people endorse more paternalistic policies for the “average citizen” than for themselves, but then re-adjust towards paternalism for themselves after they are primed to recall shortcomings in their own decision-making. (Schroeder, Epley, and Waytz 2017: paper, presentation)
How does economic insecurity affect growth and development?
Scarcity occurs when resources are more limited than people’s need for them. People in conditions of material scarcity will need to sacrifice certain things they need in order to afford the most important things. However, scarcity affects poverty beyond a direct income or price effect. This session explored how material scarcity can affect cognitive or psychological scarcity, potentially resulting in lower mental bandwidth to devote to other cognitively-intensive tasks, such as remembering immunizations and adhering to medications. These poor decisions can perpetuate the cycle of poverty. Key takeaways from this session are as follows:
· Conditions of scarcity can improve certain types of decision-making (i.e. people become less “loss-averse”) but effects on cognition are mixed. Kelsey Jack (UC Berkeley) presented findings from observations of 3000 farmers in Zambia over 15 months, leveraging natural variation in scarcity based on the timing of farmers’ harvests and additionally introducing design-based variation in scarcity in a lab-in-field experiment. Decision-making was measured using a game: farmers were given a good (washing powder) and later had the opportunity to trade it for other goods (mimicking Kahneman et al.’s famous mug experiment). Decision-making preferences were measured by farmers’ unwillingness to trade the washing powder for a higher value item, a principle known as “loss aversion.” Cognition was measured through tests of fluid intelligence. Results were mixed: conditions of scarcity reduced loss aversion and actually improved decision-making. The effect of scarcity on cognition was less clear — individuals performed better on some tests, but not others. These findings suggest that scarcity, in its different forms, may affect various decisions or cognitive functions in distinct ways. (Jack, Fehr, and Fink. 2019: presentation)
· Using money examples in basic math questions may not be the best way to get poor students to learn math. Claire Duquennois (UC Berkeley) shared experimental results in which students from lower socio-economic backgrounds scored 0.1–0.2 standard deviations lower on tests when they were given a test booklet containing more basic math questions using money examples. They were less likely to correctly answer the question containing the money example and were also less likely to get the subsequent four questions correct. The implication is that bringing money matters to the top of mind can affect performance via attention capture. (Duquennois 2018)
How do scarcity and changes in income affect psychology and decision-making?
Income shocks — positive or negative changes to income caused by an unexpected external source — and resource scarcity can influence cognition and individual choices through a variety of mechanisms. Using experimental methods, researchers explored how both expected and unexpected financial strain impacts personality traits, parent-child interactions, and worker productivity, among other outcomes. They also discussed processes for systematically replicating studies on the psychological effects of scarcity and the implications of this work. Key takeaways include:
· Economic shocks can alter (short-run) personality traits. Personality affects economic outcomes: traits like conscientiousness, “grit,” self-control, and emotional stability predict success in school, income, and job performance. Shikar Mehra (UC San Francisco) and colleagues measured the short-term impacts of both a positive income shock (income-generating asset distribution) and a negative one (exposure to drought) on personality using the “Big Five” personality traits among ultra-poor households in Uganda. They found that the former increased traits representing socialization and stability by up to 0.21 SDs while drought decreased scores on these traits by up to 0.26 SDs. Neither affected traits linked to personal growth and plasticity. (Mehra, Stopnitzky, and Alloush 2018: paper, presentation)
· Parents talk to their kids less in periods of monetary scarcity and when they are reminded about scarcity. Poor kids hear 30 million fewer words by age 3 compared to wealthy ones; this deficit contributes to an early gap in vocabulary between poor and rich children. Monica Ellwood-Lowe shared results from a lab study in which parents who were primed to think about resource scarcity and then reunited with their children spoke with them less, compared to parents in the control group. In a field study, authors listened in on parent-child interactions via LENA recorders over a calendar month (financial scarcity peaks late in the month). Initial results suggest that fewer conversational turns (utterances by child/parent followed by parent/child) occur later in the month. This effect is driven by fewer child vocalizations. (Ellwood-Lowe 2019: presentation)
Alleviating financial strain makes people work harder (and smarter!)Supreet Kaur (UC Berkeley) described results from a field experiment in which factory workers in India were given piecemeal rates for each woven plate they produced. Some were randomly assigned to receive an early payout for their work; others received the payout on the last day of the study. Initial results found that early cash group increased productivity by 5 percent, with larger increases seen among poorer workers. This increase was driven by workers making fewer mistakes, suggesting that reducing financial worry boosted productivity via increased attention and focus. (Kaur et al. 2019: presentation)
Replicating studies on scarcity present interesting and particular challenges. Ruthe Foushee (UC Berkeley) and colleagues shared results from an ongoing systematic analysis and attempt to replicate results from studies on the psychological effects of scarcity. Adding to the conversation on open science and replicability, they found that some results were replicable, while others were not. Their findings lead us to further questions about how to measure scarcity as well as avenues for continued study replication and the interpretation of those results. (Foushee et al. 2019: presentation)
In addition to the takeaways above, the convening has illuminated some important open questions at the intersection of psychology and development economics. For example:
1. To what extent is the relationship between poverty/scarcity and cognitive bandwidth a direct one? Or does scarcity affect cognition primarily through indirect channels our metrics are not capturing as well, such as nutrition, interpersonal conflicts, or fatigue from lack of sleep?
2. Should there be a policy response and if so, what should it be? Is the policy response simply: alleviate financial constraints and all of these positive psychological externalities we previously didn’t know about will follow? Or as in Wydick’s paper, should we also intervene on internal psychological processes like “hope”? I [Crystal] found an off-session discussion with the psychologist Juliana Schroeder fascinating: she mentioned Maslow’s hierarchy of needs is a misconception. We often think that people need to meet some baseline physiological requirements first before they can consider higher-level needs of esteem and self-actualization, but in actuality people can experience parallel needs: both hunger and need for self-respect. This relates broadly to a new literature on mental health and depression policies in low-income conferences (not covered in this conference) — like this upcoming AER paper that shows cognitive behavioral therapy can have huge and long-lasting effects on economic decision-making. My sense is that the research for low-income countries is still nascent and more evidence is needed on how much mileage we can get out of these interventions (on internal processes) relative to other policies, before more widespread adoption.
We look forward to seeing how CEGA’s new Psychology and Economics of Poverty initiative — and the related body of evidence — continues to evolve.