Digital loans are a source of fast, short-term credit for millions of people. While digital credit broadens market access and reduces frictions, default rates are high. We study the role of the speed of delivery of digital loans on repayment. Our study uses unique administrative data from a digital lender in Mexico and a regression-discontinuity design. We show that reducing loan speed by doubling the delivery time from ten to twenty hours decreases the likelihood of default by 21%. A number of the plausible mechanisms could explain these results. The one that best support our data is that following an unexpected delay, borrowers switch their preferred use of loans in a way that improves the chances of repayment. Our findings suggest that waiting periods used to selectively slow down credit could improve lender profitability and help consumers avoid default.

We conduct a field experiment with low-income shoppers to study how be- havioral interventions can improve the effectiveness of healthy food subsidies. Our unique design enables us to elicit choices and deliver subsidies both before and at the point of purchase. We examine the effects of two non-restrictive changes to the choice environment: giving shoppers agency over what subsidy they receive and introducing a waiting period before the shopping trip to prompt deliberation about their purchases. Combined with healthy food subsidies, these interventions increase healthy food spending by 61% more than a healthy food subsidy alone, resulting in 199% greater healthy spending than in our control group.

Digital credit that uses non-traditional scoring techniques has already expanded credit access to many new populations. A point of policy debate is whether digital lenders should be fully integrated into information sharing systems, such as those offered by credit reference bureaus (CRBs). We study an example of a digital lender in Mexico adopting credit bureau scores into their screening process. Using unique administrative data, we estimate a regression discontinuity in time around the lender’s integration of credit bureau scores. We find that the acquisition of credit scores reduces defaults, with the likelihood of borrowers’ repayment increasing by 10-13%.

We study the impact of deliberation on intertemporal choices. Using multiple experiments, including a field study in the Democratic Republic of Congo, we show that the introduction of waiting periods—a policy that temporally separate information about choices from choices themselves—cause substantially less myopic decisions. These results cannot be captured by models of exponential discounting nor present bias. Comparing the effects of waiting periods to making planned choices over future time periods, the former has a larger impact on reducing myopia. Our results highlight the role of deliberation in decision-making and have implications for policy and intervention design.

Previous research shows that the way transfer income is disbursed can affect what house- holds purchase with that income. In this paper, I provide evidence that disbursement tech- nique can affect the timing of purchases as well. I examine the U.S. Supplemental Nutrition Assistance Program (SNAP), which switched on a state-by-state basis from cash-similar food coupons to Electronic Benefit Transfer (EBT) –a secure debit card– from 1993-2004. I find that EBT mitigated boom-and-bust cycles in food spending associated with SNAP disbursement. This effect is entirely driven by households with children (about two-thirds of the SNAP pop- ulation), who experienced more severe cycles prior to EBT. The effect operates only through the intensive margin –the amount spent on food during a shopping trip– and not at all on the extensive margin –the likelihood of going food shopping.

Time preferences have been correlated with a range of life outcomes, yet little is known about their early development. We conduct a field experiment to elicit time preferences of over 1,200 children ages 3-12, who make several intertemporal decisions. To shed light on how such primitives form, we explore various channels that might affect time preferences, from background characteristics to the causal impact of an early schooling program that we developed and operated. Our results suggest that time preferences evolve substantially during this period, with younger children displaying more impatience than older children. We also find a strong association with race: black children, relative to white or Hispanic children, are more impatient. Finally, assignment to different schooling opportunities is not significantly associated with child time preferences.

Disentangling effort and luck is critical when evaluating outcomes. In a principal-agent experiment, we demonstrate that principals’ judgments of agent effort are biased by luck, despite perfectly observing the agent’s effort. This bias can erode the power of incentives to stimulate effort when there is noise in the outcome process. As such, we explore two potential solutions to this “outcome bias” – control over irrelevant information about luck, and outsourcing judgment to independent third parties. We find that both are ineffective. When principals have control over information about luck, they do not avoid the information that biases their judgment. In contrast, when agents have control over the principal’s information about luck, agents strategically manipulate principals’ outcome bias to minimize their punishments. We also find that independent third parties are just as biased as principals. These findings indicate that outcome bias may be more widespread and pernicious than previously understood. They also suggest that outcome bias cannot be driven solely by disappointment nor by distributional preferences. Instead, we hypothesize that luck directly affects principals’ beliefs about agents even though effort is perfectly observed. We elicit the beliefs of third parties and principals and find that lucky agents are believed to exert more effort than identical, unlucky agents.

We report experimental results for a twice-played prisoners’ dilemma in which the players can choose the allocation of the stakes across the two periods. Our point of departure is the assumption that some (but not all) people are willing to cooperate, as long as their opponent is sufficiently likely to do so. The presence of such types can be exploited to enhance cooperation by structuring the twice-played prisoners’ dilemma to “start small,” so that the second-stage stakes are larger (but not too much larger) than the first-stage stakes. We compare conditions where the allocation of stakes is chosen exogenously to conditions where it is chosen by the players themselves. We show that players gravitate toward the payoff maximizing strategy of starting small in a twice-played prisoners’ dilemma. Intriguingly, the salutary payoff effects of doing so are larger than those that arise when the same allocation is exogenously chosen.

Monthly welfare programs such as the Supplemental Nutrition Assistance Program (SNAP) produce consistent cycles of expenditure and consumption amongst recipients. Food insecurity, health status, crime, poor behavior and test scores track these cycles. This paper leverages new data from the USDA –the FoodAPS survey—to better understand these cycles in three ways. First, I find that expenditure and consumption cycles are correlated within households–a fact not previously established. Second, I study diet quality over the benefit month, and find that it worsens, potentially compounding the harmful consequences of cyclical food insecurity. Third, I find that kids bear less of the burden of the consumption cycle than adults, and that much of this difference may be driven by school-meal programs. This finding suggests large potential gains in child welfare from expanding summer meal programs.

A large literature in psychology studies the effects of the immediate decision environment on behavior, and conceptualizes both cognitive capacity and self-control as scarce resources that can be depleted by recent use, and replenished by factors like rest and nutrition. We assess the relevance of resource-depletion models for intertemporal financial decisions by estimating the effects of three interventions –prior impulse-controlling activity, consumption of a sugared drink, and consumption of a placebo (sugar-free) drink-- on intertemporal monetary choices in a cash-advance framework. These manipulations have large impacts on the demand for advances, but contrary to resource-based models prior impulse-controlling activity and placebo drink consumption increase patience. To understand these effects, we estimate treatment effects on the three parameters of a decision utility model for every subject in our sample. All treatments reduce utility curvature and present-bias, and these movements are highly correlated. Together, we argue that these patterns suggest that the treatments are acting not on subjects’ fundamental utility parameters but on subjects’ tendencies to frame financial decisions narrowly (within the frame of the lab experiment) versus broadly (in the context of their other financial options). Thus, while decision environments have large effects on intertemporal financial decisions, both the direction and the mechanisms underlying these effects appear to be quite different from those suggested by resource-depletion models.

Eliciting time preferences has become an important component of both laboratory and field experiments, yet there is no consensus as how to best measure discounting. We examine the predictive validity of two recent, simple, easily administered, and in- dividually successful elicitation tools: Convex Time Budgets (CTB) and Double Mul- tiple Price Lists (DMPL). Using similar methods, the CTB and DMPL are compared using within- and out-of-sample predictions. While each perform equally well within sample, the CTB significantly outperforms the DMPL on out-of-sample measures.

We propose a new organizing scheme for classifying types of experiments. In addition to the standard categories of laboratory and field experiments, we suggest a new category: “extra-laboratory experiments.” These are experiments that have the same spirit as laboratory experiments, but are conducted in a non-standard manner. We also suggest some organizing principles to distinguish amongst the three types of experiments.

In this article we explore the issues that surround within-subject and between-subject designs. We describe experiments in economics and in psychology that make comparisons using either of these designs (or both) that sometimes yield the same results and sometimes do not. The overall goal is to establish a framework for understanding which critical questions need to be asked about such experimental studies, what authors of such studies can do to ameliorate fears of confoundedness, and which scenarios are particularly susceptible to divergent results from the two approaches. Overall, we find that both designs have their merits, and the choice of designs should be carefully considered in the context of the question being studied and in terms of the practical implementation of the research study.