holly.dykstra
Faculty website
CV
Dr. Holly Dykstra is an Assistant Professor of Economics at the University of Konstanz. Her research studies how people navigate complex decisions and institutional settings using field and lab experiments, administrative records, and survey methods. She works in behavioral economics, public economics, and household finance.
She is a research affiliate of the Centre for Economic Policy Research (CEPR), as well as a faculty affiliate of the Harvard University STAR Lab and the Thurgauer Wirtschaftsinstitut. In her research, she often runs field experiments in cooperation with partner organizations. She has conducted research with the Behavioral Insights Team, the Mayors Innovation Project, a number of private companies, as well as city, state, and national government agencies in the U.S. and Europe.
She holds a Ph.D. from Harvard University and a B.A. from Columbia University. Prior to graduate school, she worked at the U.S. Federal Reserve Board in Washington, D.C.
[Manuscript]
[Appendix]
Behavioral interventions often focus on reducing friction to encourage behavior change. In contrast, we provide evidence that adding initial friction can increase follow-through behavior. In a pre-specified field experiment with the Oregon Department of Transportation (*n* = 27,227), we tested whether adding friction during sign-up for a carpooling platform increases usage. While a more effortful sign-up process led to 25% fewer sign-ups, overall usage increased. Importantly, these results were only partly explained by selection: using an intent-to-treat analysis, participants randomly assigned to the more effortful sign-up took 1.6 times more carpool trips per week over four months, leading to more overall trips despite fewer users. In a second, pre-specified experiment with online task work, participants assigned to a more effortful sign-up were 37% more likely to return the next day and complete more work overall. These results suggest that adding friction may be an overlooked strategy when follow-through, rather than initial uptake, is the primary goal.
[Manuscript]
Income-based rents, common in public housing, create an earnings disincentive. We study a policy designed to counteract this effect by returning part of the rent induced by higher earnings to residents. Importantly, the program automatically enrolled households and used a behaviorally informed design to make the increased payoff to working salient. Using a difference-in-differences approach, we estimate that annual household-head earnings rise 17% (~$1,370/year) and use of public assistance falls 7.5%, suggesting a salient intervention can successfully offset the earnings disincentive found in prior work. We document employment impacts on both the intensive and extensive margins.
[Manuscript]
[CEPR Discussion Paper]
[Appendix]
Agents insist on agency for many reasons, including a desire to obtain their preferred choice and a desire for control. Agents may also forgo agency because they desire to avoid the costs of considering a decision. In a large experiment, we present decision-makers with a menu of investment options and investigate whether they insist on agency or let someone else choose for them. We find that requiring individuals to consider their choice makes them more willing to forgo agency. This consideration effect arises even among experienced decision-makers and across a range of decision-maker characteristics. While we observe few differences in participants' willingness to forgo agency across demographic groups, people expect many differences to arise, including that women would be more willing to forgo agency than men.
[Manuscript]
[Appendix]
Individuals often behave impatiently when making financial decisions for the future. This paper proposes and tests that the timing of decisions relative to payday, which leads to temporary but recurring conditions of scarcity, influences their choice. In a large pre-registered experiment, I ask participants to adopt a commitment device that binds them to being patient. Participants who make this decision eight days before their payday, rather than one day after payday, are 34% more likely to take up commitment. This coincides with when individuals experience the most financial scarcity, and provides evidence that intertemporal decisions are affected by current psychological states.
[Publication]
Dramatic reductions in carbon emissions must take place immediately. A human-centric method of reducing environmental impacts is to “nudge” people away from single-occupancy vehicles (SOVs) toward more sustainable commuting options. While an abundance of research has focused on external determinants of mode choice, we know much less about the behavioral determinants. The field of behavioral science is overdue for a focus on transportation. This paper is meant to facilitate communication between researchers, practitioners, and policymakers in part by developing a behaviorally-informed framework that can be leveraged by policymakers, government, and organizations worldwide. We also describe the founding of our multidisciplinary team and outline lessons learned.
60% of Americans eligible for unemployment benefits are not receiving them (Forsythe and Yang, 2021; BLS, 2023; Lachowska et al., 2025). Amid widespread low and differential take-up rates of social programs (Ko and Moffitt, 2022) and lack of new solutions to address this (DellaVigna and Linos, 2022; Linos et al., 2022), we design a tailored generative AI assistant to overcome the barriers workers face in applying for benefits. We have assessed its effectiveness in a series of pilots.