Here you will find a full list of seminar speakers, along with the title, description, and presenter biography.
Please check back in early September for the most up to date list.
We evaluate the effect of introducing a behaviorally informed notification, that uses salience, framing, social norms, and loss aversion, on the timeliness of responses of civil servants in the City of Buenos Aires to information requests. We also examine the effect of a civil service training program and how it interacts with the treatment. The results indicate that the behavioral intervention changed the distribution of responses and increased the share of requests answered exactly on the deadlines. The training workshops decreased timeliness for both the treated and control groups. In particular, treated participants who attended workshops with a larger share of control participants took longer to respond.
What information would people like to have? What information would they prefer to avoid? And how does the provision of information bear on people’s welfare? Nationally representative surveys of eleven nations find that substantial percentages of people do not want to receive information even when it bears on personal health, on sustainability, and on consumer welfare. At the same time, people’s willingness to pay for information, contingent on their wanting it, is almost always higher than people’s willingness to pay not to receive information, contingent on their not wanting it. We develop a model and estimate the welfare effects of information for selected items. We find substantial benefits and costs, with the former outweighing the latter.
This study examines the mechanisms underlying long-run reductions in energy consumption caused by a widely studied social nudge. Our investigation considers two channels: physical capital in the home and habit formation in the household. Using data from 38 natural field experiments, we isolate the role of physical capital by comparing treatment and control homes after the original household moves, which ends treatment. We find 35 to 55 percent of the reductions persist once treatment ends and show this is consonant with the physical capital channel. Methodologically, our findings have important implications for the design and assessment of behavioral interventions.
In three large-scale field experiments with over 32,500 individuals, we investigate whether public transport uptake can be influenced by behavioral interventions and by economic incentives. Despite their effectiveness in other domains, we find a tightly estimated zero for social norms and implementation intentions on ridership. Increasing the economic incentive significantly increases uptake and long-term usage. This increase is sustained for months after removing the incentive. The effect is mainly driven by initial low users, which is evidence for habit formation and highlights the heterogeneous effects of the policy. While there is scope for long-term behavior change, nudging might not be the right approach.
This paper uses a natural field experiment to examine the effectiveness of specific nudges on tax compliance amongst firms and the self-employed in the Dominican Republic. In collaboration with the Dominican Republic’s tax authority, we designed messages for more than 28,000 self-employed workers and over 56,000 firms. Leveraging administrative tax data, we find evidence that our nudges (increasing the salience of prison sentences or public disclosure of tax evaders) have large effects on increasing tax compliance, primarily working through the channel of decreasing claimed tax exemptions. Interestingly, we find that firms are more impacted than the self-employed, and that firm size is critically linked to nudge effectiveness: larger firms are considerably more influenced by nudges than smaller firms. We find this latter result noteworthy given the paucity of evidence showing significant behavioral impacts of nudges amongst the largest players in a market. Overall, our messages increased tax revenue by $193 million (roughly 0.23% of the Dominican Republic’s GDP in 2018), with over $100 million constituting income that the government would not have received without our field experimental nudges.
Public recognition is a frequent tool for motivating desirable behavior, yet its welfare effects are rarely measured. We develop a portable money-metric approach for measuring the direct welfare effects of shame and pride, which we deploy in a series of experiments on exercise and charitable behavior. In all experiments, public recognition motivates desirable behavior but creates highly unequal emotional consequences. High-performing individuals enjoy significant utility gains from pride, while low-performing individuals incur significant utility losses from shame. We estimate structural models of social signaling, and we use the models to explore the social efficiency of public recognition policies.
In most organizations, promotion into leadership typically requires self-nomination and competition via an application. However, past research on gender differences in self-promotion and preference for competition suggests that the default opt-in process might result in fewer women choosing to compete for promotions. Drawing from well-established findings on behavioural nudges, we hypothesized that changing promotion schemes from a default where applicants must opt-in (i.e., self-nominate) to a default where applicants must opt-out (i.e., those who pass a qualification threshold are automatically considered for promotion, but can choose not to be considered) will attenuate gender differences in participation in competition. Across a series of studies, we find that women are less likely to compete and put themselves forward under the traditional opt-in framing. However, when the choice is described using opt-out framing, gender differences are eliminated, and men and women choose to remain in the competition pool at equal rates. These results suggest that organizations could make use of “opt-out” promotion schemes as a behavioral intervention to reduce the gender gap in promotion rates and ascension to leadership positions.
We use evidence from two large-scale field experiments in Mexico (N=131,300) to test the effectiveness of a previously undocumented behavioral intervention designed to improve retirement savings contributions: framing retirement savings as a way to ‘secure your family’s financial future’. In our first experiment (N=34,151), we find that a family security SMS intervention significantly improves the likelihood of making a voluntary contribution relative to a status quo, control condition by nearly double. In our second field experiment (N=97,149), we benchmark the effectiveness of a family SMS intervention against SMS’ that contain other previously successful behavioral interventions (e.g., basic alerts, pennies-a-day, fresh start, individual goal security). Once again, we find that a family security SMS intervention significantly increases voluntary contribution savings overall, and prove either just as effective, if not more so effective than these previously used behavioral approaches. However, we find that while our treatment significantly improves contributions in the aggregate, there is significant heterogeneity in treatment response. Using recently developed machine-learning techniques to quantify the heterogeneity in the treatment, we find that women under the age of 29 were nearly 67% less likely to contribute when given this family intervention compared to no intervention at all, while individuals between 29-41 were 68% more likely to contribute, regardless of gender. The heterogeneity in the treatment effects has important managerial implications: We demonstrate that sending this intervention to all account holders would increase profits by $375,800 pesos ($19,032 USD). In contrast, a strategic application of this intervention to profitable individuals only would increase profits by $4,887,000 pesos ($246,000 USD). From a consumer welfare perspective, this strategic application would also generate 4,597 new voluntary retirement contributors (the blanket approach would generate 2,340 new contributors).
Inequality in access to high-quality teachers is an important driver of student socioeconomic achievement gaps. We experimentally evaluate a novel nation-wide low-cost government program aimed at reducing one of its causes: teacher sorting. Specifically, we tested two behavioral strategies designed to motivate teachers to apply to job vacancies in disadvantaged schools. These strategies consisted of an "Altruistic Identity" treatment arm, which primed teachers’ altruistic identity by making it more salient, and an "Extrinsic Incentives" arm, which simplified the information and increased the salience of an existing government monetary-incentive scheme rewarding teachers who work in underprivileged institutions. We show that both strategies are successful in triggering teacher candidates to apply to such vacancies, as well as make them more likely to be assigned to a final in-person evaluation in a disadvantaged school. The effect among high-performing teachers is larger, especially in the "Altruistic" arm. Our results imply that low cost behavioral strategies can enhance the supply and quality of professionals willing to teach in high-need areas.
The COVID-19 pandemic represents a global health crisis, yet certain countries have had far more success in limiting COVID-19 cases and deaths. We suggest that collective threats require a tremendous amount of coordination, and that strict adherence to social norms is one key mechanism that enables groups to do so. Here we examine how the strength of social norms—or tightness-looseness—was associated with countries’ success in limiting cases and deaths by Fall 2020. Tight cultures have strict norms and punishments whereas loose cultures have weaker norms and are more permissive. We expected that tight cultures would have fewer cases and deaths per million compared to loose cultures. We estimated the relationship between cultural tightness and COVID-19 case and mortality rates as of October 16th using OLS regression. We fit a series of stepwise models to capture whether cultural tightness explained variation in case and death rates above and beyond other covariates including measures of underreporting, wealth, inequality, government efficiency, population density and size, mandated BCG vaccination, average age, climate, and other cultural factors. The results indicated that nations with high levels of looseness had 3.09 times the cases and 6.16 times the deaths as compared to nations with high levels of tightness. Specifically, nations with high levels of looseness had an average of 12,060 cases and 357 deaths per million whereas nations with high levels of tightness had an average of 3,903 cases and 58 deaths per million. These results were robust to a number of sensitivity analyses which controlled for underreporting, demographics, geopolitical factors, other cultural dimensions, and climate. A formal evolutionary game theoretic model suggested that tight groups coordinate on cooperative norms much faster and have higher survival rates than loose groups. The results suggest that tightening social norms confers an evolutionary advantage in times of collective threat.
Political polarization has ruptured the fabric of U.S. society. The focus of this paper is to examine various layers of (non-)strategic decision-making that are plausibly affected by political polarization through the lens of one's feelings of hate and love for Donald J. Trump. In several pre-registered experiments, I examine three layers of polarization in novel ways across multiple settings: I document the behavioral-, belief-, and norm-based mechanisms through which perceptions of interpersonal closeness, altruism, and cooperativeness are affected by polarization, both within and between political factions. To separate ingroup-love from outgroup-hate, the political setting is contrasted with a minimal group setting. I nd strong heterogeneous effects: ingroup-love occurs in the perceptional domain (how close one feels towards others), whereas outgroup-hate occurs in the behavioral domain (how one helps/harms/cooperates with others). In addition, the pernicious outcomes of partisan identity also comport with the elicited social norms. Noteworthy, the rich experimental setting also allows me to examine the drivers of these behaviors, suggesting that the observed partisan rift might be not as forlorn as previously suggested: in the contexts studied here, the adverse behavioral impact of the resulting intergroup conflict can be attributed to one's grim expectations about the cooperativeness of the opposing faction, as opposed to one's actual unwillingness to cooperate with them.
Individuals often must decide not only whether to act, but also the quantity or magnitude of that action. Athletes consider not only whether to run, but for how long; shoppers consider not only whether to buy toilet paper, but also the number of rolls; individuals consider not only whether to contribute to retirement savings, but also how much to contribute. This research compares two choice formats: quantity-sequential—which asks individuals first whether they’d like to act, and then to choose the quantity—and quantity-integrated—which asks individuals to simultaneously indicate whether and how much to act. In the lab and the field, we find that the quantity-integrated format substantially increases action. We investigate the psychological mechanisms underlying this effect, and discuss how this insight can be leveraged to encourage positive behaviors or discourage harmful ones.
Carrot Rewards was a Canadian health and wellness app that rewarded users with loyalty rewards points for reaching their personalized daily step goals. They were active for three years before closing in June 2019 and mostly operated in British Columbia and Newfoundland-Labrador. The Canadian government invested $5 million to develop the app and purchase reward points, and a recent study of the program's first year revealed that users who used the app for at least six months showed an increase of 3,141 to 6,192 steps per week. However, questions remain as to whether the program would have been cost-effective if it continued to operate. In this study, we input data from the one-year study into a Markov state transition model to examine the cost-effectiveness of the Carrot Rewards Steps program. This talk will seek to answer the following questions: (a) What are the long-term health benefits of using financial health incentives? (b) How much would the program have cost, and would it have saved the public healthcare system in direct medical expenses? and (c) At what level of willingness-to-pay would it have been favourable to keep the program operating? This study has implications for employers that offer health rewards to their employees. Assumptions and limitations of our model are discussed.
The emergence of the Covid-19 pandemic led to an unprecedented effort to change human behaviour at mass level in a short period of time to promote physical distancing, hand-washing and other protective health behaviours. This paper discusses the ethics of applying ideas from behavioural science in this context. Drawing on the FORGOOD framework by Lades and Delaney (2020), we consider issues of fairness, openness, respect, goals, other options, opinions and delegation. The Covid-19 response led to a sometimes confused discussion of the definition of behavioural science, the ties between behavioural science and other interventions and the appropriate deployment of expertise. We discuss these points with illustrations from applications of behavioural science to Covid-19 responses in different countries, focusing on Ireland and the UK. We draw from the examples and the broader ethical literature to derive principles that could inform behavioural science responses to ongoing pandemic policies and future emergencies.
Receiving instrumental help—such as advice, feedback on a task, or a referral to a job—can be critical to success, but past research shows women and racial minorities are less likely to receive such support. Across two field experiments and one online experiment (total N = 5148), we test whether women and racial minorities benefit from explicitly mentioning their demographic identity (e.g., adding “As a woman”) in requests for help. Although highlighting demographic identity could increase the risk that marginalized group members face discrimination, we theorize that such statements activate prospective helpers’ motivations to avoid prejudiced reactions, ultimately increasing their willingness to help women and racial minorities. Consistent with our hypotheses, we find that women and racial minorities benefit when they explicitly mention their identity in a request for help. In related work, we present three pre-registered lab studies showing that women are more likely to be hired in decision-making contexts that make diversity more salient. Our work suggests that highlighting demographic identity—rather than obscuring it—may benefit typically underrepresented candidates.
Across 5 conceptual replications and 1 direct replication, wefail to replicate the finding that signing at the top decreases dishonesty.Given that the original finding is widely promoted as a seminal behavioralinsights intervention, we think it is important to correct the record.Furthermore, all of the original authors joined me in the direct replicationand came together to update the record.
Political identity can shape perceptions of reality and support for government intervention. We implement a multi-waves longitudinal panel survey on the same representative sample of Americans between April and November 2020, during the peak of the pandemic crisis and the presidential elections. In a set of questions we elicit respondents' forecasts of deaths due to COVID-19 in the U.S., and ask them whether they considered it as a successful outcome of government intervention. We observe large differences between Democrats and Republicans on both the forecast estimates and the perception of success. In a subsequent experiment, we ask respondents to report the number of deaths to date in the U.S., and randomly encourage half of the respondents to consult the CDC website prior to answering the question. While the treatment was successful at increasing the likelihood of answering the question correctly, political disparities remain on the assessment of success. In other two experiments on the same sample of respondents, we elicit support for government policies where we vary the saliency of the policy (mandatory masks to prevent contagion versus GMOs) and randomly show whether the policy was endorsed by a not well-known Democratic or Republican politician. We find that respondents are more likely to support a policy when a politician of their same party endorses it, with this effect being stronger among Republicans when the policy is less salient. These results offer several novel insights to the fields of behavioral political economy on how identity and polarization affect perceptions of reality and support for government policies, and provides methodological contributions to experimental economics studies by implementing a set of survey experiments on a longitudinal panel of a representative sample of a country's population.
Digital technologies such as smartphones and social media consume a large and growing share of leisure time. While these technologies provide obvious benefits, it is often argued that they can be addictive and harmful. We lay out a model of digital addiction and estimate its parameters using a randomized experiment involving about 2000 smartphone users. We also measure treatment effects of reduced smartphone use on survey measures of smartphone addiction and subjective well-being.
Many people want to reduce their smartphone usage to increase productivity and well-being, but fail to accomplish this goal. We conduct a randomized control trial with a student population (N=112) over three weeks to test the effectiveness of two widely available digital nudges for screen time reduction. Along with a tracking-only control condition, a passive digital nudge (i.e., grayscale mode) was compared to an active digital nudge (i.e., time limits). The passive nudge led to an immediate, significant reduction of objectively measured screen time compared to the control condition. Conversely, the active nudge led to a smaller and gradual screen time reduction. Those in the control condition, who simply tracked their usage, did not lower their screen time. As opposed to the popular belief that reducing screen time is beneficial, we found no immediate causal effects of reducing screen time on subjective well-being and academic performance.
The Mind, Behavior, and Development Unit (eMBeD), the World Bank’s behavioral science team in the Poverty and Equity Global Practice, works closely with project teams, governments, and other partners to diagnose, design, and evaluate behaviorally informed interventions. By collaborating with a worldwide network of scientists and practitioners, the eMBeD teamprovides answers to important economic and social questions, and contributes to the global effort to eliminate poverty and increase equity.
This two-hour session will focus on recent results from the team on socio-emotional skills focused interventions in both education and labor markets implemented across multiple countries. Presentations will present a ‘deep dive’ for one example, and lessons on implementation, evaluation and measurement from multiple settings including Peru, South Africa, Indonesia, North Macedonia, Turkey, Ethiopia and Mauritius, with a focus on embedding in the existing country delivery systems.
We look forward to a very special webinar week with the World Bank’s eMBeD team sharing research from 5 countries:
# Indonesia http://bit.ly/WB_embed_indonesia
# Peru http://bit.ly/WB_embed_peru
# South Africa http://bit.ly/WB_embed_south_africa
# Mauritius http://bit.ly/WB_embed_mauritius
# Ethiopia http://bit.ly/WB_embed_ethiopia
Does the wish to convince others lead people to persuade themselves about the factual and moral superiority of their position? We investigate this question in field experiments at two international debating competitions that randomly assign persuasion goals (pro or contra a motion) to debaters. We find evidence for self-persuasion in incentivized measures of factual beliefs attitudes and confidence in one’s position. Self-persuasion occurs before the debate and remains after the debate. Our results lend support to interactionist accounts of cognition and suggest that the desire to persuade is an important driver of opinion formation.
We study how individuals' compliance with norms of pro-social behavior is influenced by other actors' compliance in a novel, dynamic, and non-strategic experimental setting. We are particularly interested in the role that social proximity among peers plays in eroding or upholding norm compliance. Our results suggest that social proximity is crucial. In settings without known proximity, norm compliance erodes swiftly because participants only conform to observed norm violations of their peers while ignoring norm compliance. With known social proximity, participants conform to both types of observed behaviors, thus halting the erosion of norm compliance. Our findings stress the importance of the broader social context for norm compliance and show that, even in the absence of social sanctions, compliance can be sustained in repeated interactions, provided there is group identification, as is the case in many social encounters in natural and online environments.
The coronavirus pandemic resembles other scientifically-informed but controversial policy domains (e.g., climate change, vaccination) in that the success or failure of collective efforts has enormous consequences for all. But it is unusual in that neither scientific evidence nor prior lay beliefs existed before a year ago. This novel research environment affords investigations of social proof, knowledge outsourcing, and the efficacy of appeals to the wellbeing of others without the long-entrenched, domain-specific ideological commitments that complicate inferences about beliefs and behaviors in those other domains. I will discuss findings from the past nine months of The Policy Lab’s collaboration with the Rhode Island Department of Health (osf.io/w2nhg) as well as work with independent research groups and our systematic tracking of the rapidly growing gray literature (tpl.fyi/covid-research-tracker).
Anti-price gouging laws are ubiquitous and people take costly actions to report violators to law-enforcement agencies, which suggests that they value punishing price increases during emergencies. We argue with a model that consumer reports contain information about repugnance to price gouging, or willingness to prevent third-party transactions (Roth, 2007). We conduct a field experiment during the first wave of COVID-19 to measure individuals’ willingness to pay to report sellers who increase prices of personal protective equipment. The willingness to pay to report is non-negligible, polarized, and responsive to the seller's price. We also find that repugnance is partly due to distaste for seller profits, depending on the product.
Monetary payments are often used to incentivize behavior. Payment structure and amount are researched extensively within the realm of physical health behavior, but the impact of incentive schemes on mental health behavior remains underexplored. In this experiment, we tested whether choosing one’s own incentive scheme can be used to promote the take-up and maintenance of a mindfulness meditation habit.
Over a period of 36 days, participants can complete daily online guided meditation sessions (5-15 minutes) and receive monetary incentives to do so. We compare a control group (where participants receive access to the meditation sessions but no monetary incentives to complete them) to two incentive treatments where participants additionally receive monetary rewards. In the “exogenous treatment”, participants are assigned one of two incentive schemes, while in the “choice treatment” the participant can choose one of the two incentive schemes: constant or streak. In the constant incentive scheme, participants obtain a constant payment for each day that they complete a meditation session. In the streak incentive scheme, participants earn (comparatively more) for completing meditation sessions on three consecutive days.
We explore the following questions: (1) Can mindfulness meditation habits be created and sustained by providing monetary incentives similarly to other more commonly studied habits (such as gym attendance)?; (2) Does letting participants choose their own incentive schemes increase meditation frequency?; (3) How does the streak compare to the constant incentive scheme?
The COVID-19 pandemic is increasing negative emotions and decreasing positive emotions globally. Left unchecked, these emotional changes may have a wide array of adverse impacts. To reduce negative emotions and increase positive emotions, we will examine the impact of reappraisal, a widely studied and highly effective form of emotion regulation. Participants from 55 countries (expected N = 25,448) will be randomly assigned to one of two brief reappraisal interventions (reconstrual or repurposing), an active control condition, or a passive control condition. We predict that both reappraisal interventions will reduce negative emotions and increase positive emotions relative to the control conditions. We further predict that reconstrual will decrease negative emotions more than repurposing, and that repurposing will increase positive emotions more than reconstrual. We hope to inform efforts to create a scalable intervention for use around the world to build resilience during the pandemic and beyond.