Contributions in Merlin

Decision making often requires integrating self-gathered information with information acquired from observing others. Depending on the situation, it may be beneficial to rely more on one than the other source, taking into account that either information may be imprecise or deceiving. The process by which one source is selected over the other based on perceived reliability, here defined as arbitration, has not been fully elucidated. In this study, we formalised arbitration as the relative reliability (precision) of predictions afforded by each learning system using hierarchical Bayesian models. In a probabilistic learning task, participants predicted the outcome of a lottery using recommendations from a more informed advisor and self-sampled outcomes. The number of points participants wagered on their predictions reflected arbitration: The higher the relative precision of one learning system over the other and the lower the intention volatility, the more points participants wagered on a given trial. Functional neuroimaging demonstrated that the arbitration signal was independent of decision confidence and involved modalityspecific brain regions. Arbitrating in favour of self-gathered information activated the dorsolateral prefrontal cortex and the midbrain whereas arbitrating in favour of social information engaged ventromedial prefrontal cortex and the temporoparietal junction. These findings are in line with domain specificity and indicate that relative precision captures arbitration between social and individual learning systems at both the behavioural and neural level.

Social media has become the modern arena for human life, with billions of daily users worldwide. The intense popularity of social media is often attributed to a psychological need for social rewards (“likes”), which turns the online world into a “Skinner Box” for the modern human. Yet despite such common portrayals, empirical evidence for social media engagement as reward-based behavior remains scant. We applied a computational approach to directly test whether reward learning mechanisms contribute to social media behavior. We analyzed over one million posts from over 4,000 individuals on several social media platforms, using computational models based on reward reinforcement learning theory. Our results consistently show that human behavior on social media qualitatively and quantitatively conforms to the principles of reward learning. Results further reveal meaningful individual differences in social reward learning on social media, explained in part by variability in users’ tendency for social comparison. Together, these findings support the social reinforcement learning view of social media engagement and offer key new insights into this emergent mode of modern human behavior on an unprecedented scale.

Patronage is a byword for poor performance, yet it remains pervasive. We study the selection effects of patronage in the world's most successful navy - the British Royal Navy between 1690 and 1849. Using newly collected data on the battle performance of more than 5,800 naval officers promoted - with and without family ties - to the top of the navy hierarchy, we find that connected promotees outperformed unconnected ones. There was substantial heterogeneity among the admirals in charge of promotions. Discretion over appointments thus created scope for "good" and "bad" patronage. Because most admirals promoted on the basis of merit and did not favor their kin, the overall selection effect of patronage was positive.

The U.S. economy has experienced a significant drop in the fraction of the population employed in middle wage, “routine task-intensive” occupations. Applying machine learning techniques, we identify characteristics of those who used to be employed in such occupations and show they are now less likely to work in routine occupations. Instead, they are either non-participants in the labor force or working at occupations that tend to occupy the bottom of the wage distribution. We then develop a quantitative, heterogeneous agent, general equilibrium model of labor force participation, occupational choice, and capital investment. This allows us to quantify the role of advancement in automation technology in accounting for these labor market changes. We then use this framework as a laboratory to evaluate various public policies aimed at addressing the disappearance of routine employment and its consequent impacts on inequality.

Previous research points to the heritability of risk-taking behavior. However, evidence on how genetic dispositions are translated into risky behavior is scarce. Here, we report a genetically-informed neuroimaging study of real-world risky behavior in a large European sample (N=12,675). We found negative associations between risky behavior and grey matter volume (GMV) in distinct brain regions, including amygdala, ventral striatum, hypothalamus and dorsolateral prefrontal cortex (dlPFC). Polygenic risk scores for risky behaviors, derived from a genome-wide association study in an independent sample (N=297,025), were inversely associated with GMV in dlPFC, putamen, and hypothalamus. This relation mediated ∼2.2% of the association between genes and behavior. Our results highlight distinct heritable neuroanatomical features as manifestations of the genetic propensity for risk taking. One Sentence Summary Risky behavior and its genetic associations are linked to lower grey matter volume in distinct brain regions.

We combined established emotion regulation and dietary choice tasks with fMRI to investigate behavioral and neural associations in self-regulation across the two domains in human participants. We found that increased BOLD activity during the successful reappraisal of positive and negative emotional stimuli was associated with better dietary self-control. This cross-task correlation was present in medial and lateral prefrontal cortex as well as the striatum. These results suggest that neural processes related to the reappraisal of emotional stimuli may also facilitate dietary self-control. However, within the dietary self-control task itself, we did not find that prefrontal cortex (PFC) activity significantly increased with self-control success during our food choice task, in contrast to previous reports. This prompted us to conduct exploratory analyses, which revealed that BOLD activity in PFC tracks the amount of taste and healthiness at stake on each self-control challenge trial regardless of the chosen outcome. This exploratory finding also replicated in an independent dataset. We discuss the implications of this evidence that individuals track the self-control stakes in light of theories about effortful self-regulation. In addition, we discuss features of this version of the food choice task that may have reduced the need to recruit PFC to achieve self-control. In summary, our findings indicate that the neural systems supporting emotion reappraisal can generalize to other behavioral contexts that require reevaluation to conform to the current goal. Significance statement Reappraisal is a prominent strategy for self-regulation. Yet data to compare processes underlying the reappraisal of emotions and dietary self-control within the same individual is lacking. Here, we use two established emotion regulation and dietary choice tasks to compare both on the neural level. We found that increased BOLD activity in several brain regions including medial and lateral prefrontal cortex and striatum during the successful reappraisal of positive and negative emotional stimuli was linked to better dietary self-control. These results suggest that neural processes underlying the reappraisal of emotional stimuli may also facilitate dietary self-control.

We study the long-term effects of a randomized intervention targeting children’s socio-emotional skills. The classroom-based intervention for primary school children has positive impacts that persist for over a decade. Treated children become more likely to complete academic high school and enroll in university. Two mechanisms drive these results. Treated children show fewer ADHD symptoms: they are less impulsive and less disruptive. They also attain higher grades, but they do not score higher on standardized tests. The long-term effects on educational attainment thus appear to be driven by changes in socio-emotional skills rather than cognitive skills.

A random variable is difference-form decomposable (DFD) if it may be written as the difference of two i.i.d. random terms. We show that densities of such variables exhibit a remarkable degree of structure. Specifcally, a DFD density can be neither approximately uniform, nor quasiconvex, nor strictly concave. On the other hand, a DFD density need, in general, be neither unimodal nor logconcave. Regarding smoothness, we show that a compactly supported DFD density cannot be analytic and will often exhibit a kink even if its components are smooth. The analysis highlights the risks for model consistency resulting from the strategy widely adopted in the economics literature of imposing assumptions directly on a dfference of noise terms rather than on its components.

Do firms seek to make the market transparent,or do they confuse the consumers in their product perceptions? We show that the answer to this question depends decisively on preference heterogeneity. Contrary to the well-studied case of homogeneous goods, confusion is not necessarily an equilibrium in markets with differentiated goods. In particular, if the taste distribution is polarized, so that indifferent consumers are relatively rare, firms strive to fully educate consumers. By contrast, if the taste distribution features a concentration of indecisive consumers, confusion becomes part of the equilibrium strategies. The adverse welfare consequences of confusion can be more severe than with homogeneous goods, as consumers may not only pay higher prices, but also choose a dominated option, or inefficiently refrain from buying. Qualitatively similar insights obtain for political contests, in which candidates compete for voters with heterogeneous preferences.

Socially responsible behavior is crucial for slowing the spread of infectious diseases. However, economic and epidemiological models of disease transmission abstract from prosocial motivations as a driver of behaviors that impact the health of others. In an incentivized study, we show that a large majority of people are very reluctant to put others at risk for their personal benefit. Moreover, this experimental measure of prosociality predicts health behaviors during the COVID-19 pandemic, measured in a separate and ostensibly unrelated study with the same people. Prosocial individuals are more likely to follow physical distancing guidelines, stay home when sick, and buy face masks. We also find that prosociality measured two years before the pandemic predicts health behaviors during the pandemic. Our findings indicate that prosociality is a stable, long-term predictor of policy-relevant behaviors, suggesting that the impact of policies on a population may depend on the degree of prosociality.