Department of Economics

This paper studies whether political ideology affected early restaurant avoidance behavior during Covid-19 in March 2020. Early conflicting narratives about the severity of Covid-19 driven by political leaders and media outlets with liberal versus conservative views may impact the speed in which Republican and Democratic cities adopt safety measures such as avoiding dining at restaurants. I use data on seated dining rates at restaurants from OpenTable and find that cities in the United States with more Democratic voters saw faster declines in seated dining. I also find that cities with more Covid-19 cases saw faster declines in seated dining rates, but this effect is concentrated on early adopter cities. These findings have public health implications which suggest that effective public health policy should take the influence of politics on behavior into account.

I propose a methodology to isolate the effect of competition on media content using local newspaper closures as an exogenous change in competition to closured newspapers’ competitors. I define five topical metrics and construct a specialized categorization scheme to measure newspaper content over time. By comparing content across the periods before and after a newspapers’ competitor closes, I hold all factors that may affect content constant save the level of competition in the market and chronological time. Following the theoretical model proposed by Perego and Yuksel (2020), I hypothesize that decreasing competition should incentivize general content and disincentivize specialized content. I test my hypothesis on a case-study of 13 local newspapers in California over the period from 2000 to 2020. I am unable to differentiate the effect of decreasing competition from other time-related factors, and therefore cannot make a definitive conclusion from the limited data. However, my methodology and categorization scheme may be applied a more extensive dataset further test whether the effect of competition on local newspaper content can be observed and isolated from other factors.

The decision by Germany to lead the effort against the refugee crisis in 2015 sparked controversy amongmany of its native population. With over a million new refugees in 2015 and 2016 alone fleeing war andterorrism from mainly Syria, Iraq, and Afghanistan, member’s of Germany’s far-right were outraged bythe decision. The AfD used the controversy to create a platform and gain popularity. They eventuallygained the third most votes in the 2017 federal election, the first time in 60 years that an openlynationalistic party would enter parliament. In this paper, we analyze how an increasing number ofrefugees affects voting sentiment in Germany. We search for causal effects of refugee numbers on votingfor far-right political parties. We perform our analysis with multiple regression models, mainly relying onfixed effects to determine causality. Although our initial hypothesis was that there should be a positiverelationship between the refugee share and voting for far-right political parties, our results indicate theopposite. We found that an increase in the amount of refugees actually decreases the average vote sharefor anti-refugee political parties. This falls in line with contact theory, the idea being that exposurebetween natives and foreigners can help create tolerance and acceptance among both groups. Our analysiscould be improved more, as many of our results show to be insignificant. If we took a more granularapproach by analyzing county-level data, we may be able to better isolate the effect that an increase in therefugee share has on voting.

In order to understand why labor force participation (andemployment rate) have declined in the United States over the pasttwo decades, declines in the employment to population ratio wereregressed against minimum wage, labor laws and labor forcecharacteristics during the last three major recession periods. Theresults yielded little significant effects of minimum wage orunionization, but a strong negative effect of the percent of the laborforce employed in the service sector.

This paper examines the effect of market size on the votes for the National Basketball Association’s All-NBA teams. While the All-NBA teams are intended to represent the league’s best players, prior research suggests that media members vote based on factors other than performance. I hypothesize that a player on a large market team is more likely to receive votes than a player with comparable performance but on a small market team. Using All-NBA ballots from five consecutive seasons and Nielsen television market sizes, I employ a two-part model approach to determine the effects of market size, being born outside the United States, representing an East Coast team, and Team Win-Loss Percentage on All-NBA votes. I find that the effect of market size on All-NBA votes is unclear, but Team Win-Loss Percentage has a significant positive effect. Future studies could introduce other factors to the models such as age and race.

Numerous solutions have been proposed to address the replication crisis, in which numeroushigh-profile empirical research studies cannot be replicated by other research teams. One possibleexplanation is that researchers have the option to adjust their data analyses after viewingthe results, inflating false positive rates. One popular solution is study preregistration, the practiceof developing the data analysis plan before the data is collected. However, preregistrationsonly alleviate replication problems if researchers are held accountable to their analysis plans.Across two related studies, we explore the effectiveness of preregistration in its current form.In Study 1, we audit recent preregistered publications from a major psychology journal andobserve deviations in 19 of 32 papers. In Study 2, we simulate the effects of generic deviationson the false-positive rate. We find that deviations that run more or more varied tests causelarger changes, tripling the false-positive rate in the most extreme case. We note that auditingpreregistrations requires an inconsistent amount of time depending on their length and format,which we suspect contributes to the enforcement issues we observe. We suggest that researchersand journals alike adopt the asPredicted.org template for preregistrations.

The proportion of international Ph.D. candidates has been increasing in recentyears, and the growing importance of international candidates requires researchersto pay more attention to the differences between them and Americancandidates. Using information collected from CVs on the 2021-2022 economicsPh.D. job market, we find that international candidates are more likely to havea previous graduate degree before a Ph.D., and less likely to have full time RAexperiences and a background in math. The results of this study suggest thatseveral variables have positive effects on Ph.D program rankings: graduatingfrom a highly ranked undergraduate institution, having full time RA experiences,and coming from Western Europe, Eastern Europe, or Latin America,while candidates attending an unranked undergraduate institution and an liberalarts college are less likely to graduate from highly ranked Ph.D. programs.

Although the negative health consequences of tobacco use are well-informed, smokers usuallydo not have an incentive to quit smoking immediately as the smoking interventions are notcompulsory and the illness caused by smoking is distant in time. Smoking behavior is denselystudied and proved to be associated with a wide range of genetic, social, and psychologicalfactors. This study is to learn how does COVID-19 spread influence the smoking prevalence inthe United State. The results show that smoking behavior is not geographically affected by thestrictness of lockdown orders and the severity of coronavirus spread. However, the cigaretteconsumption is associated with COVID-19 with a negative significance if people encounterdepression during COVID-19. The outcome provides some important information for thecessation-related organizations: it is necessary to take care of smokers’ emotion status in theprocess of quitting during COVID-19.

This study provides an estimation and methodology update on previous paper that studies the effect of having a second major in undergraduate on future earnings. Using 2019 National Survey of College Graduates (NSCG) data and Propensity Score Matching (PSM) method, I find that double majoring increases earnings by around 3% for the general population, and this earnings premium is more remarkable for people under the age of 40, which amounts to about 4%. While the proportion of double majors in the population drops from over 20% in 2003 to slightly above 13% in 2019, the returns to double majoring increase from around 2.5% since 2003. I also compare results from OLS regressions and PSM and argue that PSM can relax some of the parametric assumptions imposed by OLS regressions and hence reduce misspecification and extrapolation bias from OLS regressions, which previous literature on this topic relies on.

This paper seeks to show that the potential to lose money as a result of theft has adifferent effect on an individual’s risk aversion than the potential to lose money due to chance.This would indicate that an individual’s risk aversion is inconsistent under different scenarios,contrasting current literature that assumes an individual’s risk aversion is independent of thesituation they are in. We attempt to show this through an experiment that frames loss in the formof theft. We use Amazon Mechanical Turk to gather responses to our experiment online. We findthat our treatment has no statistical effect, that people do not act in a way that is inconsistent withtheir risk aversion simply because of the possibility of theft.