AI x Econ

Author: Shanshan Zhang, Matthew Gomies, Narek Bejanyan, Zhou Fang, Jason Justo, Li-Hsin Lin, Rainita Narender, and Joshua Tasoff.
Journal: Journal of Economic Behavior & Organization, Volume 179, November 2020, Pages 460-474.
Coverage: Los Angeles Times | The Scottish Mail on Sunday
Abstract:
We hypothesize that clothes can affect the behavior of the wearer by influencing the person’s identity. We test this hypothesis by recruiting trick-or-treaters during Halloween, a time of year when people wear salient and extreme clothing. Because the tradition of costume-wear for Halloween evolved, in part, to hide one’s identity during “tricks” (i.e. norm violations), we measure the effect of Halloween costumes on ethical behavior. We use the lying game of Fischbacher and Föllmi–Heusi as our experimental paradigm with 2 × 3 × 2 conditions. First, we vary the stakes to price lying behavior. Second, we run three conditions with different beneficiaries of the report (self, other, and both) to test whether lying for others is perceived to be normative. Third, we manipulate the salience of one’s costume to test the effect of costume and identity on ethical behavior. Surprisingly, we find that costume salience caused “good guys” to lie more and “bad guys” to lie less. We interpret this either as a moral licensing effect or as stemming from a perception of being monitored. Our design allows for the identification of contagion effects, and although there were no direct effects of gender, we found that children lie more when children of the same gender near them lie more. We also find that stakes had no effect, people lied more for themselves than for others, and lying has an inverted-U pattern over age, peaking at age 12.

Author: Giacomo Di Pasquale, Matthew Gomies, and Javier Rodriguez.
Journal: Social Science Quarterly, Volume 102, Issue 3, September 2023, Pages 453-482.
Abstract:
Recent research on increasing income inequality focuses on recessions’ role on the income distribution and racial disparities in the United States. This study expands this research by focusing on the evolution of racial income inequality during postrecession periods. We hypothesize that differential recovery trends by race and income rank during postrecession periods exacerbate between- and within-race income inequality. Specifically, we examine postrecession trajectories of race-specific weekly earnings for the bottom and top 10 percent in the U.S. weekly earnings distribution. We apply a break-spline regression approach to quarterly weekly earnings data (2001–2018) collected from the Bureau of Labor Statistics and the Federal Reserve Bank of Saint Louis for the 2001 Recession, the Great Recession, and between-recession periods. Results show an increase in income inequality between the bottom and top 10 percent in the weekly earnings distribution during recovery periods. For all races, the bottom 10 percent recovered from recessions slower and later than the top 10 percent. We also find that Asian Americans’ weekly earnings influence between- and within-race income inequality during postrecession periods. Income inequality—overall and across races—is entrenched as racial/ethnic minorities and the poor are particularly vulnerable to recessions and are neither catching up nor recovering during upturns of the economy.

Author: Gregory DeAngelo, Matthew Gomies, and Rustam Romaniuc.
Journal: Public Choice, Volume 196, Issue 6, November 2021, Pages 2812-2823.
Abstract:
Law enforcement institutions are tasked with a complicated undertaking that involves maintaining community safety and, at times, making arrests while exercising care in their interaction with private citizens. Errors may have dramatic consequences for civilians, police and the criminal justice system. Given limited observability of law enforcement agents’ behavior, one way to mitigate the principal-agent problem is to rely on signals from civilians via complaints. At the same time, civilian complaints may result in reputational and financial losses for the criminal justice institutions. This paper empirically investigates one way in which criminal justice institutions respond to civilian complaints. Namely, criminal prosecutors can upcharge a defendant who files a civil complaint against law enforcement. By upcharging, the prosecutor can increase the likelihood that a defendant will accept a plea deal, thus preventing the defendant from seeking monetary damages in civil court (Heck vs. Humphrey, 1994). Using data on citizen complaints and criminal charge outcomes from Cook County (Illinois), we find a strong causal link between a citizen filing a complaint and the total number of charges filed.

The Sound of Cooperation and Deception in High Stakes Interactions
Author: Monica Capra, Matthew Gomies, and Shanshan Zhang.
Abstract:
Current methods for predicting cooperation and deception from communication mostly rely on analyzing transcribed messages or text. In this study, we depart from this conventional approach by focusing on the rich information embedded in the voice itself. We collected recordings of natural speech from contestants of a high-stakes popular TV show and from defendants’ and witnesses’ testimonies in actual court proceedings and used machine learning algorithms to classify cooperative and deceptive speech with an accuracy range of 58% to 79%, solely based on acoustic features. Our findings show that higher pitch and lower intonation are associated with deceptive speech, and provide insights into the potential of using voice data to predict cooperation and deception in high-stakes, natural contexts. This study has important practical implications and demonstrates the potential of using unintrusive and inexpensive voice data to gain valuable insights into social interaction.

(When) Would You Lie to a Voicebot?
Author: Monica Capra, Matthew Gomies, and Shanshan Zhang.
Abstract:
With the advent of technological advancements, interacting with voicebots has become increasingly common. However, dishonest behaviors that occur in human-human interactions may not carry over into human-machine interactions. To explore this, we conducted an online experiment using a coin-toss task and compared reported outcomes across different reporting channels: Human Voice, Voicebot, and Text. We designed a uniform online voice chat interface to standardize the reporting experience. We also tested the effect of a feminine and a masculine voice on misreporting and varied the level of sophistication of the voicebot (AI-enhanced voicebot). Our results show that, on average, there is no significant difference in the likelihood of misreporting through a voicebot and a human voice, or between verbal and written reporting. However, we found that participants who listened to a feminine voice were more likely to lie than those who listened to a masculine voice. Moreover, those who heard a feminine voice were more likely to lie to a voicebot than a human voice. Interestingly, such difference disappears with higher sophistication (i.e., AI-enhanced voicebot). In contrast, when hearing a masculine voice, there was no difference in misreporting between the voicebot and human voice treatments. These findings suggest that utilizing a masculine voice for voicebots or voicebots with higher sophistication and feminine voice could help deter or diminish dishonest reporting in human-machine interactions.

The Effect of Contemporaneous Meat Consumption on Attitudes and Behaviors Towards Animal Welfare
Author: Monica Capra, Xi Chen, Seong-Gyu Park, Joshua Tasoff, Jin Xu, and Shanshan Zhang.
Abstract:
Animal welfare in meat production is concerning for ethical reasons. Research in psychology has shown that contemporaneous consumption of meat causes people to have less moral concern for farmed animals. Following this research, we run a laboratory experiment to test whether near contemporaneous meat consumption can affect behavior directly through information choice about animal welfare, contributions to an animal charity, and a proxy measure for political behavior. We also test for the indirect effects of meat consumption on our charity outcome and political outcome by way of its effect on information preferences. Though we find that meat consumption changes attitudes towards animals, and information changes charitable contributions, we find that meat consumption does not affect our behavior outcomes. The null result casts doubt on the extent to which shifts in attitudes translate to shifts in behavior. An online hypothetical experiment finds that information preferences are consistent with expectedutility theory, and we again find no evidence of motivated thinking on behavior.

Thoughts and Wellbeing: How Thinking Process Affects Subjective Wellbeing
Author: Joshua Tasoff, Shanshan Zhang, and Zhou Fang.

Hopes and Fears: Narratives, Attitudes and Behaviors toward the COVID-19 Pandemic
Author: Monica Capra, Matthew Gomies, and Shanshan Zhang.