RESEARCH

Publication

"Social comparisons in job search" with Martin Sefton and Richard Upward, Journal of Economic Behavior & Organization, Volume 168, December 2019, 338-361.

Final Manuscript z-Tree Codes Data and Stata Codes Matlab Codes

Working Papers

Preferences for Hedging in Multiple-Choice Tests with Xing Zhang and Songfa Zhong

This paper proposes two novel scoring rules which allow the test-takers to hedge when responding to multiple-choice questions. We randomly assign 2,986 participants in an IQ test into three conditions. In the control condition, participants choose one option, and receive one point for correct response. In the treatment conditions, they can explicitly hedge by choosing k options: if the correct option is among the k chosen options, they receive 1/k point in the outcome-mixing treatment, and one point with probability 1/k in the probability-mixing treatment. We find that participants in both treatments hedge pervasively and score lower compared to those in the control. The observed differences depend on risk preferences of the participants. While scores in three conditions exhibit similar psychometric quality, we observe a significant correlation between academic performance and IQ score measured in the probability-mixing condition, but not in the other two conditions. These observations contribute to the literature on preference for hedging, the design of multiple-choice tests, and the relationship between risk preferences and cognitive ability.

Visceral influences and gender difference in competitiveness with Songfa Zhong

In a stylized experimental design, it is commonly observed that, after performing a task first under a piece rate and then a tournament incentive scheme, men are more likely to choose tournament for the next performance than women. We conduct two experiments to investigate whether this difference is influenced by the visceral state as the tournament-entry choice is made right after the first two tasks. Experiment 1 examines the responses of hormones, including progesterone, estradiol, testosterone and cortisol: men have stronger hormonal responses than women, and part of the gender difference in competitiveness is explained by the hormonal responses. Experiment 2 exogenously introduces a resting period before the choice, and shows that it reduces the competitiveness of men but does not affect women. Our results suggest that visceral influences contribute to the widely observed gender difference in competitiveness.

Lying and social norms: a lab-in-the-field experiment with children with Desponia Alempaki and Genyue Fu

Abstract:

We conduct a lab-in-the-field experiment with 567 children, aged four to eleven, in which we investigate the effect of social norms on lying and test whether norm sensitivity changes with age. Children think about a number between 1 and 6 in private, then roll a die, and report whether the number that came up is the same as the one they thought of. Just before making their report, we expose children to different empirical and normative information prescribing lying or honesty. We show that a normative intervention suggesting other children approve of honesty effectively reduces lying. We find limited evidence of the influence of our empirical interventions: information suggesting other children report honestly is effective only for younger children, while information suggesting other children report dishonestly does not influence lying patterns. We further observe that, although lying is omnipresent across all age groups, honesty significantly increases with age.

Assessing Inequality in the School Closure Response to COVID-19 with Xiqian Cai, Yi Lu, Xiangming Tang and Songfa Zhong

The spread of coronavirus disease 2019 (COVID-19) has caused school closures in more than 190 countries, affecting over 90% of the world’s student population. School closures can widen learning inequalities and disproportionately hurt vulnerable students. We collected data on the mock exam scores of university applicants in China before and after a two-month period of school closure. We observe that students from rural households and those receiving grant aid are more negatively affected by school closures compared to their urban and non-aided counterparts. To strengthen the causal interpretation of the results, we also investigate the scores in the previous graduating cohort who did not experience school closure, and find no evidence of the change in score over the same calendar period. Our study points to the urgent need to address the immediately increased and potentially long-termed educational inequality caused by school closures.

“Meeting and matching: New evidence on marketplace search in the labour market” with Martyn Andrews and Richard Upward (draft available upon request)

Abstract:

We provide the first microeconometric estimates of the hazards to matching on both sides of the same labour market, decomposed into their constituent parts, namely “meetings” (the rate at which job-seekers and vacancies contact each other) and “matching” (the probability that these contacts result in a match). Decomposing the overall hazard rate enables us to answer three questions which are central to the search and matching literature. First, whether it becomes harder for agents to match as time passes because they receive fewer contacts or because contacts are less likely to be successful. Second, whether the effects of labour market tightness operate through meetings or matches. Third, whether job-seeker and vacancy attributes which increase or decrease search effectiveness do so through meetings or matches.

In the raw data the decline in the matching rate is driven by a decline in the contact rate, and not by any fall in the probability of a match conditional on a contact. We estimate a two-sided matching model to determine whether this result is caused by omitted observed or unobserved heterogeneity in job-seekers and vacancies. It also allows us to estimate the parameters of the individual components of the matching function. We find that the same result applies as in the raw data: the decline in the matching rate on both sides of the market is driven by the decline in the contact rate.

“An empirical test of stock-flow matching model using job search contacts” (draft available upon request)

Abstract:

We provide new evidence on the mechanism by which job-seekers and vacancies meet, in an environment where both sides of the market have access to a database (or marketplace) of potential partners. In such an environment, a natural model of search is provided by stock-flow matching, and we show that the predictions of this model outperform that of a simple random matching model for the duration patterns of job search contacts. Our descriptive and econometric evidence shows that it is the inflow rate of new agents, rather than the total stock of agents, which determines the contact rates of existing agents, consistent with the predictions of the stock-flow model.