Search

What can employers learn from personality tests when job applicants have incentives to misrepresent themselves? Using a within-subject, laboratory experiment, we compare personality measures with and without incentives for misrepresentation. Incentivized personality measures are weakly to moderately correlated with non-incentivized measures in most treatments but are correlated with intelligence when test-takers have information about desired personalities or are warned that responses may be verified. We document that actual job ads provide information about desired personalities and that employers in the UK who administer personality tests are also likely to administer intelligence tests despite the potential for substitution between the tests.

Evidence that women are less likely to opt into competitive compensation schemes in the laboratory has generated speculation that a gender difference in competitiveness contributes to the gender wage gap. Using data from the NLSY79 and NLSY97, we show that women are less likely to be employed in jobs using competitive compensation. The portion of the gender wage gap explained by gender segregation in compensation schemes is small in the NLSY79 but somewhat larger in the NLSY97 – suggesting an increasing role for competitiveness in explaining the gender wage gap.

Modeling the incentive effects of competitions among employees for promotions or financial rewards, economists have largely ignored the effects of competition on effort provision once the competition is finished. In a laboratory experiment, we examine how competition outcomes affect the provision of post-competition effort. We find that subjects who lose arbitrarily decided competitions choose lower subsequent effort levels than subjects who lose competitions decided by their effort choices. We explore the preferences underlying this behavior and show that subjects' reactions are related to their preferences for meritocratic outcomes.

We test the hypothesis that locus of control - one's perception of control over events in life - influences search by affecting beliefs about the efficacy of search effort in a laboratory experiment. We find that reservation offers and effort are increasing in the belief that one's efforts influence outcomes when subjects exert effort without knowing how effort influences the generation of offers but are unrelated to locus of control beliefs when subjects are informed about the relationship between effort and offers. These effects cannot be explained by locus of control's correlation with unmeasured human capital, personality traits, and the costs of search - alternative explanations for the relationships between locus of control and search behavior that cannot be ruled out using survey data - as the search task does not vary across treatments, which leads us to conclude that locus of control influences search through beliefs about the efficacy of search effort. Our findings provide evidence that locus of control measures can be used to identify job seekers at risk of becoming discouraged and abandoning search.