Is risk aversion related to occupational choice: evidence from 1996 PSID

ABSTRACT

This paper investigates to what extent individuals’ risk aversion are correlated with the choice of occupation which differs from earnings risk by using 1996 PSID. Following Kimball et al. (2009), we first impute the individuals’ coefficient of relative risk aversion based on personal characteristics. Then, on the aggregate level, we find workers who are more risk averse will sort into occupations with low earnings risk. On the individual level, for one-unit increase in the coefficient of relative risk aversion, the probability of going to a more risky occupation decreases by 13.5%.

KEYWORDS:

• Risk aversion
• imputation
• sorting
• occupational choice

JEL CODES::

• J24
• J31
• D01
• D81

Acknowledgments

We are grateful for comments by Howard Bodenhorn, Bill Dougan, Aspen Gorry, Robert Fleck, Tom Lam, Mike Hoy, Stephen Shore, Michael Makowsky, Peter Blair, Molly Espey, Cesar Castellon, Jordan Adamson, Jonathan Ernest, Curtis Simon, Scott Barkowski, Babur De los Santos, Kelsey Roberts, and Tim Bersak. We especially thank the anonymous referee for helpful comments and suggestions. Zhou acknowledges the financial supports from the National Natural Science Foundation of China (No. 71573050 and No. 71573170).

Disclosure statement

No potential conflict of interest was reported by the authors.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

¹ Moreover, since risk preferences are changing over time (Andersen et al. 2008; Harrison, Lau, and Yoo 2019), volatility of earnings across time is a more appropriate measure in the paper if time-varying individual risk aversion parameter could be obtained. The questions of hypothetical job choices are included only in the 1996 PSID survey, thus the risk aversion parameters, estimated from the answers to the questionnaires, represent individuals’ risk aversion in 1996. Considering temporal stability of risk preferences, we use 1996 cross-sectional variation of residual wages as a measure of earnings variance.

² Kimball, Sahm, and Shapiro (2008) also develop a measure of relative risk tolerance based on responses to hypothetical income gambles in the Health and Retirement Study.

³ A more appropriate variable to use is wealth. The closest variable in the PSID to indicate wealth is asset income, which includes the income from interest, dividends, trust funds, and rents. However, wefind asset income is zero for nearly half observations. Therefore, instead of using wealth, we follow Bonin et al. (2007) to include log individual earnings as an additional control.

Additional information

Funding

Zhou acknowledges the financial supports from the the National Natural Science Foundation of China [71573050,71573170].