Effects and Channels of Smoking on College Students’ Labor Market Performance

Abstract

This study investigates the impact of smoking on the labor market performance of college graduates, utilizing representative data from South Korea. To mitigate endogeneity concerns, we employ the average cigarette price during participants’ college years as an instrumental variable for smoking behaviors. The findings reveal a detrimental effect of smoking on the labor market outcomes of college students. Smokers are significantly less likely to secure employment, and if employed, they tend to earn lower salaries. Notably, these adverse effects are more pronounced among female students. Through a behavioral and human capital lens, we discern the mechanism by which smoking impairs labor market performance. While smoking may initially contribute to positive aspects such as enhanced social activities, subjective mental health, and psychological stability among college students, it is concurrently associated with lower academic achievement, impeding human capital accumulation. These results suggest that although smoking may yield short-term benefits, it undermines long-term human capital development. The implications advocate for governmental intervention to mitigate the negative consequences of youth smoking.

KEYWORDS:

• Smoking
• labor market performance
• behavioral economics
• human capital
• instrumental variable

JEL Codes:

• C36
• I12
• J30

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 According to the World Health Organization, smoking is one of the most severe threats to public health, as over 8 million people per year die because of tobacco. The Centers for Disease Control and Prevention report that more than 16 million Americans suffer from smoking-related diseases.

2 Refer to the following website [Accessed on 19 May 2021]: https://worldpopulationreview.com/country-rankings/smoking-rates-by-country

3 In 2014, South Korea recorded the lowest cigarette prices as well as the highest male smoking rate among OECD countries. Consequently, concerns abounded that low cigarette prices are responsible for high smoking rates and the deterioration of public health. The South Korean government responded to these concerns by nearly doubling cigarette prices in 2015.

4 We include the school fixed effect in the regression as some of the unobserved determinants of the choice between colleges might affect smoking behavior.

5 They include cannabis use (Beenstock & Rahav, 2002), later health (Leigh & Schembri, 2004), future smoking (Chaloupka, 1991), and birth outcomes (Evans & Ringel, 1999).

6 As of 2015, one USD was approximately 1152 KRW.

7 For example, people born in any month of 2001 became legally allowed to smoke starting on 1 January 2020.

8 South Korea is notable in its rapid expansion of tertiary education, as more than 70% of high school graduates have advanced to college since the 2000s (Park, 2013; Choi, 2015).

9 Refer to the following website: http://www.kdca.go.kr/yhs

10 The Korea Education and Employment Panel Survey is a longitudinal study conducted with middle school and high school students beginning in 2004 with the aim of examining teenagers’ lifestyles, educational experiences, and career paths.

11 We calculated the estimation in Figure 3 as follows. For an individual $i$ at time $t$, their smoking status is as follows: $\mathrm{Smokin}{g}_{\mathrm{it}}=\sum _{t=1}^4{\beta }_{t}I(\mathrm{Yea}{r}_t=t)+{\mu }_i+{\epsilon }_{\mathrm{it}}$. In the model, $\mathrm{Yea}{r}_t$ ranges from 1 to 4, representing the year in college (e.g., a value of 1 indicates freshman year). The baseline ($\mathrm{Yea}{r}_t$ =  0) indicates the last grade of high school (Grade 12). ${\mu }_i$ is an individual fixed effect, and ${\epsilon }_{\mathrm{it}}$ is an unobservable disturbance. $\mathrm{Smokin}{g}_{\mathrm{it}}$ takes the value 1 if the individual smokes and 0 otherwise.

12 Various dependent variables were employed in this study, and the average values of these dependent variables are presented in the tables of the regression analysis result.

13 $X_{\mathrm{ist}}$ includes gender, father’s and mother’s educational level, parents’ income at college enrollment, and total months spent in college.

14 For instance, tobacco consumption and potential unemployment may be jointly determined by a common attitude (such as risk preference) that is unobserved and therefore omitted from the model.

15 Appendix A provides the first-stage results.

16 Appendix B provides the results yielded by using the alternative IV.

17 For the robustness check, we also report the regression results when imputing no income for 10,000 KRW (the lowest value that respondents were allowed to report) in Appendix C.

18 In this study, we could not directly observe objective health-related variables due to the limitations of the dataset.

19 The variables used in this section are defined as follows: (1) Satisfaction with peers is measured on a 7-point scale. A value of 1 is assigned if the score is 5 or higher, and 0 is assigned otherwise. (2) Job information from their friends is represented by a dummy variable. A value of 1 is assigned if the respondent reports that friends provide job information, and 0 is assigned otherwise. (3) Mental health is assessed on a 7-point scale. A value of 1 is assigned if the score is 5 or higher, and 0 is assigned otherwise. (4) Exercise is observed based on hours of exercise per week. (5) Subjective health report is measured on a 5-point scale. A value of 1 is assigned if the score is 4 or higher, and 0 is assigned otherwise. (6) GPA is converted to percentile scores and then standardized. (7) Raw scores on the TOEIC range from 10 to 990, and the standardized TOEIC score is applied. The variables utilized in the mechanism inference are concurrent with the number of cigarettes smoked per day. In the case of the TOEIC, respondents are expected to record the best English score achieved after entering college.

Additional information

Funding

This work is supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea [grant number NRF-2022S1A5A2A0305221911].

Notes on contributors

Hoyong Jung

Hoyong Jung is an assistant professor in the Graduate School of Public Administration at Seoul National University.

Seungwoo Chin

Seungwoo Chin is a senior deputy director at the Ministry of Economy and Finance.