Affirmative Action and Corporate Compliance in South Korea

Abstract

The Affirmative Action Act was introduced in South Korea in 2006 to increase female employment and correct discriminatory hiring practices. Using the combined data sets of survey and the Act's implementation plan, this paper provides logit estimation results to examine empirically how political perceptions or attitudes of firms influence corporate noncompliance with the Act. According to a corporate personnel manager survey, affirmative action was initially pursued as a campaign pledge by the liberal party (the Korea Democratic Party) to attract women's votes, and took on a looser shape as the government compromised with the business sector after an election. A weak enforcement structure ultimately diluted the effects of the Act. A logit analysis indicates that noncompliance is more probable in companies that perceive affirmative action as part of a design to achieve political goals, and compliance is more probable in companies that feel it is likely to improve corporate management.

Keywords:

• Female employment
• affirmative action
• corporate compliance
• political perception

Acknowledgments

We deeply appreciate the very detailed and helpful comments made by three anonymous reviewers and by the participants in the 2008 International Association for Feminist Economics (IAFFE) Conference in Torino, Italy.

Notes

¹ South Korea affirmative action benchmarks for female representation as a percentage of all employment by enterprise and the concomitant benchmarks for women in leading positions are high by international standards. However, legislation does not include any penalty for not granting equal opportunities or incentives to promote equal opportunities in the workplace.

² The term of contracts is associated with occupational segregation in Korea. Female workers with contingent contracts intensively work in the peripheral jobs, while male workers with permanent contracts intensively work in the core jobs. Occupational segregation alone is not a sufficient indicator of gender discrimination. As one reviewer of this study explained, countries like Sweden have pronounced occupational and industrial gender segmentation of employment, but women can still access top managerial posts in Sweden and most European countries. However, if the term of contracts is associated with occupational segregation as in Korea, the occupational segregation is more likely to reflect gender discrimination. This aspect of the problem requires more thorough empirical verification. We intend to investigate this further in future research.

³ US companies are usually targeted for inspection if the company's level of women's and minority employment is less than 80 percent of the average for their industry, or if women's and minority employment are concentrated too heavily in low-level jobs.

⁴ One may consult Lei Delsen and Erik Poutsma (2005) to examine how the implementation of a new labor institution can raise the transaction cost in the labor market. An abrupt adoption of affirmative action may increase the transaction cost in the Korean labor market, especially for small companies. The South Korean AA Act provided a 2-year grace period (2008) for firms with fewer than 500 employees in order to give these companies sufficient time to prepare for implementation of its policies.

⁵ In the headquarters of the Korea Ministry of Labor, in Gwachun, Gyeonggi-do, only two people work in the affirmative action section.

⁶ Ninety percent of the employers state that they started to recruit college-educated women only in the mid 1990s; as a result, there were not yet many women eligible for promotion to managerial posts.

⁷ A potential problem of selection bias may arise associated with the employer attitude toward government policy between the respondents and nonrespondents. However, the firm size distribution for nonrespondents was similar to that for the respondents. Employers with 1,000 or more full-time employees accounted for 56.7 percent of respondents and 55.2 percent of nonrespondents. Employers with 500–999 employees occupied 7.7 percent of respondents and 8 percent of nonrespondents. Employers with 499 or fewer employees represent 35.6 percent of respondents and 36.8 percent of nonrespondents. Furthermore, the compliance rate for the nonrespondents is almost the same as the compliance rate for the respondents: 41 percent for the respondents and 39.5 percent for the nonrespondents. Finally, as both the respondents and nonrespondents are members of the Korea Employers Federation (KEF), which regularly provides them with training and communication programs, we considered that the selection bias resulting from the opinion difference between the respondents and nonrespondents would not meaningfully affect the study results.