earticle

영어

This study examines the relationship between employee pay and the capital structure of publicly listed firms in Korea. Consistent with Hanka (1998) and Hovakimian and Li (2011), we find that higher leverage is associated with lower average employee pay. However, our findings contrast with the theory of Berk, Stanton, and Zechner (2010) and the theory of the compensating wage differential. When firms add more debt to their balance sheet, the probability of bankruptcy increases and their employees face a greater risk of unemployment. Studies have suggested that workers bear significant economic and psychological costs from involuntary unemployment. In this context, the economic theory of Berk et al. (2010) predicts that firms compensate for this risk by increasing employee pay to increase the leverage related to higher average wages. However, in contrast to Berk et al. (2010), several studies have documented a negative correlation between leverage and employee wages. Hanka (1998) and Hovakimian and Li (2011) find that increased leverage is associated with lower employee pay. They interpret this as being a result of the risk shifting from shareholders and managers to employees. In addition to the risk-shifting motive, Perotti and Spier (1993) argue that firms use leverage as a bargaining tool against labor, especially when in financial distress. Moreover, Benmelech, Bergman, and Enriquez (2012) find that firms tend to renegotiate employee wages downward when they are under financial distress. Thus, from this perspective, higher leverage is associated with lower average employee pay among firms that are in financial difficulty. In summary, studies that predict a negative relationship between leverage and wages have suggested that increased leverage imposes disciplinary effects on labor by reducing employee wages. In our empirical analysis, we estimate panel fixed effect models that control for the time-invariant heterogeneity among firms and general time trends. In firm fixed effect regressions controlling solely for leverage, we find that leverage is associated with lower average employee pay. Moreover, in regressions that control for other firm characteristics, we find that higher leverage is associated with lower average employee pay. The Fama and MacBeth (1973) regressions deliver the same results. These findings suggest that the disciplinary effect of leverage is more pronounced than the compensating effect of higher bankruptcy risk. As our sample period covers the periods before and after the 2008 global financial crisis, we test whether the global financial crisis had any significant effects on the relationship between leverage and wages. We conduct the same regression analyses for the pre- and post-financial crisis periods (2003~2007, 2010~2015), and our empirical results suggest that the negative relationship between leverage and wages significantly intensified after the financial crisis. Moreover, the negative relationship between leverage and wages was only significant for the post-crisis period. This result confirms that the disciplinary effect of leverage was more pronounced after the financial crisis. We also consider industry heterogeneity and the labor market conditions in our analysis of the variation in the leverage-wage relationship. We expect employees’ wage levels to vary significantly among different industries, especially technology firms and non-technology firms (Chemmanur, Cheng, and Zhang, 2013). We examine the variations across industries and markets by conducting comparative analyses of KOSPI and KOSDAQ firms, and firms in the technology industry and those in other industries classified according to the KSIC code. We find that the leverage-wage relationship is significantly more negative among non-technology firms. We also predict that the disciplinary effect of leverage is stronger in industries with lower unionization and shortage rates, as employees in these industries have less bargaining power against the firms. Consistent with our prediction, we find that the negative relationship between leverage and wages is strengthened in industries with lower unionization and shortage rates. In summary, among Korean public firms, we find that higher leverage is associated with lower employee pay. This result substantiates the view that debt disciplines labor. However, our findings do not support the theory of Berk et al. (2010) that firms compensate for increased unemployment risk by increasing wages to gain greater leverage. In addition, we find that after the financial crisis (2008~2009), the negative relationship between leverage and wages increased significantly. Our findings suggest that the disciplinary effect of debt was significantly pronounced after the financial crisis. We also examine the effects of the labor market conditions and industry heterogeneity on the leverage-wage relationship. Consistent with our prediction, the negative relationship between leverage and wages is significantly strengthened among firms in non-technology industries and those belonging to industries with lower unionization and shortage rates. One limitation of our study is that we cannot fully control for the endogeneity of leverage. We expect that future research will mitigate this problem. Our evidence contributes to the growing body of literature connecting labor economics and corporate finance and to the research on the relationship between leverage and wages by providing additional evidence to support the disciplinary effect of debt.

한국어

본 연구에서는 국내 상장 기업의 자본구조와 임금의 관계를 분석한다. 2003년부터 2015년까지의 표본을 대상으로 한 실증분석 결과 부채비율과 종업원 1인당 평균 임금은 유의한 음(-)의 관계를 갖는 것으로 나타났다. 추가적으로 본 연구에서는 2008년 금융위기의 영향과 기술기업과 비기술기업의 이질성, 산업별 노동시장 상황이 부채- 임금관계에 미치는 영향을 분석하였다. 분석 결과 금융위기 이후 국내 기업들의 부채- 임금의 음(-)의 효과는 강화된 것으로 확인되었고 또한 비기술 기업일수록, 노조 조직률이 낮은 산업일수록, 인력 부족률이 낮은 산업일수록 부채-임금의 음(-)의 효과는 강화되었다. 전반적으로 이러한 증거들은 부채 증가가 노동을 규율함을 예측한 Hanka(1998), Hovakimian and Li(2011)의 기존연구와 일치하며 부채-임금관계는 기업과 종업원 사이의 상대적 임금협상력에 크게 영향 받을 수 있음을 시사한다. 한편, 이는 균형에서 부채의 증가로 인한 실직 위험의 증가를 더 높은 임금으로 보상한다는 Berk, Stanton, and Zechner(2010), Chemmanur, Cheng, and Zhang(2013)의 기존 연구와 노동경제학의 보상 임금 이론과는 일관되지 않음을 확인하는 결과이다