坪井 美都紀

We examine how workers' collective voice affects union and nonunion wages both theoretically and empirically. We develop a simple union model with a social insurance system and show that, due to two counteracting forces in bargaining, higher bargaining power raises both union and nonunion wages when ex-ante bargaining power is low but lowers them when it is sufficiently high. Using Japanese household-level data, we document an inverted U-shaped relationship between bargaining power and wages for both groups. Applying entropy balancing, we confirm that union wages are maximized at about 70 percent bargaining power, compared with 44 percent for nonunion wages. Given Japan's current level of about 25 percent, stronger bargaining power and a rising societal willingness to improve working conditions are likely to benefit union members and also nonunion members once under union coverage.

Union workers have earned a positive union wage premium over nonunion workers. Several recent micro-level studies using regression discontinuity designs, however, find that unionization negatively affects wages. To examine whether this negative union wage effect is merely an artifact of specific research designs or reflects a plausible underlying economic mechanism, we develop a simple model of trade unions that features firm heterogeneity and a social insurance system. Our model yields analytically tractable comparative statics and offers a clear testable prediction that greater bargaining power reduces the wage of union workers. Our result thus highlights the importance of accumulating further micro-level evidence that explicitly accounts for these factors to build a consensus of the union wage effect.

To examine whether offshoring widens or narrows wage inequality, we develop a tractable North-South model featuring firm heterogeneity, foreign outsourcing, and vertical specialization. In the baseline model with exogenous firms' outsourcing decisions, we show that increased outsourcing widens wage inequality. In an extended model with endogenous firms' outsourcing decisions, however, outsourcing either widens or narrows wage inequality, depending on the initial level of outsourcing. Using Japanese industry-level data, we then show that–in contrast to most findings for the U.S.–offshoring narrows wage inequality once we account for initial industry-level differences in the extent of outsourcing.