노사관계 제도와 그 변화가 소득불평등에 미친 영향

초록

Union density in France and the United States is nearly identical, both in the range of 10 percent. Yet collective bargaining coverage stands at 98 percent and 13 percent respectively, and their levels of income inequality (Gini coefficient) at 32.0 and 41.0. Research that has treated density as the indicator of union influence has difficulty explaining this contrast. What, then, is the factor that genuinely shapes inequality? This study analyzes the OECD-AIAS ICTWSS database (57 countries, 1960–2021, 2,796 observations). Seven institutional variables are used simultaneously: union density, bargaining coverage, wage coordination, centralization, union concentration, administrative extension, and government intervention. The dependent variables are the Gini coefficient and the P90/P10 and P50/P10 income ratios, estimated by two-way fixed-effects panel regression. The design substantially widens the temporal and spatial scope of earlier work, which typically remained within 15 to 20 countries and some three decades. The results converged on a single finding. Bargaining coverage was the only variable with a consistently negative effect across all three measures. Density, centralization, and union concentration were not significant. Administrative extension compressed P50/P10, confirming its function of setting a floor for low-wage workers. Institutions do not operate alone. The interaction between coverage and wage coordination was significantly negative for both income ratios, and the equalizing effect of coverage emerged only where coordination was high. The coverage–extension interaction, by contrast, was positive. This is interpreted not as a dysfunction of extension, but as substitutability and a ceiling effect between two institutions performing the same function. The magnitude of the effects was also considerable. A one-standard-deviation rise in coverage was observed to lower the Gini coefficient by about 1.9 points, equivalent to one quarter of the within-sample standard deviation of the Gini. Britain is the concrete case in point: the collapse of its coverage after 1980 explains 40 percent of the rise in its Gini over the same period. The results held under Hausman tests, lag specifications, and alternative sample compositions. In sum, what governs income distribution is not how many workers are organized. It is how far the outcomes of bargaining reach, and how far wage setting is coordinated. This constitutes exploratory evidence supporting the theory of institutional complementarity and the discussion of organized decentralization. Policy focus, too, should rest less on raising union density than on combining representativeness-based extension with coordination mechanisms.

키워드