Working Paper: NBER ID: w13139
Authors: Donald R. Davis; James Harrigan
Abstract: How do labor markets adjust to trade liberalization? Leading models of intraindustry trade (Krugman (1981), Melitz (2003)) assume homogeneous workers and full employment, and thus predict that all workers win from trade liberalization, a conclusion at odds with the public debate. Our paper develops a new model that merges Melitz (2003) with Shapiro and Stiglitz (1984), so also links product market churning to labor market churning. Workers care about their jobs because the model features aggregate unemployment and jobs that pay different wages to identical workers. Simulations show that, for reasonable parameter values, as many as one-fourth of existing "good jobs" (those with above average wage) may be destroyed in a liberalization. This is true even as the model shows minimal impact on aggregate unemployment and quite substantial aggregate gains from trade.
Keywords: Trade Liberalization; Labor Markets; Job Destruction; Wage Distribution; Efficiency Wages
JEL Codes: F1; F16; J2; J31
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| trade liberalization (F13) | destruction of good jobs (F66) |
| trade liberalization (F13) | minimal impact on aggregate unemployment (J64) |
| trade liberalization (F13) | significant labor market churning (J63) |
| product market churning (D49) | significant labor market churning (J63) |
| aggregate real income increases (E25) | distributional effects result in job losses among higher-wage positions (F66) |
| presence of job rents and unemployment (J69) | workers care about job loss (J63) |
| aggregate gains from trade (F14) | distributional conflicts between good and bad jobs (J68) |