We investigate to what extent worker heterogeneity explains the well-known wage and productivity exporter premiums, employing a matched employer-employee data set for Norwegian manufacturing. The wage premium falls by roughly 50% after controlling for observed and unobserved worker characteristics, while the total factor productivity premium falls by 25% to 40%, suggesting that sorting explains up to half of these premiums. Recent trade models emphasize the role of within-industry reallocation of labor in response to various shocks to the economy. Our findings suggest that aggregate productivity gains due to reallocation may be overstated if not controlling for sorting between firms and workers.

Multinational production (MP) can lead to large gains through international technology sharing. However, empirical evidence suggests that geography matters for MP: Affiliate sales fall in distance from the headquarters. We introduce intrafirm trade into a standard model of exports and MP and show that the model is consistent with firm-level and aggregate evidence. Using a maximum likelihood estimator, we find that intrafirm trade plays a crucial role in shaping the geography of MP. An implication of our work is that MP and exports are very similar activities. Consequently, shutting down MP leads to relatively small welfare losses.