Exchange rate risk and the skill composition of labor

Abstract

This paper uses matched employer-employee data covering the universe of Swedish private sector firms 1998-2014 to examine the links between exchange rate risk and the skill composition of labor in firms. We use firm × export destination and firm × import origin data to calculate firm-level measures of real exchange rate risk and real effective exchange rates. Our main result is that firms facing higher exchange rate risk employ a higher share of skilled labor and that the effect is especially marked among trade intensive firms in manufacturing and in firms with fewer than 50 employees. This finding is consistent with theoretical models of flexible work systems that require higher skill levels as a way of responding to higher risk.Keywords Skill-biased organizational change · Exchange rate risk · Firm heterogeneity JEL Classification F16 · F31 · J24 Electronic supplementary material The online version of this article (https ://doi.

Policy Uncertainty, Trade, and Welfare: Theory and Evidence for China and the United States

“…The stationary price index is then P m = P ( c m , c m h , τ m ) , which evaluates (14) at the cutoffs for each state. 25 Similarly, the domestic component of the price index is P m, h = P h ( c m h ) so its change, P ˆ m, h , depends on changes in the domestic entry cutoff, c ˆ m, h . The import price index is P m, x = P x ( c m , τ m ) and therefore P ˆ m, x depends on the change in the cutoff and also in tariffs if τ m ≠ τ b .…”

“…In online Appendix D.1 we show that higher tariffs increase the equilibrium price index, d P D /dτ > 0 , reduce export entry, d c D /dτ < 0 , and increase domestic entry, d c h D /dτ > 0 . Replacing the equilibrium 25 If T ≥ 0 periods ago the tariff changed to τ m then the stationary equilibrium is given by the value of y T→∞ ( τ m ) . In a stationary equilibrium there is still exogenous death but it is exactly offset by entry thus leaving the firm mass unchanged.…”

“…We allow τ 2 ∈ [1.04, 1.38 ] so it ranges from τ 1 to the maximum value of τ 2 in our sample. We obtain the TPU component for prices using (25) and do the same for the other outcomes.…”

“…The price effect term uses the multi-industry version of (24) at fixed tariffs and the decomposition given in (25). The effect is not sensitive to alternative values of α , it ranges from 29 to 30 log points if α = 6 or 0 respectively.…”

“…We compute it using the multi-industry version of (24) at fixed tariffs and the decomposition given in(25) 43. This does not mean that changes in tariffs have little effect on the price index.…”