How does a region's average wage adjust to a shock, say trade-induced or technology-induced, on labor demand? Beaudry, Green and Sand (2010) recently demonstrated the inaccuracy of the traditionally-used shift-share analysis - a partial equilibrium exercise - in addressing this question. While they focus on shifts in the industrial composition of employment, I argue that the interplay between inter-sectoral and inter-occupational labor adjustments is fundamental in assessing the spillover effects they emphasize. I extend their search-and-bargaining model to incorporate occupations and illustrate why omitting inter-occupational labor adjustments could lead to underestimation. Using German individual-level data for 1977-2001, I estimate that omitting this dimension creates a substantial and statistically significant negative bias
representing two-thirds of the total effect.