Productivity spillovers and the productivity-compensation gap
Explainer of Shakya et al. (2022), Spatial Economic Analysis. Not the journal article.
Citation: Shakya, S., & Plemmons, A. (2022). Productivity spillovers and the productivity-compensation gap. Spatial Economic Analysis. Free PDF
Facts¶
Main finding¶
The productivity-compensation gap narrows significantly in some industries once productivity spillovers between bordering and trading states are accounted for using spatially lagged models. However, effects are heterogeneous across the 23 industries examined - gap reduction is not uniform, with substantial variation in how spillovers mediate the relationship between productivity and compensation.
One caveat¶
Results vary significantly across industries and model specifications; not all industries show gap reduction when accounting for spillovers. Uses single-year cross-sectional data (2012), limiting temporal analysis. Geographic and trade network definitions are static, potentially missing dynamic spillover patterns over time or across business cycles.
Policy hook¶
Since the 1970s, US worker productivity has grown faster than wage growth - a divergence called the productivity-compensation gap. Conventional explanations focus on within-firm dynamics, but recent economics points to spillovers: productivity gains in one region may benefit workers elsewhere through trade and knowledge diffusion. Does accounting for these cross-regional spillovers explain part of the gap?
Data and setting¶
IO-Snap state-level industry data on employment, compensation, and gross industrial output (2012); Commodity Flow Survey Public Use Microdata File containing approximately 4.5 million shipments for interstate trade analysis; analysis covers 23 industries across all 50 US states.
Research design (plain language)¶
Spatial econometrics using Spatially Lagged X (SLX) models examining two dimensions of spillovers: spatial (contiguous states) and trade-network (trading states). Compares baseline models to spatially-adjusted specifications. Instrumental variables approach addresses endogeneity concerns. Log-log regression specifications allow comparison of productivity-compensation linkages across industries.
PDF or DOI¶
Why it matters¶
The productivity-compensation gap is a central inequality story: workers are more productive but wages stagnate. Explanations range from technology replacing labor to corporate rent-seeking. This paper suggests that geographic spillovers matter - productivity gains in neighboring or trading states may accrue to capital owners there rather than being shared with workers. For policy makers, the finding implies that local wage-setting cannot ignore regional productivity dynamics. Industries integrated through trade networks face different incentives than isolated sectors. For labor advocates, it suggests that wage policy may need regional coordination rather than local-only approaches. The group differences across industries also indicates that productivity-compensation dynamics vary by sector, implying that sector-specific policy design may be more effective than economy-wide solutions.