Spatial spillovers and the productivity-compensation gap in the United States

Rising inequality in the US, and resulting economic and wage stagnation, has been attributed to the growing productivity-compensation gap. Academics and policy makers have puzzled over the source of the gap since the 1970s. We hypothesize that the observed gap is misleading because productivity spillovers only partially translate into increased compensation. We analyze productivity spillovers with a spatial dimension, such that the industrial productivity of one state spills over to that of contiguous states. To do so, we employ a unique data set and spatial econometric techniques, including a spatial two-stage least squares instrumental variable approach and the spatially lagged X framework. The data set includes state-level industry panel data on employment, compensation, and gross industrial output for sixty-six industries within the forty-eight continental US states from 1998 to 2017. Once we account for productivity spillovers within industries, the gap between labor compensation and productivity narrows significantly.