Connecting the dots on the divergence between pay and productivity

From time to time researchers have raised technical measurement issues with our research showing that the compensation of a typical worker has diverged from overall productivity. The Heritage Foundation’s James Sherk—a repeat player—has a new entry.

Last September, Josh Bivens and I published a paper that provides a comprehensive review of the measurement issues and presents our estimates showing that rising inequality, both from greater inequality of compensation and an eroding labor’s share of income, drives the productivity-pay divergence, especially in this century. It easy to get caught up in a myriad of technical issues, none of which, as Bivens ably shows in his recent analysis, actually change the overall finding that hourly productivity growth has generally far exceeded that of a typical worker’s hourly compensation since around 1973 or 1979. As Bivens points out, Sherk actually ignores that we highlight the gap between a typical (median or “bottom eighty”) worker’s compensation and productivity and not the average compensation (including that of CEOs and the top one percent), which sets aside the main driver of the gap.

The bottom line, as Bivens points out, is that an era of growing inequality of wages and compensation and a stable or falling labor’s share of income will lead to a divergence between a typical worker’s pay and overall productivity growth. No quibbling about measurement issues can obliterate this finding. One is thus left with how to interpret the gap between productivity and a typical worker’s pay. One response by our critics is to suggest that average productivity is not any benchmark for a typical worker’s pay since it does not reflect their individual productivity. In this view, the typical worker is only nine percent more productive today than in 1979, despite being far better educated and older and working with more capital and modern technologies. I would enjoy seeing our critics make that case directly to the public, policymakers and workers themselves: “your stagnant pay reflects your failure to improve your productivity over the last nearly four decades.” The other response—ours—is to note that from 1947 to 1973 a typical worker’s compensation grew in tandem with productivity. Since then, changes in policy have broken that link, and we can no longer rely on growing productivity alone to produce broadly-shared wage and income growth. Rather, we must develop policies which reconnect pay and productivity (see Bernstein’s Reconnection Agenda or EPI’s Raising America’s Pay agenda). That means, as Bivens wrote in a new paper, that we must pay attention to distribution, not just growth, and prevent the further redistribution of income upwards let alone reverse the decades-long surge of inequality. Any politician offering more growth alone as the solution to stagnant wages and middle class incomes without policies to reconnect median pay and productivity is offering an inadequate solution and ignoring the hard lessons of the last four decades.