Rising wage inequality: the wedge between productivity and worker’s wages

The New York Times’ Paul Krugman blog today cites the most recent report by Josh Bivens and Larry Mishel on the productivity-pay gap.  Understanding the Historic Divergence Between Productivity and a Typical Worker’s Pay: Why It Matters and Why It’s Real:

Wage stagnation experienced by the vast majority of American workers has emerged as a central issue in economic policy debates, with candidates and leaders of both parties noting its importance. This is a welcome development because it means that economic inequality has become a focus of attention and that policymakers are seeing the connection between wage stagnation and inequality. Put simply, wage stagnation is how the rise in inequality has damaged the vast majority of American workers.

The Economic Policy Institute’s earlier paper, Raising America’s Pay: Why It’s Our Central Economic Policy Challenge, presented a thorough analysis of income and wage trends, documented rising wage inequality, and provided strong evidence that wage stagnation is largely the result of policy choices that boosted the bargaining power of those with the most wealth and power (Bivens et al. 2014). As we argued, better policy choices, made with low- and moderate-wage earners in mind, can lead to more widespread wage growth and strengthen and expand the middle class.

This paper updates and explains the implications of the central component of the wage stagnation story: the growing gap between overall productivity growth and the pay of the vast majority of workers since the 1970s. A careful analysis of this gap between pay and productivity provides several important insights for the ongoing debate about how to address wage stagnation and rising inequality. First, wages did not stagnate for the vast majority because growth in productivity (or income and wealth creation) collapsed. Yes, the policy shifts that led to rising inequality were also associated with a slowdown in productivity growth, but even with this slowdown, productivity still managed to rise substantially in recent decades. But essentially none of this productivity growth flowed into the paychecks of typical American workers. Second, pay failed to track productivity primarily due to two key dynamics representing rising inequality: the rising inequality of compensation (more wage and salary income accumulating at the very top of the pay scale) and the shift in the share of overall national income going to owners of capital and away from the pay of employees. Third, although boosting productivity growth is an important long-run goal, this will not lead to broad-based wage gains unless we pursue policies that reconnect productivity growth and the pay of the vast majority.

* * *

[I]n this paper we not only provide an updated analysis of the productivity–pay disconnect and the factors behind it, we also explain why the measurement choices we have made are the correct ones. As we demonstrate, the data series and methods we use to construct our graph of the growing gap between productivity and typical worker pay best capture how income generated in an average hour of work in the U.S. economy has not trickled down to raise hourly pay for typical workers.

Key findings from the paper include:

• For decades following the end of World War II, inflation-adjusted hourly compensation (including employer-provided benefits as well as wages) for the vast majority of American workers rose in line with increases in economy-wide productivity. Thus hourly pay became the primary mechanism that transmitted economy-wide productivity growth into broad-based increases in living standards.
• Since 1973, hourly compensation of the vast majority of American workers has not risen in line with economy-wide productivity. In fact, hourly compensation has almost stopped rising at all. Net productivity grew 72.2 percent between 1973 and 2014. Yet inflation-adjusted hourly compensation of the median worker rose just 8.7 percent, or 0.20 percent annually, over this same period, with essentially all of the growth occurring between 1995 and 2002. Another measure of the pay of the typical worker, real hourly compensation of production, nonsupervisory workers, who make up 80 percent of the workforce, also shows pay stagnation for most of the period since 1973, rising 9.2 percent between 1973 and 2014. Again, the lion’s share of this growth occurred between 1995 and 2002.
• Net productivity grew 1.33 percent each year between 1973 and 2014, faster than the meager 0.20 percent annual rise in median hourly compensation. In essence, about 15 percent of productivity growth between 1973 and 2014 translated into higher hourly wages and benefits for the typical American worker. Since 2000, the gap between productivity and pay has risen even faster. The net productivity growth of 21.6 percent from 2000 to 2014 translated into just a 1.8 percent rise in inflation-adjusted compensation for the median worker (just 8 percent of net productivity growth).
• Since 2000, more than 80 percent of the divergence between a typical (median) worker’s pay growth and overall net productivity growth has been driven by rising inequality (specifically, greater inequality of compensation and a falling share of income going to workers relative to capital owners). Over the entire 1973–2014 period, rising inequality explains over two-thirds of the productivity–pay divergence.
• If the hourly pay of typical American workers had kept pace with productivity growth since the 1970s, then there would have been no rise in income inequality during that period. Instead, productivity growth that did not accrue to typical workers’ pay concentrated at the very top of the pay scale (in inflated CEO pay, for example) and boosted incomes accruing to owners of capital.
• These trends indicate that while rising productivity in recent decades provided the potential for a substantial growth in the pay for the vast majority of workers, this potential was squandered due to rising inequality putting a wedge between potential and actual pay growth for these workers.
• Policies to spur widespread wage growth, therefore, must not only encourage productivity growth (via full employment, education, innovation, and public investment) but also restore the link between growing productivity and the typical worker’s pay.
• Finally, the economic evidence indicates that the rising gap between productivity and pay for the vast majority likely has nothing to do with any stagnation in the typical worker’s individual productivity. For example, even the lowest-paid American workers have made considerable gains in educational attainment and experience in recent decades, which should have raised their productivity.

The graph that Professor Krugman uses in his blog post today Productivity and Pay is Figure C from the report.

The divergence between pay and productivity — a lot of productivity gains, almost total failure to trickle down — is one of the most striking features of American economics these past 40(!) years. It’s also the subject of endless attempts at debunking, of claims that the divergence is somehow a statistical artifact. What Bivens and Mishel do is take on these arguments carefully, not dismissing them completely, but showing that they explain only a fraction of what we see. Rising benefits are mainly a pre-1979 issue, explaining almost nothing since then; the “terms of trade” — consumer prices rising faster than the prices of U.S. output — is also mostly pre-1979, and in any case only a fractional concern. And so on.

One thing they don’t say explicitly, but is important: the next time you hear someone claiming that middle-class families have, in fact, seen a big rise in living standards, you should know that to the extent that this is true (which is less than claimed), it’s mainly about working more hours [i.e., the rise of the two-earner family.]

You should read the report Understanding the Historic Divergence Between Productivity and a Typical Worker’s Pay: Why It Matters and Why It’s Real, which is far to rich in detail and analysis to try to briefly summarize here.