Workers’ Pay Hasn’t Always Lagged Productivity Growth

In a recent blog post, we introduced a calculator showing how much higher your wages (for a stylized definition of “your”—more on this below) could be if they had kept pace with economy-wide productivity, defined as economic output produced in a given hour of work. Why is this relationship between wages and productivity interesting? Well, for decades following World War II, wages across the board closely tracked productivity. In recent decades, however, while average wages (almost) track productivity growth, wages for the vast majority of Americans have not. So, telling people what this wedge between productivity and wage-growth means in dollar terms seems like a useful way to make rising inequality less abstract and more salient.

The calculator itself is pretty spare, so I figured I’d take this blog post to explain a little more about our methods. First, we constructed the actual wage distribution for 2012 by using the BLS’s CPS-ORG data (Bureau of Labor Statistics Current Population Survey Outgoing Rotation Group).

Next, when somebody inputs their own wage into the calculator, we use this 2012 distribution to figure out where they fall in the overall wage distribution. Specifically, we are looking at where that wages falls with respect to the ventile cuts in that distribution—that is, for instance, between the 55th and 60th percentile of the wage distribution. Next, we look at that same point in the wage distribution of 1979.

Lastly, we then calculate what that given 1979 wage would have been had it grown at the same rate as total economy productivity until 2012 (the total economy productivity data also comes from the BLS). So, when we say the calculator shows what “your” wage could have been, it basically means what the wage of people at the same point in the wage distribution as you are today could have been had it kept up with productivity growth. A useful feature of doing the calculation this way is that the degree of wage-loss varies by initial wage-level—that is, the loss from failing to track productivity is not simply the same scalar X for everybody. And for those who were above the 95th percentile in 1979, well, wages at this point and above have tracked productivity closely enough that we basically call it a wash and say they’ve lost nothing at all from rising inequality.

A summary measure of growth in wages and productivity is illustrated in the figure below. Here, however, while we are measuring things a bit differently, the basic story is the same. The darker blue line here is hourly compensation for production and non-supervisory workers in the private sector, which represents approximately 80% of the workforce. Compensation represents not just wages, but also non-wage benefits (estimated using the BEA’s NIPA (Bureau of Economic Analysis National Income and Product Accounts)). This includes items such as health insurance, thereby addressing concerns that wages may be squeezed by rising health costs.

In this chart, we are also measuring productivity slightly differently than in the wage-productivity calculator (illustrated in the lighter blue line). Since the calculator was constructed (which saw its first life on EPI’s website), measures of calculating GDP (the denominator of productivity) have been revised in ways that increase productivity significantly, greatly boosting the wage/productivity gap. However, the revision to GDP also significantly boosts the ratio of depreciation to overall economic growth, and since depreciation essentially represents lost output, we switched our productivity measure to reflect the growth rate of net productivity, as it is more reflective of what is available in the economy to boost workers’ wages (for some more discussion of the revisions to GDP and why they make accounting for depreciation more important, see this). However, the basic story remains the same—even this slower metric of productivity growth illustrates a growing chasm between hourly compensation and net productivity over the last three decades.

What is particularly striking is how the relationship between productivity and most workers’ compensation changed dramatically. From 1948 until the 1970’s, hourly compensation tracked net productivity very closely. As workers became more productive, they were compensated for their increase in productivity. Starting in the late 1970s, hourly compensation began to lag net productivity and has shown very little growth compared to the growing productivity we’ve experienced. We maintain and describe in painstaking detail in State of Working America how this disconnect between a growing economy and growth in wages for most of the population was the results of intentional policy decisions on taxes, trade, labor, and financial regulation. But that’s the good news: if inequality is not inevitable, then it can be fixed. Check out to learn more about the policy decisions and what can be done about it.