Reasons Behind the Apparent Divergence of Productivity and Compensation

Published August 27, 2013

In some but not all industries, we can measure output; e.g., tons of this, cubic yards of that. By relating output to hours of labor, we have a measure of labor productivity.

As measured by the U.S. Bureau of Labor Statistics (BLS), labor productivity has risen by about 150 percent since 1967. However, wages adjusted for inflation have hardly risen since 1967.

The apparent divergence between labor productivity and earnings has been noticed by various progressive think tanks (e.g., the Economic Policy Institute) and is now making its way into public discourse. The divergence is easy to explain.

Reason #1. Overstatement of Inflation

The first reason for the divergent trends between labor productivity and real hourly earnings is a bias in the Consumer Price Index. Economists believe the CPI overstates the rate of inflation by something like 1.5 percent because of inadequate adjustment for quality changes and inadequate adjustment for substitution effects. Dividing earnings by the Implicit Price Deflator (IPD), instead of dividing by the CPI, accounts for a substantial part of the apparent divergence of labor productivity and compensation.

Reason #2. Ignoring Benefits and Employer-Paid Taxes

The second reason for the apparent divergence of labor productivity and compensation is that a growing portion of the cost of labor is in the form of benefits and employer-paid taxes. These benefits and employer-paid taxes include employer contributions to retirement plans and health insurance policies, paid holidays and vacation days, and the employer-paid portion of the Social Security tax. Using hourly compensation, which includes benefits and employer-paid taxes, instead of hourly earnings, accounts for most of the remaining apparent divergence of labor productivity and compensation.

Reason #3. Overstatement of Productivity

The third reason for the apparent divergence of labor productivity and compensation is an overstatement of labor productivity. There are two parts to this.

(3A) The measure used by the BLS for the productivity of labor is gross output. Gross output does not account for depreciation of physical capital and the amortization of intangible capital, what is called “capital consumption” in the National Income and Product Accounts. If the rate of capital consumption were constant, the distinction between gross output and output net of capital consumption would not be important when looking at changes in labor productivity. But the rate of capital consumption has accelerated a bit during the recent past.

(3B) Another part of the overstatement of productivity is because part of the increase in output in recent years should be attributed to the increased use of imports in production. If the amount of imported goods and services used in production were constant over time, accounting for the use of imports in production would not be important when looking at the change in labor productivity. But the use of imports in production has grown substantially during the recent past.

Compensation, properly measured, does track productivity, as it must.