The Decoupling Of Labor Productivity And Median Wages

At the OECD Insights blog, James Plunkett announces findings from the Professor John van Reenen and Joao Paulo Pessoa of the London School of Economics study funded by the Resolution Foundation, Decoupling of Wage Growth and Productivity Growth? Myth and Reality. The research gives keen insight to the problems of median wage stagnation, concluding that there has been a great amount of decoupling between labor productivity and median hourly wages in the UK. Median hourly wages were essentially flat for the past two decades but for a four year growth.

The study defines an important distinction between net and gross decoupling; one which supports the narrative and one which does not. While the first, net decoupling, defined by van Reenan and Pessao as the difference between GDP growth per hour (labor productivity) and average compensation, with GDP deflator taken into account for both, was found not to have diverged, gross decoupling has. Gross decoupling is much more critical; the relationship between GDP growth per hour and median worker wages. The authors stress the usage of wages in gross decoupling rather than compensation, as wages underline the effect on the median worker whereas compensation is more in the realm of banker’s bonuses, or pensions and health benefits. That although labor’s share of income has not fallen off, the median worker’s share has, confirming a rise of inequality between the average worker and the higher-ups.

For worker’s median wages in the UK, the authors analyzed the Labour Force Survey which measures the pre-tax wages of 60,000 UK households each quarter. In their treatment of various UK labor statistics, the authors find that net decoupling is not existent but that there is a 43% difference between median hourly wages and mean hourly compensation versus labor productivity. The statistics also show that the trend line for mean hourly wages cuts halfway through the disparity starting from the mid-late 1990’s and ending with their final data point, 2010. The authors conclude that this means about half of the divergence between median hourly wage and labor productivity can be explained by inequality. The other half is due to the doubled growth of compensation versus wages.

To quantify the driving factor of compensation in this growth versus median wages, the authors use UK Office of National Statistics data for non-wage compensation from 1999-2007. Over this period, Van Reenan and Pessao found employers’ contributions to national insurance schemes to rise 67% and employers’ contributions to pensions rose 98% while wages rose 47%. ONS and LFS data was analyzed over a four decade period as well. By 2010 gross decoupling was 42.5%, with more than three-quarters of the decoupling due to inequality and wage/compensation divergence. The other factors are a combination of GDP deflator/RPI differences, the gap between employed and self-employed earnings, ONS/LFS growth disparties.

Plunkett asserts that the difference in net decoupling and gross decoupling should not be discounted as a sort of anomaly. Although compensation is a more inclusive quantifier of benefits, wages are vastly more accrued by workers who earn less. Higher paying jobs get the lions share of compensation benefits. The nature of jobs in the lower half of earnings distribution means that their wages are much more essential for survival based on their reliance on wages to pay for every day needs. Since this measure “most accurately captures how well off people feel”, the indication is that the stagnation has primarily sat on the back of those with lower shares of labor’s share of income.

Jared Bernstein and Lawrence Mishel reached similar conclusions in their 2006 paper, The Growing Gap Between Productivity and EarningsThe paper aimed at contending the narrative set out by President Bush’s chief economist, Edward Lazear, and Federal Reserve Governor Randall Krosner that recent American productivity gains are a “success story.” Bernstein and Mishel argue against the prevailing macroeconomic theory that productivity gains beget wage gains which began rising standards of living. The authors deflate compensation data with the GDP deflator instead of CPI, since the GDP deflator takes investment goods, which have grown at a slower rate than consumption goods, into account. Use of the CPI to deflate rates of compensation growth would not tell the correct story.

Bernstein and Mishel tackle the inequality wedge as well, quoting from Robert Gordon and Ian Drew Becker’s 2005 paper, Where did the Productivity Growth Go?;

The standard link between the standard of living and productivity growth is broken by our finding that over the entire period 1966-2001, which encompasses the period of the 1965-1979 productivity growth slowdown and subsequent 1995-2005 productivity growth revival, only the top 10 percent of the income distribution enjoyed a growth rate of total real income (excluding capital gains) equal to or above the average rate of economywide productivity growth. The bottom 90 percent of the income distribution fell behind or even were left out of the productivity gains entirely.

The authors also analyzed decoupling through plotting productivity against real hourly wages of non-managerial workers, which make up about the bottom 80% income distribution in the U.S., from 1966 to 2005. The decoupling starts in the early 1970s and as productivity rises from a baseline of 100 to nearly 220, real non-managerial wages are nearly stagnant; wavering between 110 and 100, ending up near 110 by 2005. For the authors along with this inequality aspect, the difference in GDP and CPI deflators, and the increase in capital’s share of national income as opposed to labor’s share, are the three main factors that explain decoupling in the U.S.

The paper then goes on to try to explain why the productivity/wage gap is symptomatic of something within the U.S. economy. Bernstein and Mishel contend that marginal production theory, that the a worker’s marginal wage is based on his or her productivity, does not fit well with the gap since there are many externalities that make the assumption somewhat moot. To understand the economic forces at play, the authors looked at 20th percentile real wages, productivity, and unemployment for the periods of 1984-1989, 1995-2000, and 2000-2005. To preface this, the authors look at unemployment rates from 1947-1973, a time when the split between productivity and wages was non-existent, which averaged 4.8%. Even with demographic and social shifts, from 1973-2005, the unemployment rate averaged 6.3%. Examining the return to full employment in the late 1990’s is crucial to the discussion. The data shows that at the height the business cycle’s unemployment rate of 1989 and 2000, unemployment was 5.3% and 4.0% while, over the period, productivity grew at an annual rate 1.5% and 2.5%, respectively, but 20th percentile real wages from 84-89 fell 0.1% annually while growing 2.3% annual from 95-2000. The authors looked at 2000-2005 data to extrapolate the findings; unemployment went back up to 5.1%, productivity rose 3.1% annually, while 20th percentile real wages only rose 0.1% annually. Only during the spell of full employment in the late 1990s did worker’s wages in the lowest quintile of the distribution scale rise in tandem with productivity gains. The authors attribute the gap and the violation of marginal productivity theory to particular social forces such as lessened worker’s bargaining power, a drive to productivity with the least labor costs as possible, wage caps, and cheap immigrant labor .

Next, Bernstein and Mishel try to discern whether or not rising health care costs explain the decoupling. The authors steadfastly say no, citing that 48% of workers do not get health care through their job; for workers making under $15, more than 6 of 10 do not. If health care coverage were filling in some of the gap between productivity and median wages, one would expect the rate of  wages to grow faster for workers without health care coverage. They looked at 10th, median, and 90th percentile wages over 2000-2005 and compared them to employer-provided health care coverage at the same percentiles. Instead of rising costs filling in the gap, the opposite occurred, with wages growing faster for workers with a greater share of health care coverage. Exacerbating this health care wedge was the declining percentage of health care as employer’s costs; rising costs begot less coverage.

Bernstein and Mishel take fault with Lazear’s and other economist’s explanation that lags have caused the decoupling and that productivity will soon level off. But structural changes within the U.S. economy over a period where full employment is the rarely seen point to decoupling as a fixture.

Peter Harrison’s study, Median Wages and Productivity Growth in Canada and the United States, found that from 1980 to 2005, real median hourly wages rose an average of 0.33% per year while labor productivity saw gains of 1.73% per year. This gap of 1.40% was driven mainly by inequality, which explained half of the split between productivity and wages. In Canada, the gap was a bit less, 1.26%, and inequality only accounted for 25% of the gap. Canada saw greater changes in supplementary labor income and labor’s terms of trade. Both the U.S. and Canada’s decline in labor’s share of income contributed to the decoupling as well; 17% and 20%, respectively. Harrison’s research pointed directly to the rising share of the top one percent of income distribution and the leveling or declining income shares elsewhere as the main drivers of the inequality. This data supports the ongoing struggle within American politics and social discourse about inequality between the 99% and 1%.

In the U.S. as present day recessionary forces press companies to cut back, they attempt to retain the most productive and talented of their work force. In order to keep these employees they may offer better compensation packages to attract the necessary talent, but try to accomplish a higher standard of work with less hires. Also, as the talent pool that employers realistically look at shrinks, less and less people get employee training that would have happened in the decades past. So the share of workers that can be competitive within these tough hiring environments drops as a result. As a longer trend, unions have been gradually declining and are now nearly insignificant in the U.S. labor landscape. Globalization also surely plays a part in the split as productivity gains can be seen by moving production of goods to countries with cheaper labor while the company profits stay at the top of the ladder. Plunkett turns to Jared Bernstein’s analysis that the problem is even more damaging as workers’ wages were not increasing during a time when the UK’s economy is growing and productivity was rising. Now more than ever, the average minimum wage worker in the U.S. will feel the consequences of these changes as inflation drives up the cost of necessity goods. A bigger and bigger cut of wages will go towards food, transportation, and shelter.

1 comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: