Rising wage inequality continues to be a defining feature of the U.S. labor market

It’s well documented and widely understood that wage inequality has grown dramatically over the last four decades as productivity and compensation growth have become delinked. Despite an expanding and increasingly productive economy, wages have stagnated for the vast majority. Looking at the most recent data—through the first half of 2016—we see that wage inequality has continued to grow, with top earners faring far better than those in the middle or bottom of the wage scale. First, the data paints a striking picture of growing wage inequality since the last business cycle peak in 2007. Second, average wage growth overall is slow, and any significant real wage growth continues to be driven by low (and below target) inflation—not meaningful acceleration in nominal wage growth. Last, strong payroll employment growth the last couple of months suggests positive future trends for not only wage growth, but also declining unemployment and rising labor force participation.

The figure below shows the disparate growth across the wage scale from 2007 to 2016 (using the most recent data, for the first six months of 2016, and comparing to wages at each decile and at the 95th percentile to those of the comparable period in 2007). I’m comparing the first half (FH) of both years to maintain the same seasonality in the data. Plus, it gives us a glimpse of what’s happened through the first half of this year. Except for the lowest wages—at the 10th percentile—what you see is a clear increase in growth as one moves up the wage distribution, from negative growth at the 20th percentile (-2.8 percent) to the fastest growth at the 95th percentile (10.4 percent). At the median, real wages grew a total of only 0.8 percent over the nine-year period. So, not only did the labor market since 2007 perpetuate the wage stagnation and inequality of the previous three decades, it has actually exacerbated it.

Figure A

Wage changes by decile and the 95th percentile, FH2007–FH2016

Percent change FH2007-FH2016
10th 2.6%
20th -2.8%
30th -0.4%
40th 0.0%
50th 0.8%
60th 1.0%
70th 2.5%
80th 3.9%
90th 7.8%
95th 10.4%
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: The xth-percentile wage is the wage at which x% of wage earners earn less and (100-x)% earn more.

Source: EPI analysis of Current Population Survey microdata

Copy the code below to embed this chart on your website.

Why did wages at the very bottom fare better than wages in 20th through 60th deciles? The answer is that the federal minimum wage was increased in both 2008 and 2009, and there has been a string of (legislated and indexed) state-level minimum wage increases that supported wage growth at the bottom of the wage distribution.

For those interested in the nitty gritty details, the table below provides the data on real wages across the distribution for FH2007, FH2015, and FH2016. The tables also displays two key wage ratios that measure inequality, the 50-10 and the 95-50 wage ratios, and the changes in wages and wage inequality over the last year and since 2007. Again, this demonstrates the wage bolstering effect of the minimum wage at the bottom (keeping the 50-10 ratio in check) and the pulling away at the top (increasing inequality as measured by the 95-50 wage ratio). These data do not allow a look inside the top 5 percent, but other data sources point to the top of the top as the real winners in today’s economy, taking home increasing shares of wage growth.

Table 1

Hourly wages by wage percentile, FH2007–FH2016 (2016 dollars)

Wage by percentile Wage ratio
Year  10th  20th  30th  40th  50th  60th  70th  80th  90th  95th  50-10  95-50
FH2007 $8.95 $10.93 $12.76 $14.99 $17.47 $20.69 $24.45 $30.15 $40.21 $52.36 1.95 3.00
FH2015 $9.03 $10.23 $12.41 $14.98 $17.22 $20.34 $24.86 $30.98 $42.38 $56.07 1.91 3.26
FH2016 $9.18 $10.62 $12.70 $15.00 $17.60 $20.90 $25.06 $31.34 $43.35 $57.83 1.92 3.29
Percent change
2015-2016 1.7% 3.8% 2.4% 0.1% 2.2% 2.8% 0.8% 1.1% 2.3% 3.1% 0.01 0.03
2007-2016 2.6% -2.8% -0.4% 0.0% 0.8% 1.0% 2.5% 3.9% 7.8% 10.4% -0.03 0.29

Note: The xth-percentile wage is the wage at which x% of wage earners earn less and (100-x)% earn more.

Source: EPI analysis of Current Population Survey microdata

Copy the code below to embed this chart on your website.

I would be remiss if I didn’t remind readers that much of the real (that is, inflation-adjusted) growth in wages, such as it has been for large swaths of the population, is driven in part by historically low inflation in the past few years. In fact, the largest across-the-board increases in real wages occurred between FH2014 and FH2015, when price growth was essentially zero. Very low inflation is also partially responsible for the real wage growth over the last year. Relying on low inflation to increase living standards is a poor long-term strategy because (1) declines in commodity and energy prices are not sustainable and (2) low permanent inflation would eventually slow nominal wage growth.

Taken together, this reinforces the importance of the Federal Reserve keeping interest rates low in the foreseeable future to allow the economy to tighten and wages to grow across the board. By my calculations, a continuation of the strong payroll growth we’ve seen the last couple of months into the next year will surely mean a tighter economy in the near future, with a lower unemployment rate, higher labor force participation rate, and stronger wage growth.

That’s the story the data tell. For those interested in some minor changes in EPI’s calculation of wage deciles, read on. EPI’s real hourly wage series are the hourly wages of 18-64 year olds, which includes all wage and salary workers with valid wage and hour data, whether paid weekly or by the hour, adjusted for inflation by the CPI-U-RS. If a valid hourly wage was reported, that wage was used throughout our analysis. For salaried workers, the hourly wage is their weekly wage divided by their hours worked. That hasn’t changed. Unless otherwise specified, that’s how we at EPI calculate hourly wages. What has changed, albeit slightly, is how we smooth hourly wages to compensate for wage clumps in the wage distribution. Wage clumps, which exist when a nontrivial share of the distribution sits at a particular wage, can lead to an unnatural stagnation or jump in wage decile cutoffs over time. Historically, the smoothing technique involved creating a categorical hourly wage distribution, where the categories are 50-cent intervals (also referred to as 50-cent bins). Wage decile values are calculated as a weighted average between neighboring bins. We have recently made a couple of changes to this technique. Primarily, we have narrowed the bin size to 25-cent intervals, which we believe will lead to more accurate calculation of decile values in parts of the distribution that span a narrow wage range, particularly affecting the historical series of smaller demographic subsamples. For uniformity, we are applying these changes across all our wage analysis by decile and at the 95th percentile. Below you can see the effect of these revisions on hourly wages and wage changes. In general, wage levels move within a few cents of their original value and longer term trends in wages tell a consistent story of rising wage inequality.

Table 2

Changes in hourly wage decile methodology, FH2007 and FH2016 (2016 dollars)

10th 20th 30th 40th 50th 60th 70th 80th 90th 95th
Original method FH2007 $8.98 $10.88 $12.73 $14.95 $17.50 $20.69 $24.47 $30.15 $40.45 $52.75
FH2016 $9.19 $10.68 $12.72 $15.00 $17.68 $20.87 $25.11 $31.23 $43.55 $57.70
Percent change 2.4% -1.8% 0.0% 0.3% 1.0% 0.9% 2.6% 3.6% 7.7% 9.4%
Revised method FH2007 $8.95 $10.93 $12.76 $14.99 $17.47 $20.69 $24.45 $30.15 $40.21 $52.36
FH2016 $9.18 $10.62 $12.70 $15.00 $17.60 $20.90 $25.06 $31.34 $43.35 $57.83
Percent change 2.6% -2.8% -0.4% 0.0% 0.8% 1.0% 2.5% 3.9% 7.8% 10.4%

Note: The xth-percentile wage is the wage at which x% of wage earners earn less and (100-x)% earn more.

Source: EPI analysis of Current Population Survey microdata

Copy the code below to embed this chart on your website.