Real wages from 1995 in quintiles

- by New Deal democrat

A week ago I published a surprising graph that showed that average (i.e., mean) wages for nonsupervisory workers had gone up almost 15% since 1995, even after adjusting for inflaiton. Here it is again



Since owners and managers are not included in the statistic, you can't just lay it off on the 99% vs. the 1%. Among the questions were, what has happened to the median (50th percentile) as opposed to mean wage? And what happens when we divide the workforce into quintiles by income?

It turns out it makes a lot of difference. [Note: special thanks to my co-blogger SilverOz for producing the below graphs]

To begin with, here is median income for the middle quintile since 1995 -- the low point of the graph above:



Measured at the median, there has been only about a 2% increase remaining as of 2010.

As you move up the income scale, things get better. Here's the quintile including the more affluent part of the middle class:



Even after the "great recession," this slice of the population is earning about 5% more than they were in real terms in 1995.

At the top end, the comparison is even more favorable:



This group is still about 10% ahead of where they were in real terms in 1995.

At the other end of the pay scale, the lower income working class is actually earning about 1% less than they were in 1995:



And for the poor and working poor, the compaison is even worse. They are earning 7% less than they were in 1995:



Measured from 1995 through 2010, the +$4500 increase in the affluent middle class and the $15,000 increase for the upper middle class and top income earners overwhelm the -$400 loss of the lower working class and the -$900 of the poor and working poor. This accounts for the difference between the average real wage and the median real wage.

It's also interesting to not that the tech boom of the late 1990s really was a rising tide that lifted all boats, whereas the Bush expansion was limited to the more affluent two quintiles. Since then, the great recession and its aftermath have affected the incomes of the highest and lowest quintiles the most.

The trend in real wages, an update

- by New Deal democrat

I want to thank the thoughtful commenters to my post the other day concerning the surprising upward trend in real income measured since 1995.

One suggestion was to use the "billion price project" instead of the CPI. The problem there is that the billion price project's data only goes back a couple of years, and so far has closely followed the CPI.

Another suggestion was to change the measure of wages. There is a measure of median rather than mean wages, that is calculated once a quarter. It has shown lower wage increases than the monthly average wage measure, but it is only 11 years old.

But it does make a significant difference. Here is a comparison over the life of the median wage series (in red) vs. average nonsupervisory hourly wages (blue):



Both tracked similarly until the middle of the last recession, but ever since the late 2008 deflation in gasoline prices, average wages have shown a much more positive trend. Over the entire 11 year period, median wages are up less than 2% vs. average nonsupervisory wages, up over 6%.

Another suggestion was to break the trend down into income quintiles. The typical graph of these since 1974 already adjusts for inflation, and shows an actual and significant decrease in the real wage of the bottom 20%. I'll see if I can find a way to post data starting from the mid-1990s and if I can, I'll follow up with a further comparison.