Monday, August 3, 2015

Working age real median household income (very nerdy)


 - by New Deal democrat

After my post yesterday on whether average Americans are better off than in 2008, I had a brief exchange with Doug Short.  Doug pointed out that the Census Bureau will be releasing its detailed information next month for 2014 median household income, including by age cohorts.  He wrote about this nicely last September, and his post included this graph (through 2013), breaking down median household income by age:



In my article yesterday, I included the following graph as an effort to estimate median household income for the prime working ages of 25 - 54, in which I took inflation-adjusted median wages, and subtracted out the percentage of unemployed:



Upon further reflection, probably a better adjustment is by using the employment-population ratio for ages 25 - 54, which gives me the next graph:



By this measure, median working age household income in 2015 so far has been slightly higher than 2008, although 1% or so under 2007 and before.

Still, looking at Doug's graph above, which shows prime working age household income generally stalled at about 3%-4% under their 2008 levels through 2013, that seems like a lot to ask of one year. But surrounding circumstantial data does suggest that there will be a significant increase in 2014.

First of all, not only did real median wages increase, due mainly to a big decline in gas prices, but the employment-population ratio for ages 25  54 jumped higher in 2014 and so far this year:



Also, as of last August, the National Employment Law Project reported that beginning in late 2013, and even moreso in 2014, job growth accelerated to the point where higher paying jobs were close to 40% of all jobs created.

And Professor Saez's preliminary look at income reported to the IRS in 2014 suggested that the bottom 90% had their best year so far this Millennium, with an increase of about 3.8% (although even that is pitiful compared to the 58% share of income growth taken by the top 1%).

 What the Census Bureau will report next month is also affected by how they perform their calculations.  Do they compare year-end with year-end? In that case, real income will be higher, since the late-year deflation will be fully reflected. Or will they compare  the average over 12 months, with the average over the previous 12 months? In that case, there will be a more modest gain.

 I continue to think that adjusting the real median wage by the employment to population ratio for the prime working age group is the best way to estimate working age median household income on a timely basis.  So I will go ahead and forecast that the Census Bureau will report a solid gain for that group next month, and we'll see if the forecast pans out.