Courtesy of Doug Short.
My virtual acquaintance New Deal Democrat has posted an interesting article on real (inflation-adjusted incomes) based on annual IRS tax data through 2013. His discussion includes some comparisons between the Census Bureau’s median Household Income data and the Family Unit average income. A Family Unit is the term used for an IRS designated Tax Unit (e.g., a couple with dependents, or a head of household with dependents, or a single person).
At NDD’s suggestion, I’ve created a chart that overlays the annual data from 1967-2013, the timeframe for which we have annual data from both sources.
The highest income series in the chart above is for the average (mean) household. It reflects the negative skew that high income households have on the median (e.g., the top 5%, 1%, 0.1%). The median (middle) household income closely matches the two IRS series. Interestingly enough, the median household income more closely matches Family Unit income including capital gains in the earlier years until the mid-1990s (aka the “roaring ’90s”), at which point the those cap gains formed twin peaks with the Tech and Financial bubbles. Since the turn of the century, the Median household income is a pretty close match to the IRS’s Family Unit.
We’re now into 2015, but it will be many months before we have official government data for 2014. The CB won’t publish the 2014 data until mid-September, and the IRS data won’t be available until months after that.
Fortunately, however, Sentier Research provides timely data for median household income on a monthly basis. See my latest analysis of their findings here. The good news is that the trend in 2014 has been upward.
NDD and I are indebted to Professor Emmanuel Saez of UC Berkeley for compiling the IRS data.
