Wednesday, September 13, 2017

Today’s Census data on poverty, income, and heath insurance. [feedly]

Today's Census data on poverty, income, and heath insurance.
http://jaredbernsteinblog.com/link-to-my-comments-on-todays-census-data-on-poverty-income-and-heath-insurance/

A solid report, showing gains across the spectrum. But inequality's up too, and median earnings, not so much…

My data dive in the WaPo underscores the clearly favorable results in the report, but here are a few other factoids to consider:

–While this isn't the best data for inequality analysis, for reasons I note in the WaPo, my piece points out the relative difference between gains at the 10th and 95th percentile. That observation is correct, but the 10th %'ile is a bit of a negative outlier. Better to look at a more stable statistic, the average real income gain for the bottom fifth, up 2.6% last year, compared to a 5.6% gain among the richest 5% of households. The bottom half gained last year, but not as much as the top.

–It's also true that incomes shares going to the middle and low income households are at all time lows, as the figure reveals. (See note in WaPo piece, however, re the impact of the 2013 survey change on comparisons like this. I think it's a legit comparison, and it comports with other, better inequality data–where better means inclusive of more data sources, including taxes, more transfers, and capital gains–showing even more growth in inequality.)

Source: Census Bureau

–The lack of change in real median earnings for full-time, full-year workers last year is worth noodling over a bit. It surely reflects a composition effect as lower-paid were drawn into the sample last year, pulling down the median (see here for how this works). But even considering that reality, look at this series for men since 1960:

Source: Census Bureau

Sure, there's composition effects embedded in there, but they don't explain away the very long-term stagnation of the series. I mean the median full-time guy earns about the same in 2016 as in 1970!

The trend for women is considerably more positive, but it too hasn't gone much of anywhere since around 2000.

Source: Census Bureau

So, it's a really solid report, no question, but structural problems persist.


 -- via my feedly newsfeed

No comments: