The spring’s most anticipated event for me was not the Strokes’ new EP but the new income estimates from the Congressional Budget Office. CBO’s last release was in late 2014, and it presented income estimates only through 2011. The latest numbers are updated through 2013.
The CBO figures are invaluable because they combine tax data — which captures top incomes very well — with data from the Current Population Survey (which captures income below the top much better than tax data does). The CPS data also includes estimates of the value of non-cash benefits, including employer- and government-sponsored health insurance. CBO also estimates a variety of federal taxes and the distribution of their incidence.
Here are the highlights from the new data:
The Top One Percent
The top one percent’s share of pre-tax post-transfer household income was 15.0 in 2013. That was actually depressed from the value it would have taken if not for the tax increase on top earners that year.
Those at the top strategically realized income in advance of the hike in 2012, raising the top share above trend to 17.3 percent and depressing the 2013 share. Figure the top one percent share would have been between 16 and 17 percent in 2013 if there had been no tax increase. That would still leave it below the 2007 peak and probably below the 2000 peak too.
But income concentration is still headed upward, so it’s entirely possible we’re more unequal today than in 2007. There are big reasons to question the magnitude of the increase in inequality since 1979, but it’s very likely the trend is upward and continues to be.
After-tax and -transfer household income concentration, at first glance, was no higher in 2013 than in 1998 or 2004, at 12.4 percent. The trend since 2010 is downward, save the bump up in 2012. But again, this trend is an artifact of changing top tax rates. Figure that the no-tax-change estimate in 2013 would have been maybe 14 to 15 percent, almost certainly still below the 2007 peak of 16.7 percent.
The ubiquitous inequality estimates of Thomas Piketty and Emmanuel Saez (Excel), which apply to tax units rather than households, are based on pre-tax and -transfer income. They show the top one percent’s share as 21.5 percent in 2000, 23.5 percent in 2007, and 20.1 percent in 2013. The corresponding CBO estimates for pre-tax and -transfer income are 18.9, 21.3, and 18.2. The CBO numbers are lower mostly because they are based on households. Two roommates constitute two tax units but one (richer) household. Two parents and a burger-flipping teenage daughter constitute two tax units (at least one of them with very low income) but one household. Examining tax units pads the low end of the income distribution.
Back to the CBO data: between 2009 and 2013, the top one percent received 50.5 percent of the total increase in pre-tax post-transfer income. If that sounds high, consider that they incurred 56.8 percent of the income losses from 2007 to 2009. (This is just another way of saying that income concentration hadn’t returned to its 2007 level by 2013.) Two-thirds of post-tax and -transfer losses were absorbed by the top one percent from 2007 to 2009, and two-thirds of the gains accrued to them from 2009 to 2013. Neither average income nor the average income of the top one percent were back to 2007 levels in 2013.
The Middle Class
The best way to view the state of the middle class is to look at the median household incomes CBO reports. The pre-tax and -transfer median was only 12 to 20 percent higher in 2013 than in 1979 (depending on whether or not incomes are adjusted for household size).
But you should remember that the baby boomers have been retiring, and retirees get a large chunk of their income from transfers (Social Security and Medicare, primarily). As retirees become a bigger share of the middle class over time, their lack of pre-transfer income pulls down the median.
To see that this makes a difference, the pre-tax and -transfer income among households in the middle fifth rose by 4 percent from 1979 to 2013.* Among households with nonelderly heads in this middle fifth, though, the increase was 17 percent. Unfortunately, CBO doesn’t report the median income trends for nonelderly households separately.
In the nearer term, the distinction between elderly and nonelderly households makes less of a difference. Median pre-tax and -transfer income was lower in 2013 by 2 to 3 percent than in 2000 and lower than in 2007 by 8 to 9 percent. It was higher than in 2009 by 1 to 2 percent.
After taking transfers into account — which mostly involves Social Security, Medicare, and unemployment insurance within the middle class — the increase from 1979 to 2013 was from 29 to 37 percent, and median income rose by 7 to 8 percent from 2000 to 2013. It was lower by 3 percent than in 2007 and higher by 1 to 2 percent than in 2009. The 2013 median was higher by almost $18,000 than the 1979 median.
Finally, after accounting for taxes, which have declined significantly since 1979, median income rose 40 to 47 percent. From 2000 to 2013 the increase was 13 percent. The median was lower than in 2007 by 1 percent and lower than in 2009 by less than 1 percent.
This post-tax and -transfer measure is what really counts — it’s the closest measure we have to disposable income. From 1979 to 2013, the median grew by just shy of $20,000. Cite that number the next time the Economic Policy Institute tells you the middle class is doing terrible.
Incidentally, if you were to use the Piketty-Saez estimates (which use an inferior inflation adjustment and mix the elderly and nonelderly while examining pre-tax and -transfer income) you would conclude the bottom 90 percent was worse off by over $3,000 in 2013 compared with 1979.
After taxes and transfers, the bottom fifth of households was richer by 46 percent in 2013 than in 1979 (by $7,700). This should be considered something of an upper-bound estimate because it depends heavily on the way that Medicaid and Medicare benefits are valued. In the last set of estimates CBO provided, they included a tab in the spreadsheet that showed the same estimates but using an alternative valuation for health benefits. Rather than rising by 48 percent from 1979 to 2011, the alternate estimates showed a 26 percent increase. Still!…
Occasionally, you’ll see the claim that even the pre-tax and -transfer (market) income of the bottom fifth rose substantially over time. From 1979 to 2013, for instance, among households with children in the bottom fifth, the rise in market income was 44 percent.** This figure is not wrong, but it is important to remember that CBO includes non-earners in these estimates. The market income of households at the bottom would increase less over time if households without earnings were excluded. Much of the increase at the bottom may be traced to an increase in work, particularly among single mothers, after welfare reform.
From 2000 to 2013, post-tax and -transfer income in the bottom fifth rose by as much as 18 percent (or perhaps by as little as 10 percent using the alternative health benefit valuation). It rose by 1 percent between 2007 and 2013 (or was flat). And it fell by at least 2 percent between 2009 and 2013.
That gets you up to speed. Carry on and keep using the PCE deflator.
* Technically, the middle fifth of the pre-tax income distribution. Ideally we’d have these estimates for the middle fifth of the pre-tax and -transfer distribution, but alas.
** Again, this is technically the bottom fifth of the pre-tax income distribution.