David R. Henderson  

From 2007 to 2012-13, The Income Share of Top 1% Fell

The housing bubble: Perception... How much for a book dedication...
The share of income (including capital gains) held by the top 1 percent grew from 10 percent in both 1960 and 1980 to 21.5 percent in 2000. Since then, it fell to under 17 percent in 2002 before rising to 23.5 percent in 2007. With the onset of the Great Recession, the wealthiest people actually had their incomes decline by the greatest amount as the share of the top 1 percent fell to 18 percent in 2009 before rising to 21.4 for the 2012 and 2013 on average. (These two years have to be combined because changes in tax rates for the very rich led to many people moving income and capital gains to 2012 -- which resulted in much lower incomes in 2013). The movements of the top .01 percent followed a similar trajectory.

Mr. Saez, in a September 2013 paper titled "Striking It Richer," created a splash when he reported that 95 percent of the gains of the growth between 2009 and 2012 were captured by the richest 1 percent. Other analysts have done similar calculations. Together, they have created the impression that the rich were increasing their share of the economic pie at an alarming rate.

But these "shares of growth" can be misleading over short time spans. The top 1 percent share's rose from 2009 to 2012, but only after it declined more from 2007 to 2009. The finding that inequality has exploded recently is based on the bounce-back from the unusually low level of capital gains in 2009. If we look at the whole period from 2007 to 2012-13, the share of the top 1 percent decreased as the income loss for the top 1 percent was higher than for other groups.[italics added]

This is from Stephen J. Rose, "Measuring Inequality Trends During the Great Recession," New York Times, February 17, 2015.

Rose is a research professor at the George Washington University Institute of Public Policy.

Of course, these data don't track the same people over time. There is much movement in and out of income categories. That's a big caveat. Still, the data are interesting given how much we've heard about the top 1% getting almost all the gains.

HT to Alan Reynolds.

Comments and Sharing

CATEGORIES: Income Distribution

COMMENTS (15 to date)
Robin Hanson writes:

Well it makes sense that if the very rich were getting most of the gains when things get better, they'd suffer most of the losses when things get worse.

JLV writes:

Other things that are true:

the top 1% share increased from 2008-2013.
The top 1% share increased from 2007-2012.
The top 1% share increased from 2001-2007.

Because of the fluctuations in the series including capital gains, its relatively easy to find 5-7 year windows where the top 1% share either increased or decreased. Its much harder to find a 10-15 year windows after 1970 where the top share decreased.

(And, by Rose's own admission, the dip in 2013 is likely temporary, the result of tax shifting.)

JLV writes:

I eyeballed the graph wrong, so those dates are a year off, I think: should be 2009-2013, 2006-2012.

John Thacker writes:
And, by Rose's own admission, the dip in 2013 is likely temporary, the result of tax shifting.

Ah, but once you put tax shifting into the equation, there's also a fair case that the shift in monetary gains to the 1% overestimates inequality.

Suppose we have a world where taxes are very high, and much of the 1% chooses to take their compensation in non taxed "three martini" business lunches, company cars, executive washrooms, and the like. Also assume that high marginal tax rates cause them to invest more in tax shelters, muni bonds, and other things that decrease the (measured) before-tax income while preserving after-tax income.

Now suppose we eliminate those loopholes, and the wealthy choose to take more of their compensation in taxed money. Also suppose that more small business owners choose to form S corporations that are pass-through to personal income taxes, instead of locking up small business revenue in C corps and the like. Measured income inequality should increase.

Therefore, we see that the PPACA tax on "Cadillac" health plans, if successful in shifting pay from expensive health plans to more wages, would also likely increase measured inequality.

Japan has very low income inequality, but even being a salaryman in a connected firm (or a connected bureaucrat) earns you a lot of perks and real consumption that money can't buy.

Kevin Erdmann writes:

John Thacker:
"Therefore, we see that the PPACA tax on "Cadillac" health plans, if successful in shifting pay from expensive health plans to more wages, would also likely increase measured inequality."

I think it might be the other way. Cadillac or not, the health plan for a $100k manager is going to be a lower proportion of total compensation than it is for a $30k line worker.

Health inflation has probably made inequality look worse. Giving everyone a cash wage instead would lower measured inequality, I think. Wouldn't it?

MikeP writes:

Now suppose we eliminate those loopholes...

You mean suppose 1986 happened?

JLV writes:

In re: benefits, at least in the CPS data (which I've worked with a bit), its possible to include employer contributions to insurance premiums as part of income (which has been asked in the March CPS since 1992). It doesn't move inequality much, though, if I remember correctly.

V writes:

I think the big issue is being missed here. Maybe the top 1% were hit the hardest in terms of income (I think probably in capital gains, as they have a grossly disproportionate allocation of total assets). However, I hardly assume they 'lost it all' or lost their livelihood. Whereas many 'ordinary' people saw almost all of their wealth diminish drastically (as most people's wealth is the equity in their house) and many people lost their job and could not find a new one. Then due to the one-percenters' ability to seek rents from government policies they were able to capture almost all of the recovery growth (I think simply most of their assets were undervalued during the crisis and then prices stabilized over time). Meanwhile many people struggled just to get by after the crisis, with real wages for the bottom 80% mostly stagnant for decades: http://cdn.billmoyers.com/wp-content/uploads/2013/09/Motherjones2.png

[broken html fixed--Econlib Ed.]

Mark V Anderson writes:

Mike wrote --

Now suppose we eliminate those loopholes...

You mean suppose 1986 happened?

I think that was John Thacker's point, although I wish he had been more explicit. The 1986 tax reform act did eliminate a lot of loopholes, and in turn lowered the tax rate. Both of these resulted in more taxable income, irrespective of any actual change in income, making it appear that inequality increased more than it did.

I think John is absolutely right, but the question is how big an effect was this?

BC writes:

"Of course, these data don't track the same people over time....That's a big caveat."

I don't think it's a mere caveat, and it doesn't undermine David's point at all. The top "1%" actually consists of about 12% [http://marginalrevolution.com/marginalrevolution/2014/04/mobility-in-and-out-of-the-top-one-percent.html]. When the stock market does very well, the top 1% probably consists of people that happen to own a lot of stock. When the stock market tanks, the top 1% consists of (a different set of) people whose income derives from non-stock sources. So, the so-called trends in income shares are really comparing 1/12 of the top 12% at one time with some other 1/12 at some other time. Top 1% discussions seem to be about 99% narrative, 1% data.

R Richard Schweitzer writes:


Go back and read Kuznetz's address when he became President of the American Economic Association in 1994; published in 1995, still available as a PDF.

Do we expect to understand history through a series of pictures: Kodak, "Brownie," Polaroid and now "Smartphones?"

Snapshot, "framed" economic graphics tell us what?

Michael Moran writes:

I find the income inequality debate using the tax data interesting. As a tax lawyer, does anyone know of studies that adjusted "share of income" data for the passive loss rules, the repeal of general utilities doctrine, changes to fringe benefit rules or restrictions on deductibility of semi personal expanses? In a similar vein, the immigration since 80 and greater movement of individuals in and out of the top !% should be examined.
Has anyone done this?
If they have not any such studies are worthless.

ColoComment writes:

pointer from cafehayek.com, an interesting look at income sources. Check out the info for pensions.


Tom Sparks writes:

greater movement of individuals in and out of the top 1% should be examined.
Has anyone done this?
If they have not any such studies are worthless.

I think this is the KEY point. These are generally not the same people. I was in the top 1% in 2009, but lost a lot of money in 2007-08, so obviously was not in that group.

Also, how does income distribution data handle losses? Is there a "sixth quintile" for losses? (grin)

Also, how did income distribution change after Bill Clinton changed the deductibility of "excessive" compensation? And he carved out an exemption for his Hollywood buddies for "performance-based" compensation and birthed the outrageous practice of excessive CEO option granting?

Mostly tho, I think the data need to be "risk-adjusted" to reflect that it's not the same people year after year. Perhaps a running 5 or 10 yr average income number might capture most of this? Noting what percentage of people in this year's Top 1% were also in last year's, or any of the previous 5 or 10 year's Top 1% would also help to shed some light on this.

Michael Moran writes:

Tom, while movement in and out of top 1% is important, and has increased (in part due to ZIRP), the biggest difference is the passive loss rules. Top 1% may have had the same "income in 1985, but in 1990, after the passive loss rules were enacted and phased in, his reported income may have been much higher.
This was alluded to by Mark V.
The 86 reform act which lower the top rate from 50% to 28% was tested and was tax neutral across the various income groups. The only way to raise as much money from top 1% at a 28% rate as with a 50% rate is increase income dramatically, which the passive loss rules did.
Rich people did not pay more in taxes pre 86, they just had tax shelters available to them which artificially depressed their income.

Comments for this entry have been closed
Return to top