Home Mail Articles Stats/current Supplements Subscriptions Links

The following article appeared in Left Business Observer #78, July 1997. It retains its copyright and may not be reprinted or redistributed in any form - print, electronic, facsimile, anything - without the permission of LBO.

Measuring privilege

So it looks like wealth didn't get that much more concentrated in the 1980s after all; we had to wait until the 1990s for that to happen.

One of the wonders of modern social science is the Survey of Consumer Finances (SCF), conducted by the Federal Reserve in cooperation with the IRS. While the U.S. Census Bureau collects and publishes plenty of information on income, and even some on wealth, its money questions are usually a smallish part of a broader population survey, and they also miss lots of action at the upper end. The SCF, though, is based on 90-minute interviews, and the surveyors pay special attention to the rich. It's probably the best picture of household finances done anywhere in the world.

The survey has been taken every three years since 1983, though 1986's was fairly rudimentary. Results from the 1983 edition were reported in rich detail in the Federal Reserve Bulletin, the central bank's flagship periodical. Of three articles on the SCF, one was devoted to the very richest. Since 1989, though, only an overview article has been published in the Bulletin, and data on distribution made available only in highly technical unpublished working papers.

Those technical papers are essential; assembling the raw data of the survey into a representation of 100 million households is intensely complex - especially if you're seriously interested in a reliable portrait of rich people. But it's not very user-friendly to present the wealth distribution numbers only as an appendix to a complex paper explaining how they were arrived at. As a public service, LBO presents the crucial figures at the below.

Who owns, who owes (percent of totals), 1995

bottom 90%

top 10%

top 10%, percentile breakdown
90­98.9 99­99.4 99.5­100
assets 37.9% 62.1 31 6.9 24.2
principal residence 66.4 33.6 25.7 2.7 5.2
other real estate 20.2 79.7 43.9 8.8 27.1
stocks 15.6 84.4 42.2 10.7 31.5
bonds 9.7 90.2 34.5 9.2 46.5
trusts 13.1 86.8 42.6 10.4 33.8
life insurance 55.0 45.0 27.8 6.1 11.0
checking accounts 57.7 42.3 26 4.6 11.6
thrift accounts 43.1 56.9 40.9 8.2 7.8
other accounts 37.9 62.1 35.2 7.0 19.8
business 7.7 92.2 20.8 11.5 59.9
autos 77.6 22.4 17.8 1.9 2.6
other 29.3 70.6 39.2 4.8 26.6
liabilities 70.9 29.1 19.3 3.0 6.7
principal residence (mortgage) 78.4 21.6 17.2 1.8 2.6
other real estate 25.2 74.7 41 10.2 23.5
other liabilities 80.6 19.4 9.6 1.5 8.3
net worth 31.5 68.4 33.2 7.6 27.5
nonresidential 22.5 77.5 34.2 9.1 34.2
income 68.9 31.1 19.6 3.4 8.1
dollar values
income $33,273 134,933 94,690 293,565 700,677
average net worth 39,252 1,217,375 596,974 2,844,934 10,757,046

These distributions are based on sorting households by their total net worth. Columns show share of each category held byeach net worth grouping. The first two columns are the bottom 90% and the top 10%; the next three break down the top 10%, with (reading from right to left) the richest 1/2% in the last column, the next 1/2% in the fourth column, and the next 9% in the third. The last two rows show the dollar values for income and net worth. For example, the richest 10% of the population has 68.4% of net worth and 31.1% of income; their average income was $134,933, and their net worth, $1,217,375. Rows are mostly self-explanatory; nonresidential net worth excludes the equity in the principal residence (that is, its value less the mortgage on it). Source: Arthur B. Kennickell and R. Louise Woodburn, "Consistent Weight Design for the 1989, 1992, and 1995 SCFs, and the Distribution of Wealth," unpublished Federal Reserve technical paper, June 23, 1997. (Click here for the paper, and here for its appendix; these are big files, especially the appendix - 310k and 4,700k respectively.)

Wealth is concentrated far more densely than income. The richest 0.5% of the population claimed 8% of income in 1995 - but 28% of net worth, almost as much as the bottom 90% of the population (32%). (Other, earlier studies show that the wealth of the bottom 90% is almost entirely accounted for by its upper third; the bottom 40% of the population has just 1% of total wealth.) If you strip out the principal residence, the major repository of middle-class wealth (but one not easily redeployable in the capital markets: most people won't sell the house to buy stocks), the top 0.5% pulls well ahead of the bottom 90%. The richest 10% of the population - about 10 million households - owned 84% of the stock and 90% of the bonds held by individuals (including that held indirectly through mutual funds). The democratization of ownership supposedly brought about by mutual funds has a long way to go.

One important disclosure in the technical part of the paper - written by Arthur Kennickell, of the Fed, and Louise Woodburn, formerly of the IRS and now with Ernst and Young - is that revisions to the 1989 survey now show no substantial increase in the concentration of wealth in the 1980s (see graph nearby). But the 1990s have brought concentration to record levels, with the share of the top 1% rising by nearly five percentage points. That gain came mostly at the expense of the next 9%; the share of the bottom 90% continued a long erosion. It should be said, though, that while the 1989, 1992, and 1995 surveys are very similar, the earlier ones are different enough that comparisons should be made skeptically.

In real dollar terms, while income and net worth fell for virtually every group between 1989 and 1992, between 1992 and 1995 the rich flourished. Incomes of the bottom 90% rose 0.9% after inflation (total, not per year) - but those of the top 0.5% were up 44%. Nonresidential net worth was up 1.5% for the bottom 90% - but 26% for the very richest. Still, for all income classes, household income had not yet regained 1989's peak by 1995.

There's one area in which the bottom 90% outdid the top 0.5%: debt. Liabilities for the bottom 90% rose 11%, while falling 19% for the richest. Assets were a virtual mirror image, with the masses' assets barely rising (+2%) while those of the richest rose ten times as much (22%). This complementarity is no surprise, since consumer and mortgage credit boil down to the very rich lending to the middle class and the poor.

Cite statistics like these and apologists will immediately respond that they don't include private pension accounts. This is true enough, and including them would lessen the distributional skew a bit, but there are good reasons for excluding them. Even in orthodox economics, pensions are seen as deferred wages - a smaller paycheck today in exchange for an income in retirement. Nonpension wealth, though, usually provides an income today, and can be sold and transformed into anything a rentier desires - a vacation house, a German government bond, a short position in the Thai baht, or even a real business. You can't do that with your IRA.

Home Mail Articles Stats/current Supplements Subscriptions Links