
It's a good thing they didn't choose 95%
Everywhere you look nowadays, someone’s talking about the problem with income distribution in the U.S. There’s a sense that the “rich are getting richer and the poor are getting poorer.” The rally cry of the protesters on Wall St. is “We are the 99%!”
Nowhere is anyone really challenging the assumptions behind those statistics.
I got all of the following numbers from Emmanuel Saez’s compilation of income statistics up to 2007, found here. I suspect that a lot of people are getting their talking points from these data, but I can’t be sure because hardly anyone cites sources.
Facts: the richest 0.5 percent of Americans (with incomes of more than $632,000 per year) earned 19.3 percent of all income in 2007. The richest 0.1 percent (earning more than $2 million per year) earned 12.3 percent of all income in 2007.
Facts behind the facts:
1) Income statistics invariably aggregate people each year based on tax returns, but they don’t follow the same individuals. For example, it makes no difference to these statistics if the top 0.1 percent of income earners were the same people in 2007 as they were in 2006, 2005, etc.
2) The statistics go back to 1913 in this case. Now, we can rightfully assume that the top 0.1 percent of earners in 1913 are not the same as the top earners in 2007. Those 1913 earners are almost definitely dead. The completely human act of getting old and dying causes a problem for these types of statistics. We can assume that people earn more as they get older, and that someone just beginning their career will probably graduate in percentile as they gain more experience. The mere fact that people are traveling among the percentiles from year to year doesn’t bode well for the idea that there is a “class” of rich people that are getting more rich.
3) The statistics most often cited include wages, bonuses, exercised stock options and capital gains. In the fine print of Mr. Saez’s compilation of statistics, he says “because capital gains are such a volatile component and tend to be realized in a lumpy way, we focus primarily on series excluding capital gains. Series including capital gains are useful to assess the sensitivity of the results to including capital gains.” That hasn’t stopped people from using the capital gains data- more on this below.
4) For the purposes of those statistics, income is defined as market income before government transfers. That means before taxes and before any redistribution already on the books. Welfare, unemployment, Social Security- don’t get counted. Payroll, income, capital gains, property, estate taxes- don’t get counted.
What are the implications of the facts behind the facts?
First, that people earn more as they get older accounts for a great portion of the income distribution curve. The following example comes from Thomas Sowell’s The Vision of the Anointed:
Imagine a perfectly evenly distributed income system, where each person earned $10,000 a year from age 20 to age 30, and then got a $10,000 a year raise when they turned 30, and so on every 10 years until age 70, when he retires and makes no income. Each year this person spends $5,000 a year on subsistence and saves 10% of the remaining income for the future. When such a person turns 30, they will be making $20,000 a year and saving $1,500 a year. Their lifetime savings will be $5,000. When he turns 50, he will make $40,000 and have saved $45,000 over his lifetime. At age 70 he makes no income but has saved $125,000 over his lifetime.
This system seems to be perfectly fair. Every single person is on the same pay scale, they only get paid more for being more experienced. And yet, at any given moment, the following is true in this hypothetical society:
1) The top 17 percent of income earners earn five times more than the bottom 17 percent
2) The total savings of the top 17 percent of savers is 25 times that of the savings of the bottom 17%.
3) 17% percent of people control 45% of the wealth.
Each of those statistics is technically true, but together they are misleading. The top 17 percent I reference in each statistic aren’t necessarily the same people mentioned in the previous statistic- it matters if I’m saying income, savings or wealth. It’s also misleading that I’m talking about a specific moment in time. It doesn’t account for the fact that the bottom 17% of earners are not being treated unfairly, because they can expect to earn more as they get older.
Second important implication that might change your assumptions:
Counting capital gains, as Mr. Saez admits in the fine print, is tricky business. Capital gains that register on a large scale tend to be infrequent, like exercising stock options upon retirement, or selling an investment property. For wealthy people that make large sums of money this way, the vast majority can only expect to have a really large payout once in their lifetime. Another hypothetical situation: Imagine someone who has worked his whole life for a small business, accruing stock options and getting raises the whole time until he is an executive vice president. When he retires he makes around $100,000 a year. He also sells his stock options- his nest egg- and rakes in a bonanza of $2.2 million. For that year, he is in the 0.1 percentile of wealthiest Americans. It makes no difference to the statistics that he only earned that much income one time, and that for clarity of analysis it would probably be better to spread the stock option income over his entire working career and then adjust for inflation. The next year, he makes nothing except some income off of investments that he purchased with his $2.2 million. His neighbor down the street, however, is retiring this year and sells his restaurant for $2.2 million.
To the statistics, the man and his neighbor might as well be the same person. They are the 0.1 percentile of the richest people in the country, the richest of the rich.
Final important implication:
Income is counted before any government transfers. The curve would flatten somewhat once you factor in the already very progressive tax scheme and the benefits received by people in the lower percentiles, but of course it would still be a top heavy graph (it’s inevitable because people earn more as they get older).
More importantly, any further redistribution of income would have absolutely no effect on the curve, unless it simply discouraged rich people from exercising capital gains and making money. The stated purpose of those that favor “taxing the rich” is to make the distribution less top heavy, but more taxes wouldn’t even be counted. In fact, more redistribution would provide an incentive for lower income brackets not to make more money, which would have the opposite effect on the distribution.
To put it bluntly, the only way to fix the curve is to make everyone earn less money. Fair’s fair!
Some other stats that you aren’t likely to see on a cardboard sign in lower Manhattan:
1) If you use estate tax returns and the estate multiplier method, the wealth share of the top 0.1% of Americans controlled approximately 9% of the nation’s wealth in the year 2000. Sound like a lot? It was 12% in 1965, and 23% in 1930.
2) Once you exclude capital gains, the total share of the top 0.5% in capital income and dividends has been decreasing steadily since 1916. In 1931, the top 0.5% received 60% of the nation’s capital income, and 40% of the nation’s dividend income. In 2001 it was 12% and 5%, respectively. It 2006 it went back up a bit to 20% and 10%. Excluding capital gains is appropriate, by the way, as explained above.
3) In 2000, the top 0.1% by income excluding capital gains earned 60% of their income in wages, 25% in entrepreneurial income, and 12% in capital income. In 2007, the top 0.1% earned 40% in wages, 35% in capital income, and 28% in entrepreneurial income. Over the past decade, the richest of the rich have tended to be more and more people who earned their money off of capital income and entrepreneurial income. In 2007, our richest people were more likely than not to be people who got rich off of risking their own capital in job creating activities. In 2000 they were more likely to be people with big paychecks. That trend seems like a step in the right direction.
4) When looking at “wage income,” which is salaries including bonuses and exercised stock options, the 90-95 percentile earned about 10.5% of the nation’s total. That percentage has stayed between 9 and 11 percent for 80 years, with the high being 10.9% in 1938. These people are the most likely to bear the brunt of any additional redistribution taxes, because they probably don’t make enough to take advantage of capital flight.
5) Excluding capital gains, the bottom 99% of earners earned on average $8,000 a year in 1933, $30,000 a year in 1963, and $45,000 a year in 2007. Yes, that’s adjusted for inflation. There isn’t a percentile where real incomes haven’t increased over the past century.
6) When you add up the income, payroll, corporate income, gasoline, alcoholic beverage, tobacco, diesel fuel, air transport, excise, customs, duties, estate and gift taxes, the top 20% of earners paid 70.20% of their income in federal taxes in 2007. The bottom 20% paid 38.9%. Source: The Tax Foundation.
7) In 2007, the top 10% by income paid 70% of all income taxes. The top 50% paid 97.3% of all income taxes. The bottom 50% paid 2.70% of income taxes. If you say “we are the 99%,” than at least you paid the majority of income taxes: 62%. Good thing the rallying cry isn’t “We are the 95%,” because they only paid 41% of all income taxes. Source: The Tax Foundation.
The point isn’t that rich people aren’t really rich. The point is that statistics should be taken with a grain of salt. If you’re looking through the numbers for something to confirm what you’ve already concluded, it isn’t hard to find that perfect stat that will look good on your cardboard sign.