Now let's see why there would still be enormous income inequality.
1. Suppose everyone had identical lifetime wage income and inherited no money. Also assume all investments were risk-free. Incomes would still vary greatly due to different preferences regarding current vs. future consumption (i.e. different saving preferences.)
2. Suppose everyone was exactly like me; in the poorest quintile from age 18-26, then the second, then the third, eventually the top, before falling back on retirement. Also assume these identical people were born in different years. Again, no economic inequality (lifetime) but large annual income inequality.
3. Imagine everyone was a farmer with similar land, and there were alternating good years and bad years, which varied by location. When farmers made money in the east they lost money in the west. Over any five year period income averaged out to be roughly identical, but each year farmers made large incomes in some areas, and losses in others. No economic inequality in a long run sense, but plenty of annual income inequality.
4. Let's say everyone made an identical unrealized capital gain of $5000 each year, but they differed as to when they actually sold their asset and realized the capital gain. Again, no economic inequality but lots of measured income inequality, as only realized capital gains go into the income data.
5. Let's say some people went to school much longer than others. They earned a compensating differential in income to reward them for the disutility of their extra studying. Again, no economic inequality in a utility sense but lots of income inequality.
6. Let's assume some people make more money, but only because they live in an areas with a higher cost of living, perhaps due to higher heating bills in the winter. Again, no economic inequality but lots of income inequality.
7. Let's assume that preferences differ, and that some people are less materialistic. They earn less money because they prefer to maintain more spiritual values. Here I'm thinking of groups like the Amish.
8. The underground economy.
Is all this nitpicking, or are these distortions important? The first item suggests that we really ought to be looking at the present value of wage income, or the present value of consumption. In that case capital income would not count. I don't think I need to convince anyone that capital income hugely distorts the numbers of the top one percent.
Item two suggests age is an important factor, which might distort our results. The second poorest town in America is Athens, Ohio, which is full of middle class college students with low incomes. Are they really poor? That's a judgment call. But I'd guess that their lifetime incomes are quite respectable.
The first poorest town in America is almost entirely composed of Hasidic Jews, which (according to the NYT) are choosing to live a rather non-materialistic lifestyle. That's item 7.
And yes, the fact that we measure realized capital gains rather than unrealized cap gains makes income look much more unequal than it is. Bob Murphy asks me how I can hate income and love NGDP (national income.) Do capital gains go into NGDP?
Many (liberals) object that this exercise misses the point. BUT WHAT ABOUT THE BANKSTERS!!??!! They say even if income data has problems, there is no better alternative, and in any case inequality is clearly getting much worse by any reasonable measure. I don't accept that view. We do have a better measure---consumption. But even consumption is not perfect because of the lifecycle problem I discussed. It will still overstate economic inequality.
Nonetheless, if we use consumption data then the change in inequality is much less clear. Some studies show increased consumption inequality, some don't. FWIW, in my view the consumption gap between the middle class and rich has widened over my lifetime, and the consumption gap between the middle class and poor has narrowed. Since I care much more about poverty than the middle class/rich gap, I'm not particularly dismayed by the changes we've seen in recent decades.
Some liberals claim it's about political power, which correlates with income more than consumption. I doubt it. Public policy is less pro-rich than when Reagan was president (for instance MTRs on the rich have risen sharply since 1987), and yet income is much less equal than when Reagan was President. I see no evidence that income inequality correlates with more political power for the rich, as compared to consumption inequality. (That's not to deny that the rich have more political power, they do.)
However I do favor public policies that would lead to greater economic equality (for utilitarian reasons):
1. Wage subsidies for low wage workers.
2. Less barriers to entry (less zoning, weaker IP restrictions, no occupational licensing, etc.)
Normally when inequality widens it should be self-correcting. The relatively poor will see a bigger gap, and work harder to boost their standing in society. Thus if the gap between the consumption bundle of college grads and non-college grads gets wider, then more people will get college degrees. This may not work for two reasons:
1. Barriers to entry.
2. Talent gaps.
You may have noticed that one of my policy suggestions addresses barriers to entry, and the other addresses talent gaps. And here's the final irony. A wage subsidy that reduces economic inequality actually raises income inequality, at least if you measure income net of the subsidy (which is how we do it.) If you are convinced that I am wrong and want to hold on to the income inequality data, then you ought to oppose my wage subsidy proposal---it will make inequality "worse."
PS. Josh Barro has an excellent post on a supply-side agenda for the 21st century. The emphasis is on lots of deregulation and cuts in implicit marginal tax rates. However I found this remark unfortunate:
We're going to need a new supply side economics that encourages people to work, invest and innovate.
That doesn't mean conservatives' agenda of tax cuts and federal deregulation, which aims at problems that were addressed in the 1970s or never existed at all.
Barro will find very little support for his agenda among politicians of either party. However he will find lots of support among both liberal and conservative intellectuals.
PPS. Whenever I do posts like this I get commenters saying "Sumner's trying to deny America has lots of inequality," or "Sumner's saying the typical poor person in America is Hasidic Jew." I'm used to it by now.