I’m never going to win a Nobel Prize. Maybe in literature. I don’t know why Joseph Stiglitz’s new Vanity Fair piece on inequality is so off-base. But it is. And it’s incredibly frustrating (1) to see someone so intelligent be thwarted by ideology and (2) to watch as his views are propagated on the basis of his name recognition.
What’s a lonely uninvited-to-Davos blogger to do? Blog. Herewith, my fact check of the VF article. Stiglitz writes (henceforth in italics),
The upper 1 percent of Americans are now taking in nearly a quarter of the nation’s income every year. In terms of wealth rather than income, the top 1 percent control 40 percent. Their lot in life has improved considerably. Twenty-five years ago, the corresponding figures were 12 percent and 33 percent.
Stiglitz doesn’t cite any of his figures (possibly a limitation of the outlet), but the Piketty & Saez estimate of the top one percent’s income share in the most recent year (2008) was 18 percent, which is just a hair closer to “nearly a quarter” than it is to “just over a tenth”. Their data says that share was 9 percent in 1985, but that should be adjusted upwards to 13 percent. Similarly, CBO says the top one percent’s share was 17 percent in 2007 for after-tax income, up from 11 percent in 1989. Saez’s estimate of the top one percent’s share of wealth is 21 percent for 2000, 21 percent for 1990, and 22 percent for 1985. Edward Wolff’s is 35 percent for 2007, up from 34 in 1983 (which I doubt is statistically different from 35 in this case). The top appears to have experienced income and wealth losses from 2007 to 2009 while the bottom experienced gains. Taken together, the top one percent’s income share rose from 11-13 percent twenty-five years ago to 17-18 percent according to the most recent data. The top one percent’s wealth share basically hasn’t risen.
[UPDATE: See MIT economist Erik Brynjolfsson's comment below, which led me to add this paragraph. Brynjolfsson raises an important point (though I wouldn't call it a mistake) in noting that Stiglitz may have been referring to the Piketty and Saez numbers that include realized capital gains in "income". I chose the series excluding capital gains because the timing of when capital gains are realized has everything to do with tax law, the strength of the economy, and when people retire. The P&S series including capital gains still doesn't account for all the unrealized gains accruing to people (most importantly, those accruing to people in their retirement accounts). Capital gains realization is "lumpy" in a way that makes trends problematic.
But I will concede that the level of the top's income share (including realized capital gains) is closer to 25 percent than the P&S numbers I cite above suggest. Now whether their share of income including unrealized capital gains is closer to 25 percent or 17 or 18 percent is an open question. And I still say the series excluding capital gains is the way to go for trend estimation. But look, all this aside, the CBO series includes realized capital gains (but also considers taxes and other things the P&S series leaves out). And it shows the same basic trend and level as my conclusion above.]
While the top 1 percent have seen their incomes rise 18 percent over the past decade, those in the middle have actually seen their incomes fall. For men with only high-school degrees, the decline has been precipitous—12 percent in the last quarter-century alone.
The 18 percent figure looks to be from Piketty and Saez (the change from 1998 to 2008). The claim about median incomes falling is incorrect if one takes into account the value of employer- and government-provided health insurance. (Majorities of workers with employer coverage say they prefer more generous coverage to higher wages, so it turns out employers aren’t crazy in substituting ever-more-costly insurance for wages over time.) The decline in earnings (not income) for men with just a high school diploma is probably less than 12 percent. Based on some analyses I’ve been working on using the Current Population Survey, I find that men with a high school diploma but no four-year college degree saw a 12 percent decline in earnings over the roughly 33-year period from 1971-73 to 2003-2007, but that doesn’t take into account the caveats I mention in this post. And earnings among women with the same level of education rose by over 50 percent, so that’s inconvenient for Stiglitz.
The change in household or family income among men with just a high school diploma was, I’d wager, positive even before factoring in the caveats. And while I can’t cite the paper yet, research I’ve seen using the PSID rejects the conclusion that wives have been forced to work more due to stagnant husband earnings—the biggest increases in work were among wives with the best-educated husbands, and while the hours of married men declined, those of single men did not (suggesting that the decline among married men was a reaction to increased work among their wives). I’ll update this post when I can cite the paper (though that won’t be for a couple months anyway). But think about it–did all these women increase their college-going simply in anticipation of marrying men with stagnant earnings, or did they prefer the fulfilling professional options that a college degree afforded them? Or consider–is declining fertility, delayed marriage, and increased college-going among women in developed countries around the world all somehow related to rising American inequality? You can get the basic trend on work by sex by marital status from Table 1 of this paper while you anxiously await my update.
All the growth in recent decades—and more—has gone to those at the top.
Nope, not if “the top” refers to “the top 1 percent” cited two sentences earlier. According to the Piketty and Saez data, depending on whether one uses the share of nominal or real (inflation-adjusted) gains and whether one includes or excludes capital gains in “income”, the share of income growth going to the top one percent from 1998 to 2008 was between 22 and 33 percent. If you go back to 1988, the range is from 19 to 32 percent of gains since then. And keep in mind that when you start from an unequal distribution, if everyone experiences the same rate of income growth, a disproportionate share of gains will go to the top.
In terms of income inequality, America lags behind any country in the old, ossified Europe that President George W. Bush used to deride. Among our closest counterparts are Russia with its oligarchs and Iran.
Compared to nearly all of the major nations of western and central Europe, the U.S. does have higher inequality (but it may not be that far off from the U.K. or Canada). The only numbers I could find for Russia and Iran are from the CIA World Factbook (the quality of which I can’t speak to). Out of 136 countries, the U.S. is ranked 40th worst. Iran is ranked 43rd and Russia 52nd. So that sounds bad, right? Meh. Hong Kong and Singapore rank worse than the U.S., and Indonesia, India, and Ethiopia rank much better than Russia. Stiglitz will have to do better than this if he wants to argue that American inequality is a big deal.
First, growing inequality is the flip side of something else; shrinking opportunity….Second, many of the distortions that lead to inequality—such as those associated with monopoly power and preferential tax treatment for special interests—undermine the efficiency of the economy.
OK, so now Stiglitz is trying to tell us why we should care about the inequality that he exaggerates. But these are just assertions. The best evidence suggests that opportunity for men to move from the bottom to the top over the course of a career hasn’t changed much over the past 35 to 40 years, and it has unambiguously increased for women (see Figures 15A and 15B). Across generations, the evidence is extremely thin, but it doesn’t point to an unambiguous increase or decrease in opportunity over the past few decades. As for inequality and efficiency, my dissertation advisor, Christopher Jencks, has found that there is little correlation between economic growth and inequality levels, which doesn’t exactly help those who believe inequality promotes growth but is equally problematic for Stiglitz and others who believe that inequality is inefficient.
When you look at the sheer volume of wealth controlled by the top 1 percent in this country, it’s tempting to see our growing inequality as a quintessentially American achievement…
Here Stiglitz is conflating income inequality (growing) with wealth inequality (basically flat and at a historic low in the U.S.). Whatevs.
America’s inequality distorts out society in every conceivable way. There is, for one thing, a well-documented lifestyle effect—people outside the top 1 percent increasingly live beyond their means.
So document it! The share of families with any debt rose from 72 percent in 1989 to 77 percent in 2007, though note that the share with assets also grew. Median net worth (assets minus debt) rose from $75,500 to $120,600. In the wake of the housing bust, it fell, but it was still around $92,000 in 2009. Among people with debt, median debt payments rose from 15.3 percent of family income in 1989 to 18.6 in 2007. These are pretty small changes in indebtedness, and I’m not sure how Stiglitz could empirically link them to inequality.
Inequality massively distorts our foreign policy.
Ummm…going for the Peace Prize next?
[T]he chances of a poor citizen, or even a middle-class citizen, making it to the top in America are smaller than in many countries of Europe.
What little evidence there is suggests that upward mobility is lower in the U.S. only for men and only for those who start out poor. [UPDATE: Just to clarify, I'm talking about only men who start out poor, not men plus all people who start out poor. See the linked paper for details, but we're talking about 12 to 13 percent of the population, roughly.]
All of this is having the predictable effect of creating alienation—voter turnout among those in their 20s in the last election stood at 21 percent, comparable to the unemployment rate.
Oh boy, the shift to political science by economist pundits is always fraught with danger. The 2010 election is a single data point (and an off-year election, when voting rates are much lower). I’ll just quote from a fact sheet from a Tufts research center that studies civic engagement among youth: “The 2008 election marked the third highest turnout rate among young people since the voting age was lowered to 18.” What any of this has to do with inequality is anybody’s guess.
In recent weeks we have watched people taking to the streets by the millions to protest political, economic, and social conditions in the oppressive societies that they inhabit….The ruling families elsewhere in the region look on nervously from their air-conditioned penthouses—will they be next?…As we gaze out at the popular fervor in the streets, one question to ask ourselves it this: When will it come to America?
My guess is never. By the way, Joe, be honest–were you using a pseudonym here?
Originally published at The Empiricist Strikes Back.