The Case for Worrying About Inequality is Still Overstated
Ezra Klein is in trouble. In recent months, Klein has emerged as one of today’s most intellectually honest and independent-minded liberals, and last week he concluded that the evidence fails to support President Obama’s claim that inequality is “the defining challenge of our time”
. Klein was not even saying that the evidence provides little reason to prioritize inequality at all—a case I have made in the pages of National Affairs. But it was enough to draw the ire of the Inequality Patrol on the left, led by Paul Krugman writing in today’s Times.
Krugman wants to believe that inequality hurt economic growth in the aughts and has dampened income growth among the poor and middle class in recent decades, and he wants you to believe it too. He claims, “inequality is rising so fast that over the past six years it has been as big a drag on ordinary American incomes as poor economic performance, even though those years include the worst economic slump since the 1930s.” If you want his evidence for this, see his blog post from over the weekend, in which he calculates that income below the top ten percent is “about 8 percent lower than it would have been if inequality had remained stable” since 2000.
Krugman is working with a spreadsheet from economist Emmanuel Saez’s website, and though he says his calculations exclude capital gains, they really include them (it’s just people are ranked by their income excluding the gains). If he’d consistently included gains (so that people are ranked using the income measure with gains), he’d have found the income of the bottom 90 percent was 5 percent lower because of inequality. Even accepting Krugman’s implicit assumption that the size of the pie would not have shrunk if inequality had been prevented from growing, this is some fragile stuff. He reports 2012 income if the bottom 90 percent’s income share hadn’t fallen from its 2000 level. But 2010 income if the income share hadn’t fallen from its 2000 level would have been just 0.8 percent lower than it was. That’s because the top ten percent’s share was essentially 52 percent in both years.
The Piketty-Saez data excludes most cash and noncash transfers and doesn’t take taxes into account. In other words, it ignores the very tools we use to mitigate inequality. The Congressional Budget Office recently updated its own inequality figures to 2010. It finds that after redistribution, the bottom 90 percent’s share rose from 64 percent in 2000 to 66 percent in 2010. The bottom 60 percent’s share rose too, as did the bottom 20 percent’s. So holding post-distribution inequality at its 2000 level, incomes in 2010 would have been lower than they were.
In a follow-up post, Krugman extends his argument back to earlier decades. He compares the income share of the bottom 80 percent in the CBO data in 1979 and 2007 and concludes that income growth for this group was slower by 0.7 percentage points per year because of rising inequality. This is in the ballpark whether one uses pre- or post-tax income. The questions for evaluating whether this computation ought to worry us are then (1) whether in the absence of rising inequality growth would have been as strong and (2) whether the income gains that mostly accrued to the top should have gone to everyone else to a greater extent than they did.
The first question is empirical. If higher inequality led to more growth, then we cannot assume with Krugman that in the absence of rising inequality, aggregate household income would have increased as much as his calculations indicate. The bottom 80 percent would have received a constant share of household income, but household income would not have risen as much as it did, so Krugman’s calculation would be off.
If preventing the 16 percent decline in the bottom 80 percent’s share of household income would have ended up shaving 16 percent off of aggregate household income growth, then holding the share constant would not have benefitted the poor and middle class at all. If preventing inequality from rising would have shaved 10 percent off of aggregate household income growth, then because of rising inequality the bottom 80 percent had an average income of just $49,800 instead of $53,400. And depending on the strategy used to thwart rising inequality, that extra money would not necessarily have gone to the poor or even the middle fifth of households, as opposed to, say, the second-richest fifth of households.
The Inequality Patrol thinks it is obvious that, say, median income growth would have been higher in the absence of rising inequality. But Lane Kenworthy, another straight-shooting liberal, finds that across industrialized countries, the impact of rising inequality on median household incomes has been small and possibly nonexistent. For that matter, median household income growth was slower in the 1970s than it was in the 1980s and 1990s (and maybe than in the 2000s—much depends on how Medicaid and Medicare are valued as income). This is relevant because the top one percent did not begin pulling away from everyone else until the late 1970s. It is worth repeating: the slowdown in median income growth (and income growth at the bottom) predates the rise in the top one percent’s income share.
(It is also worth saying again that there is a chance that the Piketty-Saez data have greatly overstated the rise in inequality. I give it even odds that this research will hold up.)
Even if it is true that in the absence of rising inequality, more income would have gone to the poor and middle class than actually accrued to them, that need not mean that we should object to rising inequality. We all have our own ideas about who deserves to make what. Again, Krugman and the Inequality Patrol think it’s obvious that the top doesn’t deserve what they’re making. Median earnings have not kept up with productivity growth in recent decades. The rich are siphoning off what is not theirs.
Maybe, but perhaps all the wage-productivity gap is signaling is the death of a patriarchal regime in which male workers were overpaid relative to their productivity because we were intent on keeping women in the home. For decades earlier in the twentieth century, men actively fought to shelter themselves from female competition—pushing for a minimum wage for women in order to limit the extent to which female workers could pull the union wage down to a market-clearing one, demanding that union contracts require women workers to be let go if they got married, promoting the famous five-dollar-day in Ford factories (only for men).*
Don’t get me wrong, the erosion of this regime has been a kick in the gut for less-skilled men, and we have done too little to help them adjust. It has been a boon for married women though, whose labor force participation rose steadily for decades from the 1940s (before men’s wages started stagnating) through the 1990s. The problem is not a general one of the rich appropriating the workers’ pay. Instead, it is possible that top earners and shareholders would have had something like 1929- or 2013-sized incomes during the Golden Age of the American economy—except that they, too, bought into the male breadwinner family-wage regime even at the expense of their own financial self-interest. In fact, if you start in the mid-1930s at the time of the Wagner Act’s passage that spurred union growth, median compensation and productivity have increased by roughly the same amount—it’s just that an initial period of too-rapid wage growth was followed by a long period of wage adjustment. Labor’s share of income has dropped at the same time, and the share of income going to the top one percent has risen.
If Krugman wants to argue that the rich don’t deserve to make as much as they do, he should just argue that. It’s a political winner, for sure. But the empirical case that rising inequality is the challenge of our time is less than compelling, no matter how hard the Inequality Patrol wishes it weren’t so.