Making Sense of Inequality
From the August 13, 2012, issue of NR

Timothy Noah, author of The Great Divergence


Not long ago, a former colleague asked me to recommend the best accessible reference on income inequality. I immediately suggested Timothy Noah’s 2010 series of essays, “The Great Divergence,” written for his then-employer, Slate. Why, then, do I find Noah’s new book of the same name — an expansion of the award-winning series — so frustrating?

The book largely follows the outline of the series, with its clearly presented statistics on the rise in inequality and Noah’s impressive explication of what social scientists have found regarding its causes. The frustrations come, in part, from the additional space a book provides for Noah to wade into the perilous business of dot-connecting. But even the original series was irksome in the way it (faithfully) conveyed the conventional wisdom about the supposed connections between inequality, economic mobility, and living standards. Both the original series and the new book are fantastic distillations not only of a messy literature on inequality and its causes, but also of a seriously flawed conventional story about their consequences.

Noah is an engaging and informative writer, and the first chapter of the book shows off his skills (even as it reveals the ideological commitments the reader will have to be wary of). It provides an improbably captivating overview of what we know about inequality trends and how we have come to this knowledge. He intertwines this history with a broad account of how living standards and economic security have changed, and it is here that he falls into the traps of convention.

The “Golden Age” of the 1950s and 1960s followed a steep decline in inequality and preceded our modern period of rising inequality. In Noah’s view, “there probably was no better time to hold membership in America’s middle class.” This widely held — but wrong — view reveals the extent to which inequality has come to dominate the evaluative standards by which liberals judge the economy. In fact, the median American family is twice as rich today as it was in 1960, if one takes into account changes in family size, government and employer benefits, and rising immigration. That much of this improvement came before 1979 — even 1973 — hardly negates this central fact. Congressional Budget Office statistics indicate that the median household was 35 percent richer in 2007 than in 1979. The bottom fifth of households was about 20 percent richer.

Noah and other liberals have a series of standard (but flawed) arguments that minimize the extent of this improvement: A huge share of gains since 1980 went to the top 1 percent. Family-income growth hasn’t kept up with productivity increases. Income growth is due solely to wives’ supplementing their families’ incomes. Each of these claims is supported by naïve interpretations of readily available data, but a more careful examination belies them.

Let’s begin with the share-of-gains estimates. Noah plays up a finding that 80 percent of the increase in income from 1980 to 2005 went to the top 1 percent. This is a figure derived from the work of Thomas Piketty and Emmanuel Saez. But their updates in recent years have used a more appropriate adjustment for inflation, and their more recent estimate is not 80 percent but 58 percent.

Also, for a number of reasons, the share-of-gains figure is not as straightforward as it would seem. For example, this figure looks at income gains after adjusting the 1980 numbers to reflect 2005 purchasing power. But if we ask what share of income gains went to the top before taking account of the fact that 2005 incomes bought less for rich and poor alike because of inflation, the answer is that the top received just 28 percent of income gains.

Piketty and Saez garnered quite a bit of attention in their latest update, when they found that 93 percent of the gains from 2009 to 2010 went to the top 1 percent. But the computation they use also found that 49 percent of the gains from 2007 to 2009 went to the top, which is interesting, because the total income received by the top 1 percent fell over those two years — by, on average, half a million dollars. It is a strange statistic that indicates that a group commanded half of all income gains when the data on which it is based shows the group losing income.

As for the income-versus-productivity comparison, that one should be shelved for good, because researchers on both left and right have shown that, when properly analyzed, median family income actually tracks productivity very well. Noah cites a 2005 paper co-authored by economist Robert Gordon to back up his claim, but in 2009 Gordon wrote a widely cited paper showing that median income and productivity actually aligned closely from 1979 to 2007, with the former increasing by 1.50 percent annually and the latter by 1.66 percent. On the other hand, male earnings have lagged behind productivity growth over the past few decades. The simple explanation is that in earlier decades (during the peak union years), male-earnings growth outpaced productivity increases, and the past few decades have seen men’s earnings fall back to earth. (Women’s-earnings growth, meanwhile, has far exceeded productivity increases.)