The reasons that have been given for changes in inequality in recent decades can be placed into four
categories: technology, globalization, sociology, and public policy. Much of
the debate over inequality has revolved around which of these is the most important factor.
Technology, in particular skill-based technical change, is often cited but this
explanation is hard to reconcile with some aspects of growing inequality,
particularly the fact that much of the growth has been at the very top of the
income distribution rather than more generally across educated workers.
In addition, the actual evidence that skill-based technical change explains the majority of the change inequality is based largely upon examination of the Solow residual - what's left over after we take away everything we know about - and this residual term will contain any hard to quantify variables relating to political and institutional change, not just the technical change we are hoping to measure. So it's entirely possible that previous estimates confound technical change with other factors on the list of explanations.
The point is that there remains a great deal of uncertainty about why inequality has been rising in recent decades. In the following, Robert Samuelson describes work by Frank Levy and Peter Temin of MIT. They focus on the sociology and public policy explanations of rising inequality, in particular, shifts in social norms, business practices, and government policy as reflected in the interaction between the "United Auto Workers and the Big Three U.S. automakers":
The Equality Quagmire, by Robert J. Samuelson, Commentary, Washington Post: You have almost certainly never heard of the Treaty of Detroit... The Treaty of Detroit is a ... label describing a series of landmark labor agreements between the United Auto Workers and the Big Three U.S. automakers. Starting with a 1948 contract at General Motors, the agreements guaranteed annual wage increases, job security and generous fringe benefits. As Detroit's present turmoil attests, the treaty is in tatters.
Now come economists Frank Levy and Peter Temin of the Massachusetts Institute of Technology, who resurrect the label and say it explains something much greater: the rise of economic inequality. .. . Until now, most economists have blamed the growing pay gap on skill differences caused by the explosion of computer technologies. Levy and Temin contend ... that this is too simple; they also blame a shift in social norms and business practices.
First, some historical background. In late 1945 President Truman summoned 36 business, union and government leaders to a conference. The aim: to forge an understanding between labor and capital... It failed; the postwar era began badly. In 1946 there were 4,985 strikes, involving 4.6 million workers (11 percent of all workers). Autoworkers, railroad workers, steelworkers all struck.
The Treaty of Detroit fashioned a crude truce that spread elsewhere. Between major contracts, the automakers got labor peace. In return workers got job security (reminder: in the 1930s, unemployment averaged 18 percent) and higher incomes. The treaty influenced other unionized industries, where ... similar contracts ... became common. Many nonunion companies embraced comparable norms: Workers should receive wage gains beyond inflation, job security and good fringe benefits.
The result, say Levy and Temin, was that "market outcomes" in pay were "strongly moderated by institutional factors" -- business practices shaped by social values and government policies. After World War II, many executives strove to refurbish the image of Big Business, badly battered in the Depression. ...[But]... By the 1980s this generation of business leaders had mostly retired. Companies increasingly paid what the market would bear or they could afford.
Pay gains diverged. In early postwar decades, compensation increases crudely paralleled productivity gains -- improvements in efficiency. From 1950 to 1973, productivity rose 97 percent. Over the same period, median compensation of male high school graduates aged 35-44 rose 95 percent (after inflation); for college graduates 35-44, the increase was 106 percent. Those in the top one-half of 1 percent received only a 37 percent gain. From 1980 to 2005, productivity increased 71 percent. Median compensation for high school graduates dropped 4 percent, and compensation for college graduates rose only 24 percent. For those in the top one-half of 1 percent, it jumped 89 percent.
Samuelson has already cited the retirement of image-conscious executives as one reason why things changed around 1980 and inequality began rising, and he now goes on to expand the list:
Comparisons such as these evoke images of greedy CEOs and hedge fund managers. But the story is more complicated. On the whole, the economy that produces these growing inequalities outperforms the one that created more statistical equality.
I need to break in. There's no evidence for the claim that the U.S. would have grown slower after 1980 if inequality had, say, remained stable instead of expanding.
The norms and practices highlighted by Levy and Temin collapsed mainly because they no longer worked. The idea that everyone's wages should reflect inflation plus a few percentage points worsened both inflation and stability. There were four recessions between 1969 and 1981; by then, inflation was 10 percent and mortgage rates 15 percent. Productivity growth had plunged.
Hang on again. He's blaming both inflation and instability (business cycles) on the arrangements between labor and big business.
As for inflation, in general, if wage growth equals inflation plus productivity growth, there is no inflation pressure. So wage growth in excess of inflation is not necessarily inflationary. But more directly, the source of inflation during that time period was bad monetary policy, not labor agreements.
And I don't know what evidence he has in mind when he says that the
labor agreements are the source of instability in the economy. During
the 1960s we had a long period of sustained growth, so that doesn't
fit. In addition, the change in the variance of output in 1984, i.e.
the Great Moderation, is much more sudden than
the kinds of changes he is talking about. The decline in the variance of output is puzzling abrupt, while the social, institutional, and public policy changes are much more gradual (he's also pushing the
evidence a bit to make his case that there have been "four recessions
between 1969 and 1981," the fourth recession doesn't begin until late in 1981
and runs through 1982). I am not saying that the collapse of unions along with other social, institutional, and public policy changes have
nothing to do with the inequality growth we've experienced in recent decades, just that
the changes didn't occur for the reasons
He goes on:
Greater competition -- from imports, deregulation, new technologies -- also doomed pattern wage-setting. Companies with lax pay practices lost sales and profits. Consider GM, Ford and Chrysler as Exhibit A. ...
Does exhibit A show the crappy cars that well-compensated management decided to build? I sometimes wonder if globalization is an excuse as much as a cause. Finally:
In 2008, economic inequality could become a political flash point, because the income gains at the top seem so outsize and gains elsewhere are so choppy. The very uncertainty means that, even amid great prosperity, Americans feel anxious. Whether the debate becomes an empty exercise in class warfare or a genuine search for ways to reconcile economic justice and economic growth is an open question.
That the question is open is a step forward. But finding a way to share the gains from economic growth more equitably won't be easy, and confusing the issue by claiming that more equitable distributions will lower economic growth, cause more frequent recessions, and be inflationary when there's no evidence to support those claims doesn't help at all.