Blue-Collar Blues

It seems fair to say that the United States is driven more by greed than envy. As measured by the employment cost index of the U.S. Bureau of Labor Statistics, private real wages and real compensation are up by just 2.1% and 5.6%, respectively.

Particularly striking has been the fact that the average wage growth of workers at all — but the very highest levels of skill and education has been equally poor.

For most blue-collar workers, the recent weak wage growth continued a longer-run trend of slow real wage increases which had been interrupted by the second half of the 1990s.

But for workers with a college education, the recent slow real wage growth is a relatively new experience because these workers had seen their real pay rise steadily between 1980 and 2000.

It is, therefore, not surprising that real wages and incomes have become a matter of concern.

Between 2000 and 2006, the big inequality story in the United States is the shift in income shares from labor to capital. Estimates suggest that had income gains between 2000 and 2006 been divided in proportion to shares in 2000, real compensation of blue-collar workers — which actually increased by 4.9% — would have been higher by another 5.2%.

Given that white-collar compensation increased at about the same pace as blue-collar compensation — by 5.4% — a roughly similar improvement would have occurred in white-collar compensation.

The aggregate data from the 2001-2006 period are certainly compatible with the view that global forces have induced a structural change in the income shares of labor and capital. But there are also reasons to be skeptical that this structural change reflects such forces.

First, while the U.S. economy has been globalizing for a long time — unlike many other countries — the share of labor compensation in national income in the United States has been relatively stable. After rising steadily between 1950 and the late 1960s, the share has basically fluctuated within a fairly narrow range of around 66%.

Globalization in the United States increased most dramatically over the 1970s — the years in which labor’s share in income increased significantly, and taking the period 1970 through 2006 as a whole, there is no evidence of a major trend in either direction.

It is only in the period since 2000 that labor’s share has declined. In 2006, the compensation share was relatively low, at 63.9%, about 2.2% below its peak share in 2001 and 1.6% below the long-run average but about the same share as in 1997.

Focusing more narrowly on the non-financial business sector provides a similar picture. In 2006 labor’s share of 73% of net value-added was considerably below its peak in 2001 (77.4%) — but above levels recorded in 1997 (71.5%), 1988 (72.8%) and 1984 (72.6%).

If global forces are responsible for the recent shift in income shares, the character of recent wage pressures has to be very different from that prevailing earlier

One source recently could be the India effect. While earlier trade pressures affected only unskilled workers, trade now puts downward pressures on the earnings of all kinds of workers because of the increased ability to offshore services electronically.

A second source that could be boosting profits is the China effect — the increased ability to offshore manufacturing production.

In contrast to the experience of the 1990s, when parent company employment actually increased more rapidly than employment in foreign affiliates, since 1999, employment in U.S. corporate parents has been declining, while affiliate employment has been growing.

But the recent employment pattern in U.S. multinational corporations actually reflects developments outside of manufacturing, in industries such as wholesale and retail trade in which off-shoring is not likely to be a major factor.

The data indicate that U.S. firms have increased their overseas operations but more to service the non-traded parts of the foreign economies rather than to source traded goods abroad.

More generally, if increased offshoring is responsible for depressing the share of compensation, one would expect to see these pressures operating particularly strongly in the tradable goods sectors.

The share of compensation should have been especially depressed in manufacturing, for example. But this has not been the case. In 2005, for example, the ratio of the share of compensation in manufacturing to the share in services was the same as in 2000.

In addition, one might have expected these pressures to be operating by weakening the bargaining power of organized labor in manufacturing. Yet compensation of unionized workers — especially those in manufacturing — has grown relatively rapidly during this period.

Instead, it has been the financial sector in which profit shares have been growing particularly rapidly and in which labor’s share stand historic lows.

It is likely, therefore, that a significant proportion of the low compensation share is cyclical. Profits are far more volatile than wages — and the share of labor fluctuates over the cycle, although not in a manner that is precisely pro- or anti-cyclical.

Instead, labor’s share is at its highest in the more mature phase of expansions and at the start of recessions. While labor’s share falls as the recovery sets in, productivity accelerates and profits surge.

By contrast, profit shares are highest in the middle of expansions and fall as the expansion matures and wage pressures build up.

In sum, U.S. workers of all skill levels are understandably concerned about their slow income growth — and the idea that global wage arbitrage has been exerting downward pressure on income growth is certainly plausible.

But the sectoral patterns are not compatible with this interpretation. There are reasons to believe the depressed share of labor compensation has a strong cyclical component.