Volume 10, Issue 2, Spring 2000
Inequality in America: The Recent Evidence
Robert Z. Lawrence
For the period from 1979 through the early 1990s, the story is a very familiar one. Whether the data
are arrayed by household, by earnings decile, or by indicators of skill and education, they all suggest
two fundamental results. First, growing inequality: the more advantaged in the U.S. did particularly well
while those at the bottom did poorly. Second, on average, workers fared poorly. Real wage growth was
extremely sluggish and by some measures-such as average hourly earnings-actually declined over the period.
These results were particularly striking in the face of two developments: the inequality trends had
persisted despite (a) the long expansion of the eighties and (b) the relatively rapid increase in
the supply of educated workers.
The literature advanced three lines of explanation for these developments: First, the force of
globalization. The U.S. economy ran large trade deficits in the 1980s. In addition, there was
a rise in the participation of developing countries in U.S. trade, partly occasioned by the
shift in developing countries such as China to more outwardly oriented policies. It has been
argued that these developments had particularly adverse effects on the demand for less-skilled
U.S. workers, who were increasingly forced to compete with low-wage workers. In addition, the
threats of outsourcing and increased international competition were said to reduce labor's
bargaining power-a factor which was also held responsible for the slow rise in average wages.
Immigration also contributed a larger number of less skilled workers.
Second, the force of technological change. It was also argued that there had been an acceleration
in "skill biased" technical change. Computers and/or new management approaches emphasizing lean
production skewed demand towards skilled workers and away from the less skilled.
Third, the force of institutional and structural changes. In the 1980s, de-unionization and
deregulation weakened the power of workers. In addition, the declining share of employment
in manufacturing reduced high-wage opportunities for blue-collar workers.
To be sure, these explanations are not mutually exclusive. Indeed they may be interrelated.
International competition may have stimulated technological change. International competition
has also been invoked to help explain the weakening of union bargaining power and the declining
share of manufacturing employment. While economists differ on the relative importance of these
forces-a majority arguing that technological change was dominant-most assign at least some role
to all three. I'd like to examine what has happened in the U.S. economy since the mid-1990s.
Let's ask first what has happened to these forces. Next let's look at the recent evidence on
what has happened to various indicators of inequality, and then I'd like to reflect on what
lessons we might draw.
The Forces at Work over the Past Five Years
First, what has happened on the globalization front? New trade agreements have been implemented-in
particular NAFTA and the Uruguay Round. Trade has continued to rise as a share of the U.S. economy.
In the aftermath of the Asian crisis, the U.S. trade deficit in goods and services increased
dramatically. In addition, particularly since 1997, the prices of imported goods have fallen
relative to the Consumer Price Index, and the prices of manufactured goods from developing
countries, which are typically labor-intensive, have fallen even more rapidly than those from
developed countries. By any measure therefore-trade agreements, trade volumes, trade prices,
and trade deficits-globalization's forces were more powerful in the nineties.
Second, what has happened to technological change? One very powerful feature of the strength
of this expansion has been investment in plants and equipment, particularly computers and
information technology. This can be seen in the dramatic rise in the share of real investment
in GDP and the share of high-tech investment in total investment, and, of course, the diffusion
and development of the Internet. A second feature has been the heartening acceleration in
productivity growth over the past three years. While productivity growth had been slow in the
early part of this expansion, we've seen business sector productivity up 1.5 percent in 1997,
2.4 in 1998, and 3.0 in the year prior to mid-1999. Productivity levels are now well above the
trend rate since 1973. By most measures of both inputs and outputs, therefore, the pace of
technological change driven by the information revolution and other factors appears to have
Third, what has happened on the institutional and sectoral front? Union membership has continued
to decline in 1990s. Membership in the private sector has fallen by over 400,000 since 1994. The
share of employment in manufacturing has declined steadily and, after rising through 1997, the
number employed has fallen. In addition, welfare reform has added new workers, many with limited
work experience, to the labor force.
In the light of these earlier explanations, most would surely have predicted even greater inequality
in the 1990s, and a particularly tough experience for those at the bottom end of the income distribution.
Yet that has not happened. Indeed, by numerous measures there has been no increase, and by some,
inequality has actually been reduced.
The Recent Experience
Real wages and labor's share. Since 1995 we have seen significant increases in real wages. In
addition, the share of output received by workers in the form of compensation has been rising
since 1996. It shows almost no change over the longer run, suggesting that there has been no
significant change in the shares of the pie going to labor and capital. There does not seem to
be strong recent evidence that international competition has weakened labor's ability to bargain
in this recent period. The most recent wage agreements concluded in the aerospace and automobile
industries have been particularly strong.
College premiums. There has been no acceleration in the premium over the past five years for
either men or women. The ratio of average weekly wages of college graduates to those of high
school graduates had increased from 38 percent in 1979 to 74.3 percent in 1994, but between
1994 and 1998 it actually declined to 71.3 percent. A similar story emerges in data for average
hourly earnings, although the hourly premiums are somewhat lower.
Earnings by decile. Over the past few years, the spread between the hourly earnings at the 90th
percentile and those earning at the 10th percentile has stopped rising. The same story is evident
in weekly earnings. Between 1979 and 1996 the ratio increased to 4.45. Between 1996 and 1998,
however, it declined to 4.36. In the weekly data, likewise, the 90/50 spread also rose from 1.86
to 2.12 in 1996 and then subsided. An even more dramatic picture emerges in the hourly data in
which we array the changes between 1994 and the first half of 1999. Here, when men and women are
grouped together, we see increases that are highest at the bottom and lowest at the top.
(They're up 8.6 percent since 94 in the lowest decile, up 4.7 at the top.) Similar images of faster
growth at the bottom than at the top emerge for both men and women when the data are separated by
gender. Thus the hourly earnings data actually indicate that there has been an increase in wage
equality over this period.
More comprehensive indicators of inequality-income gains by quintile and an aggregate measure
of inequality (the "Gini coefficient" for household income)-support the view that inequality
has not increased. Since 1993, incomes have grown by between 9.9 and 11.7 percent for every
quintile of the income distribution. After two decades of rising inequality the Gini coefficient
measure showed no statistically significant change between 1993 and 1998. Here too, it appears
as if the Gini is back in the bottle.
The least fortunate. The poverty data show excellent progress since 1994. The 1998 poverty
numbers suggest poverty is lower than at any time since 1979. The poverty rate has fallen
from 15.1 percent in 1993 to 12.7 percent in 1998. Over the last five years typical families
have seen their incomes rise by 12.1 percent and African American families by 21.0 percent.
Likewise, unemployment gains at the bottom are particularly impressive. Since 1993, African
American unemployment has declined from 14.3 percent to 7.8 percent-the lowest on record-while
Hispanic unemployment has fallen from 11.5 to 6.5 percent-also the lowest on record.
The overwhelming impression, therefore, is that the inequality has stopped rising. It is
certainly incorrect to claim that we have actually reversed the drift towards inequality
of the 1980s, and hard to make the claim that inequality is now on a downward trend-and
some see trouble in that. But this is convincing evidence that inequality has not continued
to worsen. And that is highly significant. It is important, therefore, that this new evidence
be introduced into the discussion.
Searching for Explanations
How can we explain what has happened? Why is it that despite the apparent strength of these
forces-globalization, technological change, and institutional and structural changes-the
inequality outcome is not what might have been expected? There are both macroeconomic and
One obvious candidate is the role of the high-pressure economy and the associated low rates
of unemployment. There is certainly compelling evidence that a high-pressure economy tends
to increase opportunities for workers, particularly at the lower end of the scale. It may
set in motion a positive virtuous circle in which new opportunities lead to new skills,
higher productivity, and improved competitiveness, which in turn may allow for labor markets
to be run at higher levels without upward pressure on wages. As the labor market tightens,
there will be more opportunities and higher pay for those with fewer qualifications.
Ironically in this regard, forces which were seen as working towards increased inequality
may well have contributed to allowing us to have a high-pressure economy. First, globalization
and increased competition can lead to a onetime reduction in prices and margins which improves
the short run inflation/output trade-off. Similarly price pressures may reduce profits and raise
the relative share of labor. And lower import prices help contain inflation in the short run.
In addition, when there is excess capacity in the rest of the world, we can draw on imports
rather than strain domestic capacity utilization, again resulting in fewer price pressures.
Second, as Alan Greenspan has emphasized, when we get positive shocks to productivity due to
technological change, there is a powerful positive macroeconomic influence allowing firms to
pay higher wages without them passing through into inflation.
A second possibility is that the trends themselves have been brought to a halt by offsetting
microeconomic forces. Perhaps "technology" is responsive to relative wage costs. Skill bias
does not come down from heaven but eventually shifts in response to incentives. Firms may
increasingly find ways to employ relatively less well educated workers more effectively because
these workers have become relatively cheaper. Or maybe the downward pressures from "trade" or
globalization have run their course. In particular, the issue of substitutability between
domestic and international production is relevant. It could be that eventually U.S. firms
either are driven out of low-wage activities, or they figure out how to compete using competitive
strategies which offset cheaper foreign labor costs. This means they no longer compete head to
head with low-wage countries and thus are less affected by competitive pressures. Similarly,
vulnerable unions may be forced to reduce members so that those who remain are actually the
more powerful ones.
While the macroeconomic explanation implies that inequality could reverse in the face of higher
unemployment, the microeconomic considerations suggest that we may have a more resilient economy.
We may finally be reaping the benefits of having adjusted to some major structural challenges.
But it is also possible that our explanations were defective in the first place, that we actually
have never had a very firm grasp on what was shifting the trends originally. There were always
problems with the story assigning a large role for trade. The numbers simply didn't seem to add
up to anything big enough to explain a large share, and the other implications of the theories
didn't seem to be corroborated.
There was also a problem with the skill-biased technology story. Why so much change and so little
progress? Why in the eighties did we have changes in technology that were large enough to wrench
the labor market and yet too small to boost productivity? Ironically, over the past five years we
are getting the "progress," i.e., faster productivity growth, without the relative wage changes.
Finally, credit for the differences in behavior should also be given to policy-on the macroeconomic
front to both fiscal and monetary policy. The monetary authorities have been vigilant in avoiding
inflation and yet willing and able to provide sufficient liquidity to finance non-inflationary
growth and deal with financial crisis. Fiscal policy has been prudent with a long-run trend towards
surplus, while at the same time resources devoted to education, training, and civilian technological
development have been increased. As National Economic Council chief Gene Sperling has recently
pointed out, strong deficit reduction has not hurt poverty reduction. Increased funding for the
earned-income tax credit and for other education, training, and research has played a role.
Important, too, have been increases in the minimum wage. Certainly this has raised wages at the
bottom and it does not appear to have had a noticeably negative impact on the employment of the
least well paid workers.
To conclude: We should not be complacent. We have not reversed the previous shifts towards
greater inequality. Ideally we would like to see sustained gains for all and particularly
large gains at the bottom. But we have made progress in arresting trends that were disturbing;
overall these data suggest we have been doing the right things. In addition, it provides an
optimistic message. We remain masters of our fate and are not, as some suggest, condemned to
be buffeted by negative global or technological forces in the face of which we are helpless.