The Myth of Middle-Class Job Loss

By STEPHEN J. ROSE
October 24, 2007

Economic change is a messy process. New technologies open up many opportunities for those prepared to take advantage of them. At the same time, old firms and their workers are displaced and forced to start over. In 1900, for example, 40% of the U.S. work force was involved in agriculture. Today, that figure is less than 2%, and no serious observer would argue that we are worse off as a result of this transformation.

Yet many of today's most prominent politicians and pundits are making an updated version of precisely this argument. They claim that the decline in the number of manufacturing jobs has led to the replacement of good middle-class jobs by low-skill, low-pay "hamburger-flipping" service jobs.


This kind of populist dogma is bad politics and even worse economics. The assertion that the American middle-class is disappearing along with manufacturing jobs is, put simply, based on an outdated view of how the economy operates, and is empirically wrong. Nonetheless, the view that the economy has failed the middle class is widespread. The outsourcing of jobs to low-wage countries is, of course, the latest culprit. Polemicists from all sides find it irresistible to blame expanding trade for middle-class decline. But how widespread a problem is outsourcing, exactly?

It is certainly true that many jobs in manufacturing clothing, steel, metal products and automobiles have gone overseas. Plant closures not only devastate the workers who are displaced, but they have also undermined the vitality of whole communities in North Carolina, Pennsylvania, New York, Michigan, Ohio and Wisconsin, to name just a few places. But while such communities are a clear sign of the decline in some sectors of the economy, there has been strong employment growth in many other sectors. In research just published by the Progressive Policy Institute, I show that incomes and employment have grown by substantial amounts in every state (even in the so-called Rust Belt) since the passage of the North American Free Trade Agreement in 1993.

In fact, there is no convincing, data-driven proof that trade has led to any overall job loss during the last 30 years. To the contrary, the economy has grown at a slow but steady rate (a few brief recessions notwithstanding) with trade and employment rising in tandem.

To prove that there has been substantial growth of middle-class jobs, I compare the situation that existed in 1979 with that of 2005. The base year is 1979 because it represents the last business-cycle peak before income inequality and the U.S. trade deficit began to grow quickly in the 1980s. To make the comparison fair, earnings in 1979 are increased by almost 150% to adjust for inflation.

Let us look at the distribution of earnings in 1979, compared with the distribution of earnings of the net new jobs created since that year. To begin with, it is necessary to assess the experiences of male and female workers separately. Unfortunately, it is still true that a large number of women are employed in occupational titles that are predominantly held by women -- e.g., teachers, nurses, and clerical workers.

Nevertheless, there has clearly been a sharp increase in female middle-class employment. As recently as 1979, 61% of female workers were in jobs that paid less than $25,000, and only 3% earned more than $50,000 a year. By contrast, more than 36% of new jobs that opened since 1979 for women pay more than $50,000 and only 17% pay less than $25,000.

Critics who bemoan the trajectory of the American economy over the past three decades somehow find it convenient to overlook or play down this historic improvement in the employment status and income levels of women. While women still lag in pay compared to men of similar educational attainment, the extraordinary rise in women's income since 1979 is a fact at odds with the notion of an overall decline in the American middle class.

For men, the change in employment since 1979 has not been quite as clear-cut, or as positive. There has been a tremendous growth in the number of men in high-paying jobs: In 1979, just 10% of male workers earned above $75,000, while fully 34% of new jobs since 1979 have paid this amount or more.

However, there was also growth in the share of male workers earning less than $25,000 a year, from 23% in 1979 to 36% by 2005. This rise of low-paying jobs hit less-educated men particularly hard. For those with just a high school diploma, 87% of the new jobs paid $25,000 or less.

Here's the bottom line: For three-quarters of the workforce (women and the top half of male earners), economic growth translated into earnings gains. But for male workers in the bottom half of the earnings distribution, the decline of unionized manufacturing employment has led to the drying up of some middle-class jobs for those with no post-secondary education.

For the clear majority of the workforce, then, the job market has become more welcoming, not less so. But where are these jobs?

Using a framework that I developed in the 1990s, I find that most of the employment gains over the last 30 years have been in business-management activities (administration, sales, finance and business services) as well as in professional services such as health care and education. While the percentage of U.S. jobs derived from manual work in agriculture, mining, timber and manufacturing has declined, the share of jobs related to low-skilled retail and personal/food services has remained steady.

Undeniably, some people have been left out of this middle-class workforce expansion and need help in making the transition to the new economy. In particular, the last six years have seen very little wage growth for the bottom 80% of the workforce. But we should bear in mind that real gross domestic product per person is up over 60% since 1979, and our goal for the job market should not be simply to keep pace with where things stood nearly three decades ago.

While the pessimists would have us go backward, we should be working today on expanding opportunities in the future. In particular, we have to address what we can do to help displaced men who lack post-secondary education. Higher levels of unionization and increasing the minimum wage would help, but they don't address the more basic need, which is to provide people with the necessary skills for the modern marketplace.

The economy can expand and provide more good jobs as long as workers have the education and training required to succeed. Talk of the "disappearance of the middle class" is actually counterproductive, because it distorts the real challenge. This is to make sure that our young men and women are better prepared to enter the workforce of the 21st century.

Mr. Rose is senior economic fellow at the Progressive Policy Institute, where he recently authored a report titled "The Truth About Middle Class Jobs." He has worked both for the Joint Economic Committee of Congress and as an adviser to former Secretary of Labor Robert Reich.