Jobs Loss From Imports: Measuring The Costs

Introduction: Focusing on
Displaced Workers

Increasing economic integration across countries (the international flows of goods, services, people, and capital commonly known as "globalization") produces large gains—particularly for the US economy as a whole. Productivity is enhanced when goods and services are produced in countries with a comparative advantage and then traded. Consumers get lower prices and improved welfare. Global competition helps restrain price inflation and spurs innovation. In turn, innovation spurs globalization. These gains, although widespread and widely touted, are not always well understood.

    But globalization also has costs. And these costs tend to be concentrated where the benefits are widespread. Producers in domestic industries whose products compete with imports (hereafter, import-competing industries) face falling profits and see their business threatened by the lower-cost competition. Workers in import-competing industries lose jobs or face downward wage pressure. Other workers fear job loss from heightened global competitive pressure and the perceived ease with which firms can relocate production. Americans are aware of both the pluses and minuses of continued globalization, as seen in public opinion polls (see Program on International Policy Attitudes 2000; Scheve and Slaughter 2001). But for many years, the "free trade" debate in the public policy arena has emphasized the benefits of trade and ignored the costs.

    This one-sided emphasis has been stopped—almost dead in its tracks—by theglobalization backlash. How can the public policy debate be started and expanded? To move forward again on the path of economic integration, it is time for the policy debate to include a better understanding of the distribution of the costs of globalization. There is broad agreement, at least in principle, on the net benefits of free trade. That says little more than that the (gross) benefits exceed the (gross) costs. Again in principle, the benefits can be used to compensate for the losses, producing a superior societal outcome. One important step in moving toward this outcome, and making these compensations, is to identify who bears the burden of the costs and to measure the size of the burden. This book—part of a larger study of globalization's costs and benefits—seeks to provide this basic information and draw policy conclusions.

    Much of the debate about the domestic labor market and free trade has been focused on the number of jobs affected. This focus is unfortunate; these accounting efforts should not form the central focus of the debate on the benefits of free trade or increased economic integration. The economy's level of employment—total number of jobs—is determined more by macroeconomic events and policy than by changes in trade policy or the trade balance.

Important Questions—and Answers

Lost in the misguided debate over the number of jobs created or destroyed by increased economic integration is the really important question: what kind of work will Americans do, as the dynamic US economy continues to change, with more trade and technological advances? Jobs are lost and created, and workers displaced and reemployed, in a dynamic economy. Rather than focus on how many jobs will be affected, we need to understand workers—who they are and how they will be affected. Specifically, who are the workers displaced from import-competing industries? What are their characteristics? How do they compare with others who lose their jobs? What happens after displacement? How do they adjust? What can we learn from the pattern of reemployment and earnings that will aid in (re)designing adjustment services?

    The research reported here focuses on individual workers, mostly in manufacturing. It builds on a foundation of more than a decade of displacement research. This research has shown that the earnings losses following job dislocation are large and persist over time (see Kletzer 1998b). To examine the costs of import-sensitive job loss, I use results from my earlier industry-level studies of the relationship between increasing foreign competition, employment change, and job loss to develop a definition of an import-competing industry and apply that definition to a nationally representative sample of displaced workers.

    This process yields a sample of import-competing displaced workers. The labeling of "import-competing"—or import-sensitive—job loss thus is according to the industry from which the job was lost. Although I make no strong claims about the precise cause of each worker's job loss, I am confident that the sample captures most of the kinds of jobs Americans feel to be "at risk" from increasing economic integration. My examination of the evidence proceeds through a series of questions, with the answers previewed here:

* Question 1: How can we define import-competing job loss? What can we learn about the scope of job loss from import-sensitive industries?

Answer: I rank detailed industries (defined at the 3-digit Census of Population Industrial Classification [CIC] level) by their percentage change in import share during the 1979-94 period (from largest positive changes to smallest). Using that ranking, industries in the top 25 percent are defined as the "most" import-competing ones. This industry definition is then applied to the worker-based Displaced Worker Surveys to draw out a sample of import-competing displaced workers, based on an individual's industry of displacement. During the 21-year period from 1979 to 1999, 6.4 million workers were displaced from an import-competing industry; about 17 million were displaced from manufacturing. Import-competing job loss is concentrated in a few large-employment industries: electrical machinery, apparel, motor vehicles, nonelectrical machinery, and blast furnaces.

* Question 2: Do import-competing displaced workers have different characteristics than other manufacturing workers? How do these workers compare with workers displaced from other sectors of the economy?

Answer: Import-competing workers are similar to other displaced manufacturing workers—slightly older, with virtually no difference in educational attainment or job tenure. The most striking difference between import-competing displaced workers and other displaced manufacturing workers is the degree to which import-competing industries employ and displace women. Women account for 45 percent of import-sensitive displaced workers, relative to 37 percent of the overall manufacturing displaced. Some industries stand out: Women account for 80 percent of those displaced from apparel, 66 percent from footwear, and 76 percent from knitting mills (part of the textiles industry). Thus we see that women dominate the group of displaced workers from these import-competing industries as a result of their strong representation in employment. In comparison with workers displaced from other sectors of the economy (e.g., wholesale and retail trade, utilities, and services), manufacturing workers are slightly older, are notably less educated, have longer job tenure, are somewhat more likely to be a member of a minority group, and are far more likely to be production oriented (blue collar workers).

* Question 3: What are the reemployment prospects of import-competing displaced workers? Does the likelihood of reemployment differ by industrial sector?

Answer: Import-competing displaced workers are a little less likely to be reemployed (63.4 percent were reemployed at their survey date) than other displaced manufacturing workers (65.8 percent were reemployed). Most of this statistically significant difference is accounted for by the lower reemployment rate of women—at 56.2 percent, 13 percentage points lower than the rate for men. Women are heavily employed and displaced from import-competing industries. More generally, manufacturing workers are less likely to be reemployed than nonmanufacturing workers. Some of this difference is accounted for by higher age, less education, longer job tenure, minority status, and gender. There is a remaining gap, with both durable-goods displaced workers less likely to be reemployed (by 4.2 percentage points) and nondurable-goods displaced workers less likely to be reemployed (by 2.7 points) than a worker displaced from a large set of nonmanufacturing industries (including transportation, communications, utilities, wholesale and retail trade, and services), where the reemployment rate was 69 percent.

* Question 4: How costly is job displacement, in terms of earnings losses? What do we know about the size and range of these losses? Who are the workers with the largest earnings losses?

Answer: Among the reemployed, import-competing displaced workers experience sizable average weekly earnings losses of about 13 percent. This large average loss masks considerable variation: 36 percent of import-competing displaced workers report earning the same or more on their new job as they earned on the old job, and 25 percent reported earnings losses of 30 percent or more. This average and distribution are very similar to those for manufacturing workers as a group. Older, less educated, lower-skilled production workers with established tenure on their old job are more likely to suffer earnings losses in excess of 30 percent.

* Question 5: Where do workers become reemployed? How accurate is the perception that all reemployment is in the services sector? Do exporting industries account for much reemployment? How important is changing industry in understanding earnings losses?

Answer: The reemployment sector matters a great deal for understanding the variation in earnings losses. Contrary to stereotype, not all displaced manufacturing workers are reemployed at McDonald's. Overall, just 10 percent of reemployed manufacturing workers are in retail trade, and this percentage is similar for import-competing displaced workers. Fifty percent of import-competing workers are reemployed back in manufacturing, and the same percentage holds for manufacturing overall. Export-intensive manufacturing industries do not account for a large share of reemployment. Regaining employment in manufacturing greatly reduces earnings losses. Earnings losses are largest for workers employed in retail trade. Reemployment in the same industry is very important for reducing earnings losses, with half of all these workers reporting very small or no earnings losses.

* Question 6: How can the pattern of reemployment inform our policy decisions on the future of worker adjustment programs?

Answer: The pattern of reemployment can reveal how the consequences of job loss vary and how some discernible transitions are better than others. Reemploying some workers in "the right places" might reduce the number of workers who need retraining. The best outcome for many import-competing displaced workers—particularly middle-aged workers with high tenure and a high school education—appears to be a return to their old industry. Their earnings losses will be smaller, for the most part. This outcome suggests tailored job search assistance, to help workers seek reemployment in old industries where skills are transferable. We know that workers can be provided with job-search assistance at a low cost. The patterns found here should also prompt a rethinking of the current focus of assistance on education and retraining (if these policy tools are predicated on reallocating workers to different jobs). Education, skill enhancement, and retraining do help some workers. But for older workers, many of whom are represented in this book, enhancing job skills and retraining will be problematic. Training programs can require higher education levels for entry than many workers have attained, and completing a course can be difficult. Furthermore, the evidence presented here shows that finding a job in a different sector can be very costly; thus, a policy response to economic structural change that depends on reallocating retrained workers across sectors (from the "old" economy to the "new") may not be an answer for all. A more broadly based labor market policy should include active assistance for finding new jobs and short-term financial assistance when the new job pays less than the old.

Informing Future Policymaking

One of the most important findings for policymaking is that, for most displaced workers, what matters is the kind of job lost and the kind of job regained. Why the job was lost does not matter much at all. If workers and consequences are alike, across differing causes of job loss (e.g., increasing foreign competition, technological change, and downsizing), then policymakers should consider adjustment policy for all displaced workers, and broaden program eligibility beyond "trade-displaced workers." A broadly based program is not only justified here by the data, but would also serve to reduce the perception that free trade is a special problem for workers, one that alone needs to be addressed by labor market assistance programs.

Thinking about Trade and the Labor Market

This book complements and extends some of the questions posed in much of the current research on trade and the labor market. During the past 25 years, as the US economy has become more open and integrated, manufacturing employment (particularly of production workers) has declined, real wage growth has been sluggish, and income inequality has increased. These coincident time trends have motivated the creation of an active research literature investigating the link between globalization and major recent US labor market changes. One core question: Is globalization a culprit in the deteriorating economic status of less-skilled workers? In a recent review of the empirical literature, Blanchflower (2000) concludes that globalization is not the major influence in recent labor market changes, but rather one of several important factors. Other key factors are skill-biased technological change, immigration, declining unionization, and declining real minimum wages.

    My approach differs from much of the trade and wages literature. First, microevidence on real displaced workers, their characteristics, and postdisplacement outcomes is used to measure the domestic impact of trade—in contrast to the most common measures, changes in net industry employment and in industry wages. Second, instead of pursuing ways to disentangle trade from technology, my operating assumption is that all manufacturing industries and their workers are affected similarly by trends in technology, outsourcing, capital deepening, and other related changes. I then infer the effects of extraordinary surges in import competition alone from the differences in outcomes between my samples of import-competing workers and otherwise-displaced manufacturing workers. Such differences turn out to be small, as summarized above.

    My questions and approach do not presume increasing economic integration to be the most important factor in domestic labor market changes. Nor do I presume that the forces of increasing trade can be separated definitively from other factors. It is very difficult to disentangle technological change from the heightened competitive pressures of globalization (i.e., it is difficult to separate trade from technology). In the main, I attempt to stay out of that fray in this book. My strategy for identifying import-competing industries does not imply that all workers displaced from these industries are displaced by rising imports. The causes of job loss in any industry are broader: technological change, restructuring, shifts in investment, changes in domestic demand. These causes are not my focus. My focus here is on the costs of (any kind of) job loss from import-competing industries.

    Before we begin, one note on limitations. This book focuses on job loss associated with trade, and not, for example, multinational investment. The trade focus is mostly on imports, not exports. That emphasis is admittedly one-sided, but it takes on directly the allegation that "imports cost jobs," and I believe this to be one of the most important perceived costs of globalization. As such, this book ignores job gains and the quality of jobs related to trade and investment. Others have written about exports and jobs (see Richardson and Rindal 1995, 1996), and about investment and jobs (see Slaughter and Blonigen 2000).

    The book proceeds as follows. I summarize my earlier industry-level research on increasing foreign competition, employment change, and job loss in chapter 2, as a prelude to developing a working definition of an import-competing industry. The trade and job loss data are then described. Applying my definition to the data, I introduce the categorization of manufacturing industries by degree of import sensitivity. The chapter concludes with a discussion of two-way trade.

    Chapter 3 examines the characteristics of workers displaced from the most import-sensitive industries, and compares them with other manufacturing workers and workers displaced from industries other than manufacturing. Chapter 4 turns to the first postdisplacement outcome, reemployment, and reports estimates from a straightforward econometric model of the probability of reemployment. Chapter 5 reports on reemployment earnings and earnings losses, and brings in information from studies that follow the same workers' experience over time.

    The industrial sectors where workers are reemployed is the topic of chapter 6. Here we see how earnings losses vary by reemployment sector. The importance of regaining employment in the same industry or sector to minimize earnings losses is also examined. Chapter 7 concludes, and discusses in some detail how these findings can inform future policymaking.

---

Excerpted from Job Loss from Imports: Measuring the Costs by Lori G. Kletzer. Copyright © 2001 by Institute for International Economics. Excerpted by permission. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.

---