The Measurable Costs and Benefits of Globalization: An Economic Analysis.
Senior Fellow, Institute for
For presentation at a conference on
Responding to Globalization:
Societies, Groups, and Individuals
Globalization and Democracy Research and Training Program
Program on Political and Economic Change
· Overview (p. 1)
· Detailed Summaries (p. 4)
· References (p.13)
starts with the
Five projects have been completed, with at least six more planned.
The ultimate aim of the GBS projects is the capability to weigh credible measures of traditional and new costs and benefits against each other. Since many of the newer costs and benefits are distributional (some gain at the expense of others), and since some cannot be quantified at all, this ultimate aim will involve judgement that goes beyond mere measurement. The so-called “capstone” volume will make this attempt.
A derivative aim of the GBS family of projects is a template for similar exercises abroad and on a global scale.
Completed and Nearly-Completed GBS Projects:
Five projects are completed or nearly so, two by Lori G. Kletzer, one by Howard Lewis and J. David Richardson, one by Kenneth F. Scheve and Matthew J. Slaughter, and one by Robert E. Baldwin,.
of Kletzer’s projects (2001a) tries to measure the effects of exceptional
import penetration on American manufacturing workers – in isolation from trends
in technology, outsourcing, unionization, and other shocks during the last
quarter of the twentieth century. The
import effects on worker dislocation and job/earnings recovery turn out to be
modest, and hard to disentangle from the special burdens that women workers
face in the American labor market. But
the effects on manufacturing workers of generalized dislocation from both global and local shocks are by no
means modest, and they bear them especially heavily compared to other American
workers. This conclusion leads Kletzer,
together with IIE and other co-authors, to propose a radical re-shaping of
Lewis and Richardson (2001) synthesize a growing body of microeconomic research that suggests that these gains may be quite large for American workers, firms, and communities – as long as they are able to engage the global economy in some fashion. Globally engaged Americans earn more, grow those earnings faster, and fail less frequently than comparable, yet insular Americans. Global engagement includes not only exports, but imports, inward and outward investment, and technology transfer. As globally engaged population shares grow, the micro-level gains rejuvenate workplaces, industries, regions, and the entire American economy. Another generation’s growth in global commitment at the same rate that the past generation globalized would raise standards of living by a quarter. But there is another face to these benefits, a distributional downside. Workers, firms, and communities that cannot or will not engage the global economy lose heavily over time. They are displaced by the globally engaged. The challenge is to empower them, or their children, to be able to engage – to engage change of all types, especially when their skills, education, and history make such engagement unlikely or very costly.
helps explain why Scheve and Slaughter (2001) find so much skepticism in their
American voter surveys about globalization in the early and mid-1990s.
A large number of voters with median skills and education (or less) do
not favor freer trade, nor freer immigration.
They fear – perhaps correctly, as Scheve and Slaughter show – that they
have little or nothing to gain from further
Emerging Policy Implications from the GBS Projects So Far:
Taken together, these findings suggest a radical re-orientation of familiar American worker-adjustment policies -- especially those that are consciously adopted to complement and enable global liberalizing initiatives, spreading the gains from global integration more widely within American society and stabilizing the uneasy popular support for them.
· Global-complement local-adjustment policies should move away from specific industry-and-job-based relief and toward worker empowerment.
· They should move toward education and skill-building experience, including on-the-job training, and towards insurance programs that preserve a individual’s lifetime earnings potential, including the potential to accumulate savings and to build human capital, including experience in particular jobs.
One such program is the combination of wage insurance and subsidized health insurance for dislocated workers, proposed by Kletzer and Litan (2001). Another is policies that allow older and less mobile dislocated workers to avoid having to change sectors, or more pragmatically, policies that facilitate mobility especially among younger workers (e.g., job search programs, benefit portability enhancements). That in turn, helps stabilize individuals’ lifetime wealth and makes them less fearful of global initiatives that expose them to capital losses (e.g., on their homes).
GBS Projects on the Horizon and Beyond:
Six more GBS projects are on the horizon.
Levinsohn (2002) on adjustment to globalization in American textiles and apparel, both costly adjustment and surprising opportunity.
Elliott, Kar, and Richardson (2002) on the identity and economic underpinnings of the globalization protest movent.
Kletzer (2002) on the special pressures faced by especially disadvantaged workers in especially import-sensitive sectors (textiles, apparel, footwear, leather) in the late 1990s.
Kletzer and Lawrence (2002) on the more detailed (political?) economics of earnings and benefits insurance proposals.
Elliott and Richardson (2002) on open trade in “worker-agency services.”
· Detailed Summaries, Completed and Horizon Projects (Alphabetical by Author)
first exams US deunionization in the raw data and in cross-tabulations. Deunionization has been especially strong in
manufacturing and among basically-skilled workers. Unions have maintained their membership
better in services sectors (especially public services) and among
better-educated workers. They have also
maintained wage differentials over comparable non-unionized workers better for
better-educated members. Deunionization
has been strong across all regions of the
In both decades, but especially in the first, pure trend declines in union affiliation – within each goods and services sector, within each region, within each worker educational group -- explain half to most all the remaining fall in the unionization rate.
for basically-educated workers, the corresponding econometric calculations of
the impact of net imports are 13 and 45 percent, respectively, for each decade,
a larger impact, though pure trend declines continue to be still larger. And especially in the first decade, the sectors
in which imports surged had a disproportionately large effect on
Elliott, Kar, and Richardson (2002) identifies some of the critical groups involved in the anti- or alternative globalization movement. It attempts to sketch the key issues and concerns that motivate them in a way that is broadly representative and intelligible to economists.
Lori G. Kletzer:
(2000) Background Research. Lori
G. Kletzer’s research over the past few years has involved pulling together the
richly detailed U.S. worker-by-worker surveys with
American import data -- and for the first time, export data, too – by sector in
Much of Kletzer’s research is summarized comprehensively
in her manuscript submission to the W. E. Upjohn Institute for Employment
Research (Kletzer (2000)). This
manuscript sets new standards for comprehensiveness within the voluminous
literature on globalization and
Kletzer’s most arresting policy conclusion from this
background research is that even-handed U.S. globalization, that is, both export and import-deepening, has
probably -- and counter-intuitively --increased
manufacturing jobs overall and reduced worker dislocation in
manufacturing from 1979 to 1994. The
industries where this is not true, and where intuition rules after all, are a
familiar and unfortunate group – apparel, footwear, leather, watches and
clocks, and a few others. But these
industries are NOT typical -- it is wrong to think their experience represents
Nevertheless, import pressure was intense over this period for important industries, that employ large numbers of American workers. And export opportunity, though welcome, did not reverse the pressure for them.
So how significant for workers was the import pressure alone for the sectors most susceptible to it? Might a narrow resistance to globalization in those sectors alone, a “targeted protectionism” aimed at their workers alone, have made much difference to them? Are these sectors that might, for the sake of their workers, stay “off the table” in future trade negotiations?
Kletzer explores some of these questions in her several
contributions to the IIE’s Globalization Balance Sheet family (Kletzer
(2001a,b); Kletzer and
Kletzer (2001a). In her nearly completed IIE project, Kletzer examines the effect of imports on American worker dislocation from 1979-99 using the Bureau of Labor Statistics Displaced Worker Surveys. She focuses on: (i) the incidence of involuntary dislocation and of subsequent reemployment; and on (ii) subsequent earnings recovery. She compares incidence and earnings effects among three groups of workers. One group is dislocated from “import-sensitive” manufacturing sectors; the second group is dislocated from other manufacturing industries; and the third group is a background control group of workers dislocated from the rest of the economy.
By comparing import-displaced manufacturing workers to otherwise displaced manufacturing workers, and both in turn to the rest of dislocated American workers, Kletzer cleverly isolates her estimate of the effects of import competition alone. She surmises that the influence of technical change, outsourcing, de-unionization, the business cycle and other important causes of worker dislocation are not systematically different between import-sensitive and other manufacturing industries. She is thus able to argue that the differences that remain between her import-displaced manufacturing workers and normally displaced manufacturing workers are due primarily to import-sensitivity. Manufacturing workers as a group experience higher dislocation and weaker earnings recovery than other American workers.
Kletzer finds that import-displaced workers do face some modest additional labor-market burdens, though not on every measure.
The clear consensus of these findings are that involuntary dislocation is a severe shock to American workers, but that import pressures play only a modest role in accentuating the shock. The implications for policy are that alleviating import pressures has no special payoff in moderating the severe distributional burdens borne by dislocated workers, but that a more target-effective program would confront the remployment dilemma and earnings losses of all workers dislocated for any reason. Kletzer proposes exactly this in Kletzer and Litan (2001).
and Litan (2001). In this
recently released IIE Policy Brief, Kletzer and Robert E. Litan of The
Brookings Institution use the Displaced Worker Surveys to estimate the costs
of a two-part proposal to ease the
burdens and shorten the duration of
In this spirit, the most arresting conclusion of their study is how cost-effective their proposed policy seems to be. Their program would cost the public budget only $3 to $3 ½ billion dollars and would entail:
· Wage insurance to replace 50 percent of a full-time dislocated worker’s earnings loss in a new job up to $10,000 of total annual reimbursement for a two-year period of eligibility.
· Health insurance premium subsidies (50 percent of the displaced worker’s contribution) to maintain affordable medical insurance for six months after a worker’s dislocation.
Even more generous provisions bring the public cost no higher than $6.5 billion.
Kletzer and Litan make no calculation of the savings to the conventional unemployment insurance system from sharpening incentives to become re-employed. Such UI savings only makes their proposal more cost-effective.
and Lawrence (2002). Kletzer
and IIE Visiting Fellow Robert Z. Lawrence plan to extend the work above on
worker dislocation policies, and more broadly evaluate the most prominent past
and prospective policies for worker training, search, mobility, adjustment, and
earnings enhancement. This work may have
a broader focus than the
(2002). Kletzer (2001a)
isolated a set of
Levinsohn and Petropoulos (2000) is background research that lays the groundwork for Levinsohn’s IIE project (Levinsohn (2002)).
In it, Jim Levinsohn of the
Levinsohn restricts his attention to the U.S. textile and apparel industry because it poses the greatest challenge to the optimistic view that globalization rejuvenates entire industries, allowing stable, high-wage, high-productivity plants and firms to grow at the expense of less successful rivals (this is the view that underlies the unappreciated benefits of globalization articulated in the GBS project by Lewis and Richardson (2001)).
Levinsohn finds mixed results in this regard. Plant-level dislocation (exit) is indeed concentrated on low-productivity plants, and newly entering plants do indeed have higher productivity than the exiters, consistent with the optimistic rejuvenation impact of globalization. But high-productivity incumbent plants do not grow much faster than low-productivity incumbent plants. So the size of the rejuvenation effect is small at best – the evolutionary turnover and growth of plants makes only a modest contribution to each industry’s productivity growth. And there is only a little evidence that workers in the those plants that have globalized most have avoided the severe downturn in wages in these industries, as implied by some of the optimistic accounts in Lewis and Richardson (2001).
More specifically, Levinsohn finds that
Both industries have been affected by overseas outsourcing, especially apparel, but both feature successful outsourcers who pay their workers roughly $1000 more in annual salaries for every 10 percent of their inputs that they outsource. In apparel, successful outsourcers also have more stable jobs (higher plant survival rates), consistent with general results in Lewis and Richardson (2001). But Levinsohn does not find the same stability payoff for textiles. And on the darker side of the outsourcing phenomenon, outsourcing-sensitive plants that historically specialized in those stages being outsourced have extremely high dislocation rates.
Howard Lewis and J. David Richardson:
Lewis and Richardson (2001) synthesizes new research that shows how globally engaged American firms, workers, and communities enjoy significant advantages compared to otherwise-identical counterparts (twins) – comparable firms, workers, and communities who are not globally engaged. Productivity, growth rates, wages, and job stability differ between the twins by significant amounts, usually from 5 to 15 percent.
These sizeable rewards for the globally engaged persist in bad times as well as good. They characterize small firms as well as large, in low-tech as well as high-tech activities, normally skilled as well as highly skilled workers, union members, minorities, small towns and other seemingly unlikely communities. There is little to no evidence in this research that global engagement helps capitalists more than workers, nor the strong and mobile any more than the small and solidarity-minded. Global engagement seems to benefit all … who partake. Only insularity seems to penalize.
The same new research shows that global “engagement” is not just export success. It is much more diverse and multi-dimensional. Global engagement includes importing and cross-border investment and technology exchange, in either direction.
Each type of global engagement generates rewards. And when this new research is pulled together, it implies that exports, imports, investment, outsourcing, and licensing of technology are an integrated family of global commitments. They “hang” together.
And the rewards from this diverse family of globally integrated activities are much larger, and much more widely dispersed across American society, than is commonly appreciated. For example, even in Appalachia, an apparel firm that exports, imports, outsources, and invests abroad has been the stable workplace for multiple generations of women workers with basic education and no foreign experience, many of whom have risen to positions of significant responsibility – but comparable rivals without global engagement have vanished along with their jobs. (This report punctuates its research survey with illustrative profiles like this one.)
In sum, the good new-news from this research is that globally engaged twins seem to enjoy special opportunities. But the bad new-news is that the gains of one twin are sometimes at the expense of the other, even when the family as a whole is better off.
In other words, rewards from global integration are there, but they are not there for every American ... at least not yet.
The reasons for these patterns and correlations are not yet clear from the new research. One interpretation is simply that those who are unusually gifted with special talent and luck are naturally able to succeed in global endeavors … and do succeed; whereas those who are normally endowed can’t and don’t and shouldn’t be urged to “go global.” But another interpretation, not inconsistent with the first, is that global engagement energizes, informs, motivates, insures, … and does all this for firms, workers, and communities who make global efforts of any variety.
central conclusion from this new research is that the
Matthew J. Slaughter and Kenneth F. Scheve:
backlash against globalization has itself gone global in the past 10 years,
even in the
Slaughter and Scheve (2001) aims to document and explain the backlash among American voters. They show first that it is widespread, and not just the vocal objection of a small fringe.
The authors then examine the backlash econometrically. They sample voter-by-voter preferences toward restrictions on trade and immigration from three surveys in the mid-1990s. They ask first whether these preferences are economically consistent, using measures of a voter’s socioeconomic status, industry of employment, location of residence, and other familiar variables.
In qualitative terms, American voters’ preferences correspond with their economic situation fairly closely.
But quantitatively, a large group of voters seem to fear exposure to further trade and immigration with far greater intensity than they welcome the benefits. Previous researchers have found similar results.
Slaughter and Scheve reach two striking and new conclusions about these patterns.
conclusion is that preferences for restrictive trade and immigration measures
become progressively weaker as a voter’s skills and education rise, and that
this correlation overwhelms almost all others.
Specifically, neither the import or export sensitivity of the voters’
jobs, nor their residence in immigration “gateways,” correlate well with the
remaining variation in preferences.
Furthermore, a majority of typical
The authors’ second novel conclusion is that the import sensitivity of a region’s output mix does have one important indirect effect — homeowners in import-sensitive regions are more likely to favor restrictive border policies, presumably to reduce the chance of capital losses on their main asset.
Taken together, these findings suggest a radical re-orientation of familiar American worker-adjustment policies. Such policies should move away from industry-and-job-based relief, and toward education and skill-building, including on-the-job training, and towards insurance programs that preserve a individual’s lifetime earnings potential, including savings.
Slaughter and Scheve document that trade liberalization with such adjustment policies does earn the support of a plurality of voters.
Baldwin, Robert E.
2002, forthcoming. Deunionization in the
Elliott, Kimberly Ann; Debi Kar; and J. David Richardson (2002) ….
Elliott, Kimberly Ann and J. David Richardson. 2002, forthcoming. Free
Trade in Worker Agency Services!
Kletzer, Lori G. Forthcoming 2002. Imports, Exports, and Jobs: What Does Trade Mean for Employment and Job Loss? 2000 manuscript accepted subject to revision by the W.E. Upjohn Institute for Employment Research, June.
_. (2001a), forthcoming. Job Loss from
Imports: Measuring the Costs.
_. (2002), forthcoming. [No
Kletzer, Lori G. and Robert Z. Lawrence. 2002, forthcoming. [No
Kletzer, Lori G. and Robert E. Litan. 2001. A Prescription to Relieve Worker Anxiety.
Levinsohn, Jim and Wendy Petropoulos (2000). Creative
Destruction or Just Plain Destruction?:
Levinsohn, Jim (2002), forthoming. Responses to Globalization in the
Howard and J. David Richardson. 2001,
forthcoming. Why Global Commitment Really Matters!
J. David. 2000. “The WTO and Market-Supportive Regulation: A
Way Forward on New Competition, Technological, and Labor Issues,” Federal
Reserve Bank of
_ _ _. 2001, forthcoming. “Designing a Market-Enhancing WTO,”forthcoming in Alan M. Rugman and Gavin Boyd, eds., Millenium Round Trade and Investment Issues: Structural and Policy Contexts, …...:Edward Elgar, New Horizons in International Business, 2001.
Kenneth F. and Matthew J. Slaughter (2001).
Globalization and the Perceptions
of American Workers.
 It is rare for economists to construct comprehensive distributional accounts of a country’s gainers and losers from global integration. Yet the identification of such gainers and losers occupies much of the political debate. Such distributional accounts are a necessary input to the GBS projects.
 The Nobel Prizes in Economics in 2000 were awarded to James J. Heckman and Daniel McFadden for pioneering techniques in handling such data and their widespread applications.
Institute is cooperating with the Deutsche Gesellschaft fur Auswartige Politik
 Inport trends alone, without netting out exports, contributed more importantly to deunionization. Exports trends alone sometimes helped unionization rates grow and sometimes not.
 And by 5-3 in all sectors.
agency services encompass primarily: collective representation and bargaining
over wages, benefits, and working conditions; workplace safety monitoring;
grievance and dispute settlement; training, apprenticeship, and employee
assistance; financial counsel (e.g., for pensions) and management of other
benefits (e.g., child care).
proposal has been given brief treatment in
 Current Population Survey (CPS), National Longitudinal Survey (NLS), and Displaced Worker Survey (DWS).
 Kletzer estimates, for example, that if American manufacturing exports and imports had both been held constant at 1979 levels, by 1994 American manufacturing jobs would have declined by 16.4 percent, instead of the milder measured decline of 13.5 percent.
 Import-sensitive sectors are defined as the top 25 percent of 3-digit SIC manufactures, rank ordered by the percentage change in their import share from 1979-1994. Kletzer experiments with other ways of defining them, finding her results robust, and also shows that her import-sensitive sectors correlate very tightly with those certified as import-sensitive for NAFTA-TAA Adjustment Assistance in the late 1990s. This gives credence to her claim that her results are up-to-date even though her data for classifying industries end in 1994. See Kletzer (1991b), however, for further testing of this assertion.
 Workers displaced from agriculture and construction are not covered.
 On sector by sector basis, of course they are different. Levinsohn, for example, in his GBS project (2001), shows how apparel has been heavily influenced by outsourcing, whereas textiles have been heavily inflenced by capital-intensive technological change. However, what would make Kletzer’s isolation of import effects wrong is a systematic difference (in technological change, outsourcing, etc.) in the group of sectors she identifies as import-sensitive compared to the remainder of manufacturing. This seems unlikely, or at least unproven, and sets a hurdle for critics of her method that they must leap to be persuasive.
 Among other influences that all manufacturing sectors seem to share in common is the effect of exports on worker dislocation. See Kletzer (2000).
 On average three years later.
 On average three years later.
 Both they and Slaughter and Scheve cite polls that show how voters shift their preferences in favor of further trade liberalization when it is tied to programs that assist workers to adjust.
 No dollar cap on wage insurance reimbursement; 70 percent reimbursement of earnings losses and worker contribution to medical insurance.
 Overall textile productivity grew 68 percent between 1972 and 1992, whereas overall apparel productivity grew only 26 percent. Most of the trend for each industry is due to productivity growth within plants that survived over the entire period.
 Kletzer (2000), Ch. 6, however, finds that across all manufacturing, industries with strong growth in overseas outsourcing also had more rapid declines in employment than other industries, especially from 1985-1994..