Return to the Institute's Homepage/Policy Briefs Section Number 01-9
July 2001
Steel: Big Problems, Better Solutions by Gary Clyde Hufbauer and Ben Goodrich Gary Clyde Hufbauer is Reginald Jones Senior Fellow at the Institute for International Economics. He is coeditor of The Ex-Im Bank in the 21st Century: A New Approach? (2000) and coauthor of World Capital Markets: Challenge to the G-10 (2001) and Steel Quotas: A Rigged Lottery, International Economics Policy Brief 99-5 (1999). Ben Goodrich is a research assistant at the Institute for International Economics. © Institute for International Economics. All rights reserved.
Once again the steel industry is distressed. And once again Washington is looking for trade “solutions” when it should be searching elsewhere. There is nothing new about either story. But, can’t we do something better this time? A Short History of Steel Trade Steel trade has been in turmoil since the late 1960s. Without exaggeration, more Washington trade lawyers work on steel disputes than any other trade issue. To recap the trade saga: o
In 1968, US steel producers filed a series of countervailing duty cases against subsidized European steel- makers. These cases led to “voluntary restraint agreements” that were terminated when the global steel market recovered in 1974.1
o
In 1977, US steel producers filed a series of antidumping cases primarily aimed at Japanese steel firms. These cases led to a system of minimum reference prices for
steel imports, known as the “trigger price mechanism” (TPM). The TPM system was soon extended to European steel. o
In 1974 and 1979, at both the launch and the ratification of the Tokyo Round of multilateral trade negotiations, the US antidumping law was amended (at the insistence of the steel industry and to conform with the Tokyo Round Antidumping Code), making it easier for domestic producers to prevail in antidumping cases.
o
In 1982, dissatisfied with the workings of the TPM system, US steel producers filed many antidumping and counter- vailing duty cases. These were resolved by new voluntary restraint agreements, which lasted until 1992.
o
In the mid-1980s, trade remedy cases were filed against “new” exporters, such as Brazil and Korea. Most of these cases resulted in high antidumping and countervailing duty penalties.
o
In 1989, the United States launched an effort to negotiate a Multilateral Steel Agreement designed to abolish subsidies. The negotiations failed and were ultimately abandoned in 1997.
o
In 1992, when the voluntary restraint agreements expired, a new set of trade remedy cases were filed. Many of the resulting penalty duties remain in effect today.
o
In March 1999, the House of Representatives passed H.R. 975, Congressman Peter Visclosky’s (D-IN) steel quota bill, 289 to 141. After a spirited debate,2 bill was defeated in the Senate. The current round of steel trade initiatives essentially renews the 1999 debate.
Why so much trade turmoil? One reason is persistent overcapacity in the global steel industry abetted by widespread market distortions. Persistent overcapacity has translated into cyclically falling prices and industry losses in every business slowdown. Another reason is the combination of rapid productivity growth and slow demand growth. Wheat farmers and steel workers share two traits: both have greatly increased their output per worker-year and both face sluggish demand for their products. The result is a painful secular decline in employment. These forces—in an effort to cushion the domestic
steel industry—have provoked a series of rearguard trade actions.
Fast Productivity and Slow Demand Between 1990 and 2000, crude steel output per worker-year increased from under 400 tons to nearly 600 tons in the European Union, Japan, and the United States. During the same period, world steel production was essentially flat—786 million metric tons in 1989 and 788 million metric tons in 1999—a decade when world GDP grew by about 30 percent.3 In combination, a 50 percent gain in productivity coupled with zero demand growth meant that steel employment in the three big industrial areas had to fall by a third—and that’s what happened (see table 1). In any industry, labor layoffs of this magnitude are bound to trigger trade friction. Overcapacity in Steel In 1998, in the midst of the Asian financial crisis, world steel overcapacity was estimated at 275 million metric tons, against production of 776 million metric tons (see table 2). By these estimates, overcapacity in 1998 exceeded a third of actual production. That was a cyclically high figure, but during the past 30 years, overcapacity has often exceeded 20 percent of production. By these estimates, overcapacity in 1998 exceeded a third of actual production. That was a cyclically high figure, but during the past 30 years, capacity has often exceeded 20 percent of production. Persistent overcapacity of this magnitude translates into downward price pressure, especially during business downturns. The economic logic is simple. Big, integrated steel firms face high fixed costs. Importantly, their fixed costs include not only the familiar cost of interest payments on corporate debt, but also the “legacy costs” of pension, health, and severance benefits promised to both current and retired workers.4
US trade barriers on steel imports may cause more job losses in US steel-using industries—such as autos, heavy equipment, machinery,
and construction—than the number of jobs saved in steel mills. The reason is that higher steel costs may force user industries to raise their prices and reduce demand.
When fixed costs are high, it makes sense for struggling steel firms to continue running their plants so long as the marginal revenues from extra production at least cover variable costs. Economic logic at the firm level ensures depressed prices—and widespread operating losses—at the industry level. To be sure, economic logic differs somewhat for mini mills that purchase scrap steel and electricity as their key inputs. For most mini mills, variable costs are high and fixed costs are low. They can shut down more quickly when prices fall. But while mini mills account for a big share of US steel production (over 45 percent), their share of global production is much smaller. The world steel industry is still characterized by integrated steel producers and their overcapacity problems.
Market Distortions In a normal industry, prolonged operating losses will weed out weak firms. Ideally, steel firms with the highest costs would be the first to close. But the real world is far from ideal. Many plants have closed and steel employment has plummeted. In the United States alone, over the last three years, 18 steel firms went bankrupt and about 23,500 workers lost their jobs.5 However, it is not production costs but rather market distortions that often determine the global “exit order” of struggling firms. Moreover, these distortions prolong the agony of failing firms by stretching the duration of depressed prices for the whole industry. One such distortion is cartel practices—private arrangements that enable some steel producers to maintain high prices in their home markets and sell abroad at low prices.6 Another distortion is public subsidies used to cover huge fixed costs (debt burdens, pension benefits, etc.).7 The United States is not the worst sinner when it comes to market distortions, but it is hardly free of guilt. Federal guarantee programs totaling several billion dollars have
absorbed the pension responsibilities of some bankrupt steel firms and staved off bankruptcy for others. As a result of these distortions, the least efficient firms are not necessarily the first to close their doors and the industry as a whole sheds its excess capacity at a very slow pace.
A New Round of Trade Actions These underlying forces—persistent overcapacity, rapid productivity growth, and slow demand growth—have again erupted, as in past episodes, in measures principally designed to limit steel imports—measures mounted by the steel industry, by Congress, and by President Bush. o o
o
The steel industry itself has launched numerous antidumping and countervailing duty actions. The Congress is debating the Steel Revitalization Act of 2001, a direct descendant of the failed Visclosky Steel Quota bill (H.R. 975) of 1999.8 President Bush has launched a new round of international steel negotiations and a Section 201 case.
Industry Answer: Antidumping and Countervailing Duty Actions One answer to market distortions is provided by the antidumping and countervailing duty laws. Indeed, 159 steel actions have either resulted in antidumping or countervailing duty orders or are under investigation. In fact, about a third of all antidumping cases launched in the United States between 1980 and 1995 were steel cases.9 Steel imports from nearly every country are now subject to one or more antidumping or countervailing duty actions (table 3). These actions cover 44 percent of finished steel imports from countries outside the North American Free Trade Agreement (NAFTA), and 11 percent of finished steel imports from Canada and Mexico. With such broad coverage, and more cases pending, are not unfair trade remedies adequate to deal with market distortions abroad? The answer is both “yes” and “no”.
The calculated consumer cost per job saved in the steel industry [by the Steel Revitalization Act] is very high, about $360,000 annually. “Yes”. In the 1970s, the United States dramatically expanded the coverage of its antidumping remedies, largely at the urging of the US steel industry, to encompass not only “classic” dumping (when an industry sells abroad cheaper than at home), but also “below cost” dumping (when an industry sells below its average cost). As a result, the antidumping laws are no longer focused on predatory pricing behavior, their original target.10 Instead, they catch both “fair” and “unfair” trade in a broad net that has become a “user-friendly” safeguard system. It is hard to argue that antidumping remedies do not adequately penalize exports from foreign firms that either operate at a loss or discriminate in their pricing policies. Indeed, most economists contend that US antidumping laws are too heavily slanted in favor of domestic petitioners.11 “No”. However easy it may be for domestic steel firms to prevail in antidumping cases, antidumping remedies cannot create competitive conditions in closed foreign steel markets, nor can they ensure that inefficient plants abroad are shut down. At most, antidumping remedies can shut particular foreign firms out of the US market. Other foreign steel firms may simply fill the market vacuum—substantially diminishing the protective effect of the antidumping order.12 Moreover, there is the serious problem of subsidies, especially when costs are absorbed by the public purse to keep a firm afloat. In the UK Bar case, decided in May 2000, the WTO Appellate Body held that subsidies received by erstwhile stateowned steel firms are effectively discharged once the firm is privatized. The US Commerce Department disagrees, and has not followed the UK Bar example in a number of other countervailing duty cases. In turn, the European Community is threatening to seek WTO permission to retaliate against the department’s refusal to follow the UK Bar precedent.13 Free traders say the United States should not complain when other countries export their steel at bargain prices, whatever the reason—dumping, subsidization, or just plain low costs. Many in Congress disagree. Urged on by the steel industry, they want to
erect more comprehensive barriers around the US market than those afforded by the antidumping and countervailing duty statutes.
Congressional Answer: The Steel Revitalization Act of 2001 The Steel Revitalization Act of 2001 (H.R.808 and S.957) is the congressional answer. This bill, which already has 219 cosponsors in the House, would restrict unfinished steel imports over the next five years to their average monthly levels between July 1994 and June 1997.14 This 36-month period was chosen because it precedes the “import surge” of 1998 and depressed conditions in the industry. Further, the Steel Revitalization Act requires foreign steel exporters to disclose extensive information on their pollution emission and wage levels. This information might eventually provide a springboard for congressional initiatives to impose minimum labor and environmental standards on imported steel.15
The real question facing President Bush and Congress is whether they would rather charge American consumers $3.5 billion annually for trade protection—and spend most of the money on steel investors and lucky importers—or only $350 million annually for meaningful benefits to displaced workers.
The bill authorizes additional Treasury guarantees for private loans to the US steel industry, enlarging the authority under the Emergency Steel Loan Guarantee Act of 1999 from $1 billion to $10 billion. Access conditions are also relaxed.16 Like kindred foreign subsidies, Treasury guarantees would prolong the life of failing steel firms and cause more distress to healthy producers. The Act imposes an excise tax on all steel for the creation of a health care trust fund to cover the “legacy costs” of unemployed and retired steel workers who are left in the cold by bankrupt firms. If necessary, the tax could rise to as high as 1.5 percent of
the value of domestic and imported steel. At current average prices and quantities (see table 4), the tax could theoretically rise to $700 million annually. Healthy US steel firms, such as US Steel, AK Steel, and Nucorp, are not enthusiastic about Treasury guarantees and taxes that would keep their inefficient competitors in business and make healthy firms bear the legacy costs of bankrupt firms. Meanwhile, the Bush Administration is rightly worried that the trade provisions of the Steel Revitalization Act would flatly violate US obligations under the World Trade Organization (WTO),17 invite foreign retaliation, and poison the atmosphere for WTO and FTAA talks.
The President’s Answer: International Negotiations and Section 201 On 5 June 2001, President Bush announced the administration’s own two-part steel strategy—international negotiations and Section 201. The President evidently wants to accomplish multiple objectives with his strategy: help distressed steel firms and workers; head off the Steel Revitalization Act; and, if possible, garner support for new “fast-track” negotiating authority (now labeled trade promotion authority) to credibly negotiate in the WTO and the Free Trade Area of the Americas.16 At an international level, the administration will open negotiations with trading partners to seek “the near-term elimination of inefficient excess capacity in the steel industry worldwide…” and to write “rules that will govern steel trade in the future and eliminate the underlying market-distorting subsidies that led to the current conditions…”
International negotiations The president’s first goal in international negotiations—to negotiate the closure of inefficient excess capacity—is bound to fail. Inefficient steel mills, almost by definition, are older mills with a long industrial history and many employees. Ultimately, market forces may force them to close, but it is virtually impossible to close them through a political negotiation. Any reasonable definition of “inefficiency” would rank some mills in
West Virginia, Pennsylvania, and Indiana below the cutoff mark. Imagine the grief facing the Bush Administration if it tried to persuade Senator John Rockefeller (D-W.VA), Congressman Phil English (R-PA), or Congressman Peter Visclosky (D-IN) to close down “their” mills. Public officials abroad will have similar problems. Just as international negotiations have never satisfactorily retired excess capacity in agricultural commodities—sugar, cocoa, coffee, and wheat—they will not succeed in steel. The president’s second goal in international negotiations—to write new rules governing future steel trade—has brighter prospects. Later in this policy brief we offer specific recommendations as to what those goals should be. Here we simply observe that new rules will not likely be agreed to in “stand-alone” talks; rather the best prospects are in the context of FTAA and WTO talks.
Section 201 Domestically, the president has requested the United States International Trade Commission (USITC) to launch a case under Section 201 of the Trade Act of 1974 to determine whether the steel industry has been “seriously injured” or “threatened by serious injury” from imports. The USITC carries out its analysis product category by product category, provoking a preliminary skirmish on the appropriate definition of steel “products”.19 Also, there’s the tricky matter of steel imports from Canada and Mexico, which enjoy preferential treatment under NAFTA.20 But in a Section 201 case, it makes absolutely no difference whether the imports are “fairly” or “unfairly” traded. All that matters is whether imports are a “substantial cause” of “serious injury” (or threat thereof) to the domestic steel industry. Under Section 201, the USITC must deliver its report and recommendation within six months, and the president then has 60 days to determine the remedies, if any.21 Under the Agreement on Safeguards adopted in the Uruguay Round, certain restrictions are imposed on Section 201 remedies, if they take the form of trade barriers.22 Among these provisions, quotas should not reduce imports below the level of the last three “representative years”, without a strong justification. Trade barriers should be progressively liberalized. Provided that these and all other provisions of the Agreement on Safeguards are met,
Article 8(3) allows the importing country (i.e., the United States) to avoid paying “compensation” to the exporting countries for the first three years of the safeguard measures.23 Section 201 does not require President Bush to impose steel trade barriers, even if the USITC finds that segments of the industry are “seriously injured” and even if it recommends trade barriers. Instead, the president can provide alternative forms of adjustment relief to firms, workers, and communities, and can condition relief on measures taken by the industry to downsize and become more competitive. In providing alternative relief, the president can draw on his existing statutory powers and budget pockets. By self-initiating the Section 201 case, however, President Bush created a strong political presumption that he would eventually restrict imports. But in our recommendations, discussed below, we urge the president to use his Section 201 powers creatively to fashion alternatives to oft-used, but very costly, trade “solutions”.
Strategic interaction While the domestic and international parts of the president’s strategy are conceptually distinct, they inevitably interact. If the USITC finds that imports cause injury in some product lines, it can propose remedies itself. However, as noted, the president can choose whatever remedies he wants to apply (consistent with the Agreement on Safeguards)—mixing trade barriers, industry assistance, and adjustment requirements. The president can select his remedies to shape developments both in Congress (on the Steel Revitalization Act and other trade issues) and in international negotiations. It should be pointed out that domestic and international political forces pull in different directions: in the congressional steel caucus, Section 201 remedies will be most acceptable as a substitute for the Steel Revitalization Act if they sharply restrict imports; but to coax foreign countries to reduce market distortions, it would be more useful to go easy on import barriers and instead emphasize industry assistance and adjustment requirements—the approach we advocate in this brief.
There is no perfect solution for the US steel industry’s problems, but there are certainly
better alternatives than top-heavy reliance on import barriers. The better approach would have two big components: wage insurance, and trade talks focused on rules to limit future market distortions in steel trade.
What Is Wrong with Trade Solutions? Most trade solutions to the problems of the steel industry— solutions in the form of import barriers—have major flaws. Import barriers are not very good at getting foreign governments to correct their own market distortions. This has been experienced not only by the steel industry, but also by numerous agricultural commodities (dairy, wheat, and corn), telecommunications, civil aviation, and financial services. In very few of these industries have US import barriers persuaded other countries to create more competitive markets. Instead, trade barriers have a way of inspiring “me-too” restrictions abroad, increasing the number of years (or decades) it takes for the industry to adjust worldwide. Moreover, US trade barriers on steel imports may cause more job losses in US steel-using industries—such as autos, heavy equipment, machinery, and construction—than the number of jobs saved in steel mills. The reason is that higher steel costs may force user industries to raise their prices and reduce demand. In a study commissioned by steel importers, François and Baughman (2001) applied a computable general equilibrium model to estimate that the quota protection envisaged in the Steel Revitalization Act would translate into two jobs lost in other industries for every job gained in the steel industry.24 An equally serious flaw is that import barriers cost US steel users a lot of money via higher prices. Higher steel costs are then passed on to the American public in a multitude of small markups for cars, washing machines, buildings, and practically everything else. Only a fraction of this money reaches distressed workers. Where does the rest of the money go? Mainly to the shareholders and creditors of steel firms, and to lucky importers.
Many Americans would be quite happy to pay an extra $50 for a car if all the money reached a steel worker’s pocket. But they would not be so happy if they knew that $34 went to shareholders, creditors, and lucky importers, while $6 was lost to pure inefficiency, leaving only $10 for steel workers. We illustrate these realities using a model that depicts the impact of the Steel Revitalization Act had it been in force in 2000.25 Table 4 presents the calculations. The cost of the Act to consumers would have been $3.5 billion. Of this amount, about $700 million would go to the 9,700 steel workers called back to the mills.26 About $400 million would be lost to pure inefficiency. This disappearing $400 million represents the costs to the US economy of not using labor and capital where they can produce output with greater value—industries like trucks, machinery, and construction. Domestic steel shareholders and creditors, together with importers, would capture the rest of the pie, about $2.4 billion, either directly through higher domestic steel prices, or indirectly by controlling the quota tickets for imported steel.27 Since the quota rents amount to $2.2 billion (see table 4), they deserve more explanation. As soon as imported steel is rationed by quantitative restrictions, the right to import steel—the quota ticket—becomes very valuable. In fact, it would be worth about $70 per metric ton of imported steel. Who gets the tickets? That depends on statutory language and lobbying skill. In practice, most of these valuable tickets would probably be awarded to traditional steel importers and to domestic steel firms that process unfinished steel (for example, turning slabs into cold rolled sheets) or augment their US-made steel products with imported goods. Of course, if policy were rational, quota tickets would be auctioned like broadcast bandwidth, not awarded to traditional importers and domestic steel mills. While this idea has enormous merit, lobbying forces have long kept it bottled up.28 Coming to the bottom line, the calculated consumer cost per job saved in the steel industry is very high, about $360,000 annually.29 This is several times the estimated annual compensation of $72,000 per steel worker (earnings of $44,000 plus fringe benefits of $28,000).30 To be blunt, capitalists gobble up the lion’s share of benefits from import barriers. That is the problem with trade solutions. In recent years, to be sure, some steel firms have suffered heavy losses. But a capitalist system works best when capitalists pay
for bad judgment or bad luck. In exceptional circumstances, the federal government may buffer them from losses. But after 30 years of “on-again, off-again” trade protection, it’s hard for the steel industry to argue that there’s anything exceptional about stiff competition from foreign producers.
Better Solutions: Wage Insurance and Sensible Trade Talks There is no perfect solution for the US steel industry’s problems, but there are certainly better alternatives than top-heavy reliance on import barriers. The better approach would have two big components: wage insurance, and trade talks focused on rules to limit future market distortions in steel trade. Wage Insurance Lori Kletzer and Robert Litan (2001) advocate an innovative approach—wage insurance—to address the core problem of distressed industries such as steel. The core problem is that displaced workers in the steel industry suffer a big loss in annual earnings when they lose a steel job and subsequently find work in another industry. Table 5 shows the statistical profile of displaced steel workers over the period 1979-99. Overall, comparing displaced workers in all manufacturing industries with displaced steel workers, the steel workers are a little older, a little less educated, have much longer job tenure, and are predominantly male. At the margin, this profile means that displaced steel workers take longer to find a new job and, when they find new work, take a bigger earnings hit than other displaced manufacturing workers. The average period of unemployment for displaced steel workers is 30.6 weeks (versus 18.2 weeks for all manufacturing), the average earnings loss is 34.8 percent (versus 18.2 percent for all manufacturing), and only 61 percent are reemployed in any job (versus 65 percent for all manufacturing). The idea of wage insurance is to encourage displaced workers to seek new employment, even at a lower wage, rather than collect unemployment benefits for the full 26 weeks normally permitted, or leave the workforce altogether. The Kletzer-Litan proposal has two components: o
The federal government would reimburse displaced workers a substantial fraction of the difference between
the wage earned on the old job and the wage earned on the new job once the worker is reemployed. o
Displaced workers would also receive a health insurance subsidy for up to six months as they search for new jobs.
These proposed benefits are in addition to unemployment insurance (UI coverage normally lasts for 26 weeks). Kletzer and Litan estimate the additional cost of wage and health insurance for all displaced workers would have been about $3 billion in 1999. The steel industry would be the ideal place to try out the wage insurance concept. Our proposal, sketched out in table 6, has somewhat more generous insurance parameters than Kletzer and Litan propose. In particular, we propose that wage insurance should cover 75 percent of the earnings loss for two years, rather than the 50 percent coverage recommended by Kletzer and Litan (2001). 31 In addition, we think the insurance program for the steel workers should address two unique features. First, some benefits should be paid to the 23,500 steel workers who were laid off in the last three years. We suggest 75 percent of the earnings loss for one year, for those who actually got new jobs (estimated at 14,000 workers). Second, something should be done about the so-called “legacy costs”. Legacy costs represent health and pension benefits that will be lost altogether if the responsible steel corporation goes bankrupt.32 In the case of pension benefits, the federal government already pays these costs under the Pension Benefit Guarantee Corporation (PBGC). But there is no safety net for health benefits. On this point, we agree with the Steel Revitalization Act: a mechanism should be created to provide a safety net for health benefits. Table 6 summarizes the arithmetic of our wage insurance and legacy cost proposal. The 23,500 steel workers displaced between 1997 and 2000 who have since been reemployed (estimated at 14,000 workers) would, on average, receive $11,500 of wage insurance. The total cost of this “make-up” wage insurance would be about $161 million. Over the next ten years, we project annual steel job losses at the rate that occurred between 1990 and 1999, about 5,000 jobs a year. With wage insurance, we assume that 75 percent of the displaced workers would get new jobs (compared with the historical average of only 61 percent). On average, each worker would be paid wage insurance of $23,000 for the earnings loss—
old steel job versus new job in ansurance costs would amount to about $86 million annually for the next ten years. Legacy costs must also be considered. Estimates are rough, but available figures suggest that legacy costs (both pension and health benefits) are $30 to $65 per ton of production by integrated mills (about 55 percent of total US steel production). In other words, total legacy costs in the integrated steel industry are between $1.7 billion and $3.6 billion. Assuming that 10 percent of integrated steel production closes annually for the next ten years (a worst case scenario), annual legacy costs would run between $170 million and $360 million. The existing PBGC would cover pension benefits, but health benefits would need to be covered by a new program. We assume the annual average total for legacy costs (both pension and health benefits) would run about $265 million. Box 1 concretely illustrates our wage insurance and legacy cost proposal, using the example of Weirton Steel. By our arithmetic, the lump sum outlay for wage insurance paid to workers displaced between 1997 and mid-2001 is about $161 million. The prospective other industry. Following this arithmetic, wage in annual expenditure on wage insurance and legacy costs for the steel industry would run about $350 million each year over the next ten years. In other words, the prospective budget expenditure is only one-tenth of the annual consumer cost of steel quotas. The real question facing President Bush and Congress is whether they would rather charge American consumers $3.5 billion annually for trade protection—and spend most of the money on steel investors and lucky importers—or only $350 million annually for meaningful benefits to displaced workers. Once the USITC finishes its work on the Section 201 case, President Bush could easily choose the sensible alternative, and ask Congress for the power and purse to implement a wage insurance program that genuinely helps displaced steel workers.
Sensible trade talks In addition to sensible decisions for the domestic steel industry, we need sensible international negotiations. These talks should be conducted, if possible, within the larger and far more important context of WTO and FTAA negotiations. Sensible steel talks should help the larger trade agenda, not hinder it.
International steel negotiations should seek to fulfill three objectives. The first is to write new subsidy rules that would reverse the WTO decision in the UK Bar case—the decision that effectively “wiped out” a history of subsidization (however long and deep), once a steel plant was privatized. Otherwise, China, Russia, Ukraine, and a few other countries will be tempted to modernize their steel industries on the public purse, “strip” the fixed costs from distressed firms, and privatize steel plants at a fraction of the historic cost. The outcome? A new wave of subsidized steel will flood world markets. In place of the UK Bar rule, new rules should attribute past subsidies to any successor firm. The second goal should be the creation of a reliable system within the WTO for affirmatively certifying, at the request of individual WTO members, that their steel markets are completely open to foreign producers, and that market access is not burdened by public or private restrictive practices. Restrictive practices include cartels of any sort, tariffs above a low rate, any limits on foreign investment, and public subsidies. Beyond the absence of restrictions, the certified country would need to positively demonstrate open market access, for example by evidence of open tendering by steel users, and substantial imports from foreign suppliers. The third goal would be to write a safe harbor, within the WTO Antidumping Code, for steel imported from certified “open market” countries—those with open trade and investment access. Imports from certified countries would be subject to antidumping orders only if they were dumped according to the classical definition—exported at a lower price than home market price. Steel sales below average costs would not, for that reason alone, be subject to antidumping duties. US steel firms often sell steel below average cost; there is no reason why foreign steel firms that operate in an open market should not be able to do the same. Reversing the UK Bar decision—together with affirmative “open market” certification and the antidumping safe harbor— would provide substantial incentives for countries to curtail the market distortions that have proved so troublesome for so long to the world steel industry. At the same time, this package of international rules should prove to be attractive to steel exporters and a positive incentive to launch WTO and FTAA negotiations.
It’s Up to President Bush! Over the next six months, President Bush has the power to skillfully conclude the steel initiatives he announced on 5 June 2001, and to make trade policy history. He can use the Section 201 case to pioneer a meaningful program of wage insurance for displaced steel workers. At the same time, he can shape steel negotiations in a way that points to the end of market distortions and the reform of antidumping laws, while giving a push to WTO and FTAA talks. Or he can fall back on the tried and failed “solution” of import barriers. When you think about it—and when the president thinks about it—there is only one right path.
References: American Iron and Steel Institute. 2000. December 2000 Selected Steel Industry Data. http://www.steel.org/stats/00december.htm, 18 June 2001. Bergsten, C. Fred, Kimberly Ann Elliott, Jeffrey J. Schott, and Wendy E. Takacs. 1987. Auction Quotas and United States Trade Policy. Washington: Institute for International Economics. Blonigen, Bruce. 2001. U.S. Antidumping Case-specific Data, 19801995. http://darkwing. uoregon.edu/~bruceb/adpage.html, 21 June 2001. Boltuck, Richard and Robert E. Litan, editors, 1991. Down in the Dumps: Administration of the Unfair Trade Laws. Washington: The Brookings Institution. Christmas, Ian. 2000. IISI Survey Reveals Renewed World Steel Consumption Growth. http:/www.worldsteel.org/trends_indicators/demand.html, 18 June 2001. Finger, J. Michael, editor. 1993. Antidumping: How It Works and Who Gets Hurt. Ann Arbor, MI: University of Michigan Press. François, Joseph F. and Laura Baughman. 2001. Costs to American Consuming Industries of Steel Quotas and Taxes. Washington: The Consuming Industries Trade Action Coalition Foundation.
http://www.citactrade.org/latest/citac_steel_quotas_and_taxes_04_30_01.pdf, 18 June 2001. Howell, Thomas R. 1998. The Trade Remedies: A U.S. Perspective. In Trade Strategies for a New Era: Ensuring U.S. Leadership in a Global Economy, eds. Geza Feketekuty and Bruce Stokes. New York: Council on Foreign Relations and Monterey Institute of International Studies. Hufbauer, Gary Clyde and Howard Rosen. 1986. Trade Policy for Troubled Industries. Washington: Institute for International Economics. Hufbauer, Gary Clyde and Kimberly Ann Elliott. 1994. Measuring the Costs of Protectionism in the United States. Washington: Institute for International Economics. Hufbauer, Gary Clyde and Jeffrey J. Schott. 1993. NAFTA: An Assessment. Washington: Institute for International Economics. Hufbauer, Gary Clyde and Erika Wada. 1999. Steel Quotas: A Rigged Lottery. International Economics Policy Brief 99-5. Washinton: Institute for International Economics. International Eco- nomics. June. International Iron and Steel Institute. 2001a. Crude Steel Production Per Employee. http://www.worldsteel.org/trends_indicators/figures_20.html, 18 June 2001. International Iron and Steel Institute. 2001b. Employment in the steel industry, 1974, 1990, 1996-2000. http://www.worldsteel.org/trends_indicators/figures_19.html, 18 June 2001. International Iron and Steel Institute. 2001c. World crude steel production, 1950 to 2000. http://www.worldsteel.org/trends_indicators/figures_1.html, 18 June 2001. Lindsey, Brink. 1999. The U.S. Antidumping Law: Rhetoric vs. Reality. Washington: Center for Trade Policy Studies, Cato Institute. Kletzer, Lori G. and Robert E. Litan. 2001. A Prescription to Relieve Worker Anxiety. International Economics Policy Brief 01-02. Washington: Institute for International Economics and The Brookings Institution. February.
Maddison, Angus. 2001. The World Economy: A Millennial Perspective. Paris: Organization for Economic Cooperation and Development. Messerlin, Patrick A. 1996. “Competition Policy and Antidumping Reform: An Exercise in Transition.” In The World Trading System: Challenges Ahead, ed. Jeffrey J. Schott. Washington: Institute for International Economics. Jeffrey J. Schott. 1994. The Uruguay Round: An Assessment. Washington: Institute for International Economics.tional Economics, 162-63. Sharkey, Andrew G. 2001. The CITAC Study: Flawed Model, Biased Conclusions. Washington: American Iron and Steel Institute. http://www.steel.org/news/oped/sharkey_citac.htm, 18 June 2001. Szamosszegi, Andrew. 2000. Cross-Border Cartels and the Steel Trade: The Impacts of the Europe, Japan Club. Washington: Economic Strategy Institute. United States International Trade Commission. 2001. USITC Interactive Tariff and Trade DataWeb. Washington: USITC. http://dataweb.usitc.gov/scripts/user_set.asp, 18 June 2001. US Department of Commerce, International Trade Administration. 2000. Global Steel Trade: Structural Problems and Future Solutions. Washington: US Department of Commerce. 18 June 2001. US Department of Labor, Bureau of Labor Statistics. 2001. Average Weekly Hours and Average Weekly Overtime Hours of Production Workers for the Manufacturing Industry, January 1999 to present. Washington: Bureau of Labor Statistics.
Box 1: How Our Arithmetic Works: A Hypothetical Case It is worthwhile applying the arithmetic of our suggested answer to one hypothetical case: Weirton Steel, the eighth largest steel producer in the United States, located in Weirton, West Virginia.1 This employeeowned mill, bought by the workers from National Steel in 1982, has 4,100 workers. Since 1905, Weirton Steel has been the dominant employer in Weirton, now a town of 20,000 people. Since 1990, output
per worker has doubled, owing in part to public guarantees for about $1.2 billion of investment in better productivity and a cleaner environment. Despite these improvements, depressed conditions have brought Weirton Steel to the brink of bankruptcy. Suppose Weirton closes, how would wage insurance work? Unemployment in two countries straddling Weirton is around 4 percent, but nearly all the jobs pay significantly less than the steel industry. Assume the average steel worker loses annual earnings of $15,300 when he takes a new job, and assume that wage insurance covers 75 percent of the earnings losses for two years. Wage insurance payments to these 4,100 workers would total about $94 million, assuming they all get new jobs. In addition, the federal government would pay “legacy costs”—health and pension benefits—of about $105 million. The pension component of legacy costs would be paid by the Pension Benefit Guarantee Corporation, while the health component would be paid by a new health care trust fund. The total federal cost of wage insurance and legacy costs would be about $200 million. 1. Most of the information and figures in this example are based on the report in the Financial Times, 13 June 2001, 8.
Notes: 1. Inflation concerns and a famous lawsuit, Consumers Union v. Kissinger (506 F.2nd 136) 1974, also contributed to the termination of the voluntary restraint agreements. 2. See Hufbauer and Wada (1999). 3. See International Iron and Steel Institute (2001b, 2001c); Maddison (2001). Steel demand was flat, despite economic growth, because many user industries engineered lighter products and products that substituted other materials for steel. 4. In Japan, Korea, and many European countries, it is legally or politically unacceptable to lay off workers; as a result, in these countries, wage payments almost amount to a fixed cost from the standpoint of the firm. However, from the standpoint of steel production technology, these labor inputs are still variable costs. 5. Sharkey (2001), 1, cites 23,500 jobs lost in the last three years. 6. See Szamosszegi (2000). 7. Likewise, when a firm discharges its obligations through bankruptcy, a similar subsidy occurs, but in this case the subsidy is paid not by taxpayers at large but by unlucky bondholders and employees. 8. See Hufbauer and Wada (1999).
9. See Blonigen (2001). 10. The vast majority of antidumping cases penalize imports sold in the US market below the foreign cost of production, and price discrimination is seldom an issue. 11. Among critical economists, see Boltuck and Litan (1991), Finger (1993), Messerlin (1996), and Lindsey (1999). Among analysts who endorse the US antidumping statutes, see Howell (1998) and Mastel (1998). 12. While domestic steel producers regret losing “market opportunities” to rival foreign steel imports, free traders regard this possibility as a saving grace of the antidumping laws. 13. The Wall Street Journal, 12 June 2001, p. A14, and Inside U.S. Trade, vol. 19, no. 24, 15 June 2001, 8. 14. For unfinished products, such as iron ore, pig iron, slabs and billets, the Act limits the tonnage of imports to the average monthly level between July 1994 and June 1997. For finished steel products, the Act limits the share of imports in domestic consumption to the level achieved between July 1994 and June 1997 (in other words, a greater quantity of finished steel imports would be permitted if US consumption rises). All quotas are established for individual products, apparently with no transfer between product categories. 15. If steel importers are required to report on foreign pollution and wage conditions, the law might also serve as a template for other US industries seeking “process protection”. 16. Because of tight access conditions, the entire $1 billion of available guarantees has not been used. 17. See Hufbauer and Wada (1999) for an analysis of the import quota provisions of the 1999 Visclosky bill (H.R. 975). The quota provisions in the Steel Revitalization Act of 2001 would, in a similar fashion, violate the WTO. The requirement of importers to report the pollution and wage conditions of foreign steel is new to the Steel Revitalization Act of 2001, and arguably creates a novel violation of the WTO. 18. Very commonly, either to launch or conclude international trade negotiations, past presidents have tossed protectionist bones to particular industries. This happened at the launch of the Kennedy Round in 1962 (new restraints on textile and apparel imports), at the launch of the Tokyo Round in 1974 (new antidumping provisions), at the conclusion of the Tokyo Round in 1979 (additional antidumping provisions and new quota protection for textiles and apparel), in the runup to the Uruguay Round (protection for autos, semiconductors, steel, and other industries), at the launch of the US-Canada Free Trade Agreement (excluding maritime and road transport) and at the conclusion of NAFTA (special deals for sugar and road transport). 19. For these and other details, see Inside U.S. Trade, vol. 19, no. 24, 15 June 2001, 4. 20. Under NAFTA, US steel imports from Canada and Mexico are
exempt from safeguard measures, unless the NAFTA member ranks among the top five foreign suppliers of the particular steel product to the US market, and unless those exports “contribute importantly” to serious injury or threat of serious injury. Overall, both Canada and Mexico rank among the top five suppliers of steel to the US market; however, they may not be among the top five for products where the ITC finds serious injury; and in any event their exports may not contribute importantly to the injury. To confuse matters further, in the Wheat Gluten case, the WTO ruled that if imports from Canada and Mexico are counted in an ITC determination of of injury, then the NAFTA partners must also be subject to the safeguards remedy. See Hufbauer and Schott (1993, 136-37) and Inside U.S. Trade, vol. 19, no. 24. 15 June 2001, 6. 21. The President can provide interim relief 90 days after the case is filed if the ITC finds “critical circumstances”. It does not appear, however, that the President’s Section 201 case will seek a finding of “critical circumstances”. See Inside U.S. Trade, vol. 19, no. 24, 15 June 2001, 4. 22. See Schott (1994). 23. In WTO terms, “compensation” traditionally refers to trade liberalization by the importing country that enables greater sales of other products sold by the exporting country. The idea is to maintain the “balance” of concessions that was in place (and “bound” under WTO agreements) prior to the safeguard measures. 24. The computable general equilibrium (CGE) model used by François and Baughman (2001) traces the impact of steel import barriers on downstream industries and the entire economy. They estimate that the Steel Revitalization Act would create 4,022 jobs in the steel industry, but destroy 10,306 jobs in steel-consuming industries (assuming full employment; see their table 3a). For a critique of François and Baughman (2001), see Sharkey (2001). Sharkey’s criticisms would also apply (if valid) to our partial equilibrium calculations. However, we do not think Sharkey’s criticisms are valid. 25. For these calculations, we use the computable partial equilibrium model, explained in Hufbauer and Elliott (1994). A partial equilibrium model only reflects events in the steel industry—not events in downstream user industries or the rest of the economy. In making our calculations, we assume that the quantitative restrictions required by the Steel Revitalization Act would apply with equal force (in terms of percentage reduction in steel imports) to unfinished and finished steel. 26. The figure of $0.7 billion reaching steel workers is based on the assumption that average compensation (earnings plus all fringe benefits) per steel worker is $72,000 annually ($36 an hour, for 2,000 hours of work per year—probably figures that are too high), and that 9,737 additional steel workers would be employed following imposition of quotas under the Steel Revitalization Act. Notice that the total producers
gain (table 4) is calculated at $0.9 billion. After subtracting $0.7 billion for additional compensation, the amount of producer gain left for the steel mills is $0.15 billion. Steel mills would probably get more— because the figure of $72,000 annual compensation is too high, and the mills would probably not call 9,700 workers back to the job. In addition, steel mills would capture a large chunk of the quota rent, discussed below. 27. The $2.4 billion is calculated as follows (from table 4): producer gain of $0.9 billion, minus the compensation cost of 9,737 additional steel workers, $0.7 billion, plus the quota rent of $2.2 billion. 28. See Hufbauer and Rosen (1986) and Bergsten, Elliott, Schott, and Takacs (1987). If policy was superrational, the funds raised by an auction of steel quotas would be dedicated to downsizing the steel industry and helping distressed workers and communities. In the practical logrolling that surrounds most protectionist policies, valuable quota tickets are instead used to reward supporters and to “buy off” losers. For example, some steel importers may bring less steel into the US market, but get a better price for each ton imported. 29. The consumer costs of $3.5 billion translates into $360,000 annually for each of the additional 9,737 steel industry jobs created by the Steel Revitalization Act (table 4). The cost per job, while extremely high, is significantly lower than the $800,000 figure cited in Hufbauer and Wada (1999). The reason is that Hufbauer and Wada (1999) assumed that the elasticity of domestic steel supply was 0.65 (based on a parameter for flat-rolled steel cited in Hufbauer and Elliott, 1994). For these calculations we assume that the supply elasticity is 3.00, a much higher supply response. Excess steel capacity is now larger than before, and mini mills have high importance in the industry as a whole, but low importance for flat-rolled steel. Both factors argue for a higher supply elasticity. 30. Based on the American Iron and Steel Institute (2000), steel workers had average hourly wages and benefits totaling $36 in 2000. Average hourly earnings were about $22, indicating average hourly benefits of $14. According to the Bureau of Labor Statistics (2001), the average annual hour figure (excluding overtime) for all production workers was 2,160. We assume 2,000 hours for steel workers (on account of slack industry conditions), suggesting average annual compensation (earnings plus benefits) of $72,000 and average annual earnings of $44,000. 31. To save space, we do not spell out the various antiabuse provisions that would need to accompany any wage insurance program. See Kletzer and Litan (2001). 32. Another legacy cost is severe environmental pollution at the sites of older steel mills. These mills were built and operated decades before the Environmental Protection Agency was created and strict environmental standards were imposed on plant operators—and all subsequent landowners. When these mills fail, the federal government will have to
assume responsibility for the cleanup costs. Otherwise, no sensible private corporation will come close to these sites.
B a c k t o: POLICY BRIEFS HOMEPAGE | TOP OF PAGE
Table 1 Employment in the steel industry (thousands) Country/group Austria Belgium Denmark Finland France Germanya Greece Ireland Italy Luxembourg Netherlands Portugal Spain Sweden United Kingdom European Union Australia Brazil Canada Japan South Africa South Korea United States Yugoslaviab Total for all listed countries
1974 44 69 2 10 158 232 0 1 96 23 25 4 89 51 194 998 42 118 77 459 100 n.a. 521 42 2,357
1990 21 26 1 10 46 125 3 1 56 9 17 4 36 26 51 434 30 115 53 305 112 67 204 69 1,388
n.a. = not available a. West Germany in 1974. b. Varying definitions. Source: International Iron and Steel Institute (2001b).
1999 12 20 1 7 38 78 2 0 39 4 12 2 22 13 31 280 24 59 57 208 55 64 153 15 836
Table 2 Overcapacity in the world steel industry, 1998 (million metric tons) Country or group European Union China United States Japan Russia and Ukraine Korea Brazil India Taiwan Canada Mexico Turkey Rest of world World
Production
Overcapacitya
157 115 98 94 68 40 26 24 17 16 14 14 95 776
n.a. = not available a. Overcapacity figures are approximate. Sources: For production: International Iron and Steel Institute, http://www.worldsteel.org/trends_indicators/countries.html. For overcapacity: US Department of Commerce, International Trade Administration, Global Steel Trade, 2000, http://www.ita.doc.gov/media/steelreport726.htm.
50 n.a. 15 17 25 n.a. n.a. n.a. n.a. n.a. n.a. n.a. 168 275
Table 3 US steel imports and antidumping actions, 2000-01 Country or group
European Union China Japan Russia and Ukraine Korea Brazil India Taiwan Canada Mexico Argentina Venezuela South Africa Australia Indonesia Malaysia Philippines Thailand Czech Republic Poland Romania Belarus Kazakhstan Latvia Moldova Rest of world Total
AD/CVD actionsa Importsb (as of June 2001) (in thousand metric tons, 2000) 36 7 20 8 14 10 6 14 4 6 5 1 6 1 4 2 1 3 1 2 4 1 1 1 1 0 159
6,510 1,531 1,981 4,590 2,457 6,402 1,012 1,158 5,079 3,097 423 1,864 1,275 825 379 135 9 494 87 320 442 96 474 230 357 2,186 43,413
a. AD/CVD (antidumping/countervailing duty) actions refer to the number of products subject to antidumping orders or investigations or suspension agreements. About 95 percent of actions are AD actions and only 5 percent are CVD actions. If an AD/CVD action applies to a product from more than one country, it is counted multiple times. b. Imports include all products under SITC code 67, “Iron and Steel”. Sources: AD/CVD actions: Fax correspondence with Willkie Farr & Gallagher, June 2001. Imports: US International Trade Commission (2001).
Table 4 Variables in the computable partial equilibrium model Notationa
Variables
Value
Elasticity parameters and market shares Value share of domestic demand Value share of import demand Elasticity of substitution between imported and domestic steelb Price elasticity of total demand Elasticity of domestic demand Elasticity of import demand Cross-price elasticity for domestic steel Cross-price elasticity for imported steel Elasticity of domestic supplyc
Sd Sm ó Edt E dd Emm Emd Edm Es
0.67 0.33 3.00 -1.00 -1.66 -2.34 1.34 0.66 3.00
Intercept terms For domestic demand schedule For domestic supply schedule For import demand schedule
lna lnb lnc
10.48 -1.29 10.10
Industry characteristics c.i.f. price of imported steel (dollars per metric ton) Landed price of imported steel (dollars per metric ton) Imported volume (million metric tons) Price of domestic steel (dollars per metric ton) Domestic production (million metric tons) Weighted average of imported and domestic steel prices (dollars per metric ton) Domestic steel employment (thousands) Estimated welfare effects from the Steel Revitalization Act of 2001 Consumer loss (billions of dollars) Producer gain (billions of dollars) Quota rent (billions of dollars) Efficiency loss (billions of dollars) Industry employment gain (number of jobs) Consumer cost per industry job gained (thousands of dollars) Landed price of imported steel (dollars per metric ton) Imported volume (million metric tons) Price of domestic steel (dollars per metric ton) Domestic production (million metric tons)
441 448 43.4 379 104.5 402 152 3.5 0.9 2.2 0.4 9,737 360 518 31.7 387 111.2
c.i.f. = cost, insurance, freight a. Notation is consistent with that found in Hufbauer and Elliott (1994), 36-37. b. François and Baughman (2001), 29. c. A high supply elasticity is adopted, based on Robert Crandall’s suggestion that mini mills can expand output at constant costs (conversation with Robert Crandall, June 2001).
Table 5
Profile of displaced workers in the steel industry compared with all manufacturing workers, 1979-99 Steel
Median age at displacement Mean education (years) Mean job tenure (years) Percent female Percent minority Share displaced from full-time jobs (percent) Mean jobless period (weeks) Percent reemployed For those reemployed, mean percent earnings loss Percent with: Earnings loss above 30 percent No earnings loss or earnings gain
All manufacturing
39.6 12.0 10.6 11.6 16.3 98.0 30.6 61.0
38.6 12.3 6.5 36.9 17.6 95.6 18.2 65.0
34.8
12.1
50.0 25.0
25.0 35.0
Source: E-mail correspondence with Lori Kletzer, June 2001; calculations from Current Population Surveys, Displaced Workers Surveys, 1984-2000.
Table 6 Assumptions for calculating wage insurance and legacy costs Average annual earnings of steel workersa Wage insurance make-up Steel workers displaced, 1997-2000 zzz Of which, estimated number reemployedb Annual earnings loss per reemployed steel workerc Wage insurance per displaced steel worker who was then reemployedd
$44,000 23,500 14,000 $15,300 $11,500
Total wage insurance for workers displaced, 1997-2000 (millions of dollars) $161 Wage insurance prospects, 2001-10 Steel workers displaced per yeare Steel workers who subsequently find jobs per yearf Wage insurance per displaced workerg Wage insurance cost per year (millions of dollars) Legacy costs Health care and pension benefits, per ton of steel produced in integrated mills Steel produced by integrated mills, 2000 (million tons) Total legacy costs (billions of dollars) Estimated legacy costs associated with closed capacity (millions of dollars annually) Total prospective wage insurance and legacy costs,h (millions of dollars annually)
5,000 3,750 $23,000 $86
$30-65 56 $1.7-3.6 $170-360 $351
a. The earnings figure does not include fringe benefits. By contrast, compensation is the sum of earnings and fringe benefits. Hourly compensation of all steel workers, according to the Bureau of Labor Statistics (BLS) is $36, while hourly earnings are $22. The calculation assumes 2,000 hours of work annually for steel workers (compared with BLS figure of 2,400 hours for regular and overtime work in all goods producing industries). b. Assuming 61 percent were reemployed; see table 5. c. The calculation assumes that the average displaced steel worker who subsequently finds a job incurs an earnings loss of 34.8 percent; see table 5. d. Assuming that make-up wage insurance pays 75 percent of the earnings loss for one year. e. Assuming 10 percent of integrated steel production closes each year between 2001 and 2010. f. Assuming that wage insurance raises the reemployment rate from 61 percent to 75 percent. g. Assuming that wage insurance pays 75 percent of the earnings loss for two years. h. Assuming wage insurance of $86 million a year and legacy costs of $265 million a year. Sources: E-mail correspondence with Lori Kletzer, June 2001; presentation titled “The Steel Problem: Causes, Effects, and Solutions,” Willkie Farr & Gallagher, June 2001, Washington.
