The Jones Act fleet is not only shrinking but also increasingly decrepit because of artificially high replacement costs. Of the mere 98 ships in service, more than a third (34.7 percent) are past the age of 20, and a quarter of them (24.5 percent) are past 30. Studies also show that these old vessels are not only inefficient but dangerous.
With fewer (and older) ships, fewer shipyards, and fewer workers in the industry, the Jones Act has undoubtedly failed to achieve its national security objectives—a conclusion evident by the fact that the U.S. military during the Gulf War and thereafter rarely turned to the Jones Act fleet (and overwhelmingly relied on foreign‐built ships) to meet its sealift needs.83
Second, higher shipping costs caused by the Jones Act increase demand for alternative forms of transportation, including trucking, rail, and pipeline services, raising those modes’ rates and inflating business costs throughout the supply chain—thus affecting the operations and finances of nearly every business in nearly every U.S. industry, especially manufacturing. The Jones Act therefore disadvantages U.S. companies relative to their foreign competitors and consumes funds that U.S. households could spend or invest elsewhere in the economy (on more productive ventures). Indeed, some of those competitors, such as Russian gas producers that service Northeastern U.S. communities due to the artificially high cost of shipping liquified natural gas from Texas and Louisiana, are in hostile territories—another unintended consequence that undermines national security.
At the same time, heightened use of trucks and freight trains increases infrastructure and maintenance costs, as well as environmental costs (surface transportation emits more carbon than ships). It also raises safety issues (e.g., transporting toxic materials on U.S. highways) and increases traffic congestion—especially on highways running parallel to U.S. sea lanes—thereby generating opportunity costs from lost wages and lost output for American commuters. Finally, the Jones Act has been a persistent irritant to important U.S. trade partners, thus discouraging U.S. exports in those markets. These economic harms further undermine, rather than support, U.S. national security.
Machine Tool “Voluntary” Restraint Agreements, Subsidies, and Buy American Restrictions. Following a 1983 petition from the domestic machine tool industry under Section 232 and an affirmative “national security” determination by the Commerce Department in 1984, the Reagan administration concluded in 1986 five‐year “voluntary restraint agreements” (VRAs) with Japan and Taiwan to limit their exports of certain machine tools and requested that nine other countries limit their U.S. machine tool market shares to certain levels. The federal government—led by the Commerce Department and the DOD—simultaneously implemented a “Domestic Action Plan” to “assist, encourage, and fund a variety of research and development activities to help modernize machine tool and manufacturing technology.“84 This included the creation of the National Center for Manufacturing Sciences (NCMS), a public‐private consortium intended to “revitalize” the machine tool industry. Finally, Congress in 1986 and 1988 imposed Buy American restrictions on the DOD’s procurement of foreign‐made machine tools.85 The VRAs were modified in 1991 and extended through December 1993, when they expired. The NCMS still exists, though it has been expanded to cover all manufacturing operations in North America (as opposed to just U.S. machine tools producers).
These trade restrictions and subsidies proved unsuccessful. First, the VRAs cost U.S. machine tool consumers (i.e., other U.S. manufacturers) hundreds of millions of dollars per year but provided much smaller benefits to U.S. producers, resulting in an estimated annual net loss of $647,892 (in 2017 dollars) per job protected.86 Among the measures’ victims was one of the United States’ largest machine tool producers, Hurco, which sourced certain hardware from Taiwan and survived because it won an exemption from the government through 1990.87 The trade measures also failed to reverse import growth (which expanded by 2.51 percent per year between 1986 and 1990) or domestic job losses (which declined by 2.09 percent per year), while delivering substantial “quota rents” to foreign producers.88 Import growth was attributable to the exemptions, lax government enforcement, and the growth of unconstrained foreign suppliers—most notably Austria and China.89
Second, neither the trade restrictions nor the subsidies revitalized the domestic industry. A 1990 Government Accountability Office report found, for example, that American companies had thus far failed to meet the Reagan administration’s domestic market share targets in four of the six machine tool categories at issue.90 The ITC in 1993 found that domestic machine tools shipments actually declined by 11.7 percent between 1989 and 1991, while employment dropped by 9.8 percent.91 According to a 1995 article for the Philadelphia Inquirer, after years of subsidies and import protection, the U.S. machine tool industry still remained a “distant third” in global production—essentially tied with Italy but well behind both Germany and Japan; had only half the volume of sales (measured in constant dollars) in 1995 that it had 30 years earlier; and saw its workforce shrink from 108,000 in 1980 to 58,300 in 1995 (as well as experience stagnant wages).92
Finally, the programs were plagued with dysfunction. The Government Accountability Office report found that the Commerce Department’s methods for monitoring quota compliance and related import volumes suffered from a lack of documented procedures and data, inaccurate calculation methodologies, and reporting delays of five months or more. It also found problems with compliance and enforcement. Moreover, U.S. restrictions on supplies from major foreign producers likely fueled the growth of new market entrants, including China, which has since become a global leader.
In 1993, the George H. W. Bush administration quietly allowed the machine tools VRAs to fade away.93 Furthermore, neither the Buy American restrictions nor the NCMS’s continued operation ever revitalized the domestic machine tools industry, which remains a DOD concern.
Semiconductor Tariffs and Subsidies. National security also undergirded U.S. support for the semiconductor industry in the 1980s and ‘90s, but it also proved costly and unsuccessful.94 Government support was primarily implemented through two measures: the 1986 Semiconductor Trade Agreement (STA) between the United States and Japan and contemporaneous subsidies to support domestic semiconductor research and production. Each measure, however, generated meager benefits for specific U.S. firms while imposing substantial and unforeseen economic costs, leaving long‐term national security objectives unmet or even undermined.
The STA. Under the STA, the Japanese government agreed to stop its producers from “dumping” dynamic random‐access memory (DRAM) and erasable programmable read‐only memory chips—enforced through production limits and export restraints that kept prices above U.S.-determined levels—and to guarantee foreign producers 20 percent of the Japanese market. In exchange, the United States suspended ongoing antidumping and Section 301 (a U.S. law intended to police foreign trading practices that allegedly harm U.S. commerce) investigations of Japanese memory chips.95
The STA’s economic harms were significant. A 1994 Peterson Institute analysis found that in 1989, the STA generated a net national welfare loss of $974 million ($2.04 billion in 2020 dollars) and cost U.S. consumers over $525,000 ($1.10 million in 2020 dollars) per manufacturing job potentially saved. After the STA took effect, domestic semiconductor prices “skyrocketed,” and a “full‐fledged shortage of DRAMs was widely felt in the United States and Europe by early 1998.“96 As a result, U.S. semiconductor users, particularly up‐and‐coming computer manufacturers such as Apple that were dependent on DRAMs, were hobbled and less able to compete with Asian and European producers that could obtain cheaper DRAMs.97 As a result, the computer manufacturing industry shed one job for every U.S. semiconductor job supposedly gained from the STA.98 Increased DRAM prices also added almost $100 to the price of a personal computer selling for $600 or $700 in 1988.99
The STA also ended up helping Japanese semiconductor producers more than their U.S. competitors because the STA allowed the Japanese to charge higher prices in the United States and elsewhere. According to one Brookings Institution study, Japan’s manufacturers earned $1.2 billion in extra DRAM profits in 1988 alone and another $3–4 billion on all products in 1989—most of which was paid by U.S. consumers and computer manufacturers.100 Other studies found similar gains for Japanese producers, in part due to collusive behavior.101
U.S. producers, on the other hand, did not increase production capacity, despite artificially high domestic prices and U.S. government subsidies.102 All but one U.S. chip maker left the DRAM market within a decade, and the STA prevented neither industry recessions nor declining U.S. market share (which shrunk from 83 percent to 70 percent between 1986 and 1992).103 One reason is that U.S. firms found ways to circumvent the STA by importing not individual chips but rather assembled circuit boards that weren’t subject to the agreement.104 The benefits of the Japanese market share targets also proved illusory: although foreign semiconductor exports to Japan in 1992 hit the STA’s 20 percent market share targets, economist Craig Parsons found that this “achievement” was caused by broader macroeconomic trends, not the agreement itself.105 Other reports at the time noted that Japanese firms dumped the semiconductors that they were forced to buy into Tokyo Bay.106 Overall, “there is little consensus on whether the STA was effective in increasing the foreign market share.“107
As a result, “for most U.S. chip makers, the main impact of the price hikes was vastly greater profits strengthening their Japanese competitors.“108 Longer term, the STA actually helped “accelerate the entrance of Korean companies onto the world DRAM scene—as with Japanese companies, the supernormal profits that were obtainable in the years immediately after the [STA] allowed Korean firms such as Hyundai, Samsung, and LG to reap unexpected returns and gain a foothold at the lower end of the semiconductor technology ladder.“109 They are now market leaders.
Finally, the STA had significant political ramifications in the United States and abroad. It encouraged collusion among Japanese producers and restored the Japanese government’s control over the sector, with U.S. help. It led to the creation of a new and powerful lobbying group in the United States—composed of injured downstream user industries—that would go on to mold U.S. trade policy for decades.110 And it demonstrated the folly of U.S. security nationalism: just as the DOD was recommending action, American companies were exiting the DRAM market, having already discerned that their future was not in the “high‐volume, low‐profit commodity” but in advanced microprocessors, specialty chips, and design.111 As a result, U.S. Memories, a private consortium to expand domestic DRAM production, was “stillborn and collapsed in January 1990 owing to insufficient financial support and an unwillingness of other major buyers … to commit to future purchases.“112 Government planners foresaw none of this.
Sematech. Sematech (short for “semiconductor manufacturing technology”) was not a DRAMs project but instead a semiconductor R&D consortium funded jointly by private industry and the federal government—very similar to the consortium now proposed in the NDAA. As chronicled by Brink Lindsey in a 1992 piece for Reason, the primary impetus for Sematech was national security: only a month before the entity’s formation, a Pentagon‐sponsored study on “defense semiconductor dependency”—prepared by the Defense Science Board, whose advisory panel conveniently included Sematech member companies—concluded that “it is simply no longer possible for individual U.S. semiconductor firms to compete independently against world‐class combinations of foreign industrial, governmental and academic institutions.” The DOD therefore recommended $1 billion in government funding for a “Semiconductor Manufacturing Technology Institute.” Congress authorized $100 million a year for five years via the Defense Advanced Research Projects Agency.113
Lindsey showed how Sematech “confirm[ed] all the darkest suspicions of industrial‐policy critics.” In its first phase, “Sematech was able to borrow technology from private companies and reproduce manufacturing results that other private companies had achieved years before—and do it with taxpayers’ money”; and in its second phase, Sematech did some “useful work, both in evaluating new equipment and improving working relations between chipmakers and suppliers”—but it was work that, while it may have helped a few favored U.S. equipment suppliers, added “very little to what private industry is already capable of doing for itself.“114
Meanwhile, U.S. semiconductor firms were staging a major turnaround but did so by “ignor[ing] just about everything Sematech’s supporters have ever said about semiconductor competitiveness.” Instead, “American companies have been thriving in those supposedly marginal ‘specialty’ markets derided by the Defense Science Board.” Even worse, Lindsey explained how Sematech actually hindered the industry’s revitalization by “favoring older, more‐established companies [i.e., Sematech’s member companies] over innovative newcomers.” He finally debunked Sematech’s national security basis, noting that the U.S. military had ample domestic and foreign supplies of both commodity DRAMs and the chips most essential to U.S. weapons systems.115
Subsequent studies have confirmed Lindsey’s contemporaneous reporting. In a 1996 paper, for example, Douglas A. Irwin and Peter J. Klenow concluded that the “U.S. government’s contributions to Sematech do not induce more semiconductor research than would otherwise occur.“116 Even Sematech proponents Kenneth Flamm and Qifei Wang concluded that the consortium’s impact on member companies’ R&D expenditures was inconclusive and could in fact have been negative on net.117 That is hardly a ringing endorsement, given Lindsey’s account of other, noneconomic harms. In 2020, the Carnegie Endowment’s James L. Schoff included Sematech among the cautionary tales of American “technonationalism” in the 1980s: