Industrial Policy: A Bad Idea Is Back

Manufacturing jobs cannot justify industrial policy.

By Scott Lincicome of Cato.

"In the wake of the COVID-19 pandemic and rising U.S.-Chinese tensions, American policymakers on both sides of the aisle have once again embraced “industrial policy” to fix perceived market failures and counter China’s growing economic clout. Perhaps the idea’s biggest fan is President Biden, who — much like his predecessor — has proposed a wide range of federal support for American manufacturers of “essential goods” and “critical technologies.” In the first half of 2021, Biden has pushed massive new subsidies (tax credits, grants, preferential contracts, etc.) for domestic producers of renewable energy technologies, electric vehicles, semiconductors, and “critical minerals,” as well as “Buy American” requirements for the construction materials and other goods needed to implement trillions of dollars in proposed infrastructure spending. Congress is eager to play along: both chambers are considering major legislation to subsidize American industrial research and development (R&D).

One could hardly blame the politicians if industrial policy advocates are to be believed. By their account, almost every major modern marvel, including basically everything involving computers and technology, all types of energy sources, the civil aviation industry, the pharmaceutical and biotech industries, as well as hybrid corn and lactose‐​free milk, is an “industrial policy success.”

However, few such innovations are the result of real U.S. industrial policy, which both advocates and critics historically understand to mean targeted and directed government interventions intended to achieve specific, market‐​beating industrial and commercial outcomes within national borders. The specificity of these targeted interventions is what makes them different from other kinds of broader, more general interventions.

Contra the cheerleaders, this excludes “horizontal” economic policies (patents, tax or trade liberalization, etc.) that apply to all sectors but might have indirect and disproportionate effects on certain industries, government funding for basic academic research or governmental goods (e.g., fighter jets) that unintentionally results in an innovation, and government contracts to purchase certain goods (e.g., the BioNTechPfizer vaccine) regardless of where or how it is made. That a random university researcher on a small federal grant stumbled on a new technology in an unrelated field does not “industrial policy” make.

THE FAILED HISTORY OF INDUSTRIAL POLICY
By contrast, real “industrial policy” has a long and ignominious history in the United States, one that honest supporters acknowledge has been riddled with “performance underruns and cost overruns,” owing to four main obstacles to these policies’ effective design and implementation.

First, past U.S. industrial policy efforts have often struggled to surmount F. A. Hayek’s knowledge problem, particularly for high technology goods. Centralized attempts to identify “critical technologies” in the 1990s, for example, failed in part because the government could not predict which technologies would be most valuable in the future or foresee how the marketplace would develop. Contemporaneous semiconductor and supercomputer protectionism picked the right industries but the wrong products and companies.

Second, even if U.S. planners can pick the right industries or products, politics thwarts their policies’ implementation — just as public choice theory predicts. Supercomputer policy in the 1990s, for example, was essentially aimed at supporting one politically powerful U.S. company, Cray, and ignored other American market entrants that offered different and arguably better products. Energy technology demonstration projects funded by President Barack Obama’s American Recovery and Reinvestment Act (ARRA) were dominated by unpromising (and now failed) clean coal and carbon capture projects, accounting for about five of every six dollars allocated, due in large part to the political influence of coal and ethanol producers and Obama’s affection for his home state of Illinois. Then, of course, there is Solyndra and the Obama administration’s green energy loan programs, which studies have repeatedly found to connect funding amounts to lobbying expenditures and campaign contributions, not scientific merit.

Most recently, Defense Production Act subsidies have gone to politically favored industries, such as shipbuilding, that have no connection to COVID-19. Even in cases where the connection to the pandemic is clearer, these sorts of industrial policy interventions have a poor track record. For example, certain vaccine supplies have been imperiled by Maryland vaccine manufacturer Emergent Biosolutions — a longtime government contractor that invested heavily in lobbying and consistently underperformed but was rewarded with an (as yet uncompleted) $628 million vaccine contract, perhaps because it had effectively “captured” the government agency awarding the contracts.

Emergent certainly isn’t alone. Politics routinely causes American industrial policies to suffer from a lack of discipline regarding scope, duration, and budgetary costs. Unlike private transactions whose success or failure is usually adjudicated — often ruthlessly — by the market, government industrial policies often live or die based on political considerations rather than their actual efficacy. Linda R. Cohen and Roger Noll documented such issues in their 1991 book, The Technology Pork Barrel, which examined six federal industrial policy programs originating in the 1960s and 1970s — the Supersonic Transport, the Applications Technology Satellite Program, the Space Shuttle, the Clinch River Breeder Reactor, Synthetic Fuels from Coal, and the Photovoltaics Commercialization Program — and found none truly successful. Four were “almost unqualified failures,” costing billions and crowding out more meritorious R&D projects yet enduring long after failure was established — a survival owed to political pressure and captured regulators.

The authors’ principal conclusion: “American political institutions introduce predictable, systematic biases into R&D programs so that, on balance, government projects will be susceptible to performance underruns and cost overruns.” Other programs — such as the Jones Act, the U.S. ethanol program, the U.S. antidumping law, and the clean coal megaprojects — permit the same conclusions. In each case, legislators and bureaucrats responded to years of failure not with reform or termination but with more funding or protectionism.

Third, industrial policies are often undermined by other government policies that have distorted the market at issue. Substantial ARRA funding for carbon capture, for example, was diverted to ethanol — a subsidized energy product with few if any environmental benefits but substantial political backing. Federal loan guarantee applicants’ compliance with the Davis‐​Bacon Act (mandating high wages and favoring politically connected labor unions), Buy American Act (mandating domestic content), and National Environmental Policy Act (requiring government review and approval of projects “significantly affecting” the environment) increased project costs, duration, and paperwork — and scuttled some projects altogether. New legislation to boost U.S. R&D spending and subsidize domestic semiconductor manufacturing has been larded with DavisBacon and Buy American rules, just as public choice predicts.

Fourth, industrial policies have costs far beyond the budget assigned to a specific project. Beyond the “seen” cost overruns (especially after considering federal borrowing costs), U.S. industrial policies create a host of “unseen” costs, such as indirect costs paid by others (e.g., consumers of tariffed goods), deadweight loss for the economy as a whole, opportunity costs, misallocation of resources, unintended consequences, moral hazard and adverse selection, and uncertainty inherent in a system dependent on politics, not the market.

Almost all these issues arose in the government bailouts of General Motors (GM) and Chrysler, which the Obama administration deemed an industrial policy “success” because they only “cost” taxpayers about $10 billion (the difference between the current‐​dollar value of funds the government “invested” and recouped). However, this rosy projection ignored not only the true interest‐​adjusted cost to taxpayers, estimated to be $14 billion, but also whether the $61 billion that the government invested could have been better spent at the time (for example, via direct payments to and retraining for autoworkers). Other neglected considerations include the long‐​term costs to GM and Chrysler because they were not reorganized via standard bankruptcy proceedings, the costs (e.g., lost business) incurred by Ford and other U.S.-based automakers who did not receive special treatment, and the costs to U.S. consumers and the economy because these companies’ better products and business models were not rewarded with additional business. On top of these are the moral hazards that resulted from encouraging the continuation of the companies’ and their union’s irresponsible practices, the costs to bond‐​holders and other investors who did not receive the fair value of their holdings, and the cost of uncertainty about whether political actors will again decide to intervene in the U.S. market and legal system, citing the bailout as precedent.

Industrial policy advocates’ responses to these criticisms are routinely deficient. Beyond the overbroad list of alleged successes, for example, rosy projections of direct economic benefits for recipient companies are rarely combined with empirical assessments of whether the U.S. economy overall would be better off due to the oft‐​claimed but usually unproven positive externalities, market‐​beating R&D spillovers, or faster economic growth. Furthermore, there is little consideration given for whether an industrial policy success would have occurred in a market without the supporting program at issue. Assessments of Department of Energy loan guarantee programs, semiconductor subsidies (SEMATECH), and cleantech startups funded by the U.S. Advanced Research Projects Agency‐​Energy (ARPA-E) all have found that government support mostly went to companies that could have obtained private funding or produced outcomes that the market could have provided (and did previously without government assistance).

Advocates also frequently claim that these economic and political costs are worth the expense if the project ultimately supports one big “winner,” such as Tesla Motors. However, even assuming that Tesla’s story is fully written and that it couldn’t have succeeded in the absence of government subsidies, this last‐​gasp argument must have limits: Would government backing of Tesla be worth a trillion dollars in waste, failure, and cronyism? Two trillion? Surely, some amount of money wasted on losers would be too much, even if the government picked one winner in the process.

“GOOD JOBS“
Finally, there is the small issue that the most common “problems” that industrial policies are supposedly needed to solve aren’t problems at all. As I explained in a recent Cato policy analysis (Manufactured Crisis: “Deindustrialization,” Free Markets, and National Security, Policy Analysis no. 907), for example, widespread claims of American “deindustrialization” are mistaken. Both U.S. manufacturing job losses and the sector’s shrinking share of gross domestic product primarily reflect long‐​term global trends shared by most industrialized nations and disconnected from specific federal economic policies, whether free market or interventionist.

At the same time, the U.S. manufacturing sector remains among the most productive in the world and has expanded since the 1990s — continuing earlier trends in output, investment, and financial performance. Between 1997 and 2018, real value‐​added for U.S. manufacturing overall and the durable goods sector in particular increased by 52.8 percent and 109 percent, respectively. Investment in the manufacturing sector — capital expenditures, R&D, and foreign direct investment — has been consistent and strong over roughly the same period. Indeed, real R&D expenditures more than doubled between 1999 and 2018, from around $127 billion to $274 billion. Pre‐​pandemic data and more recent news reports, moreover, show particularly strong investment in motor vehicles (especially electric vehicles and batteries), semiconductors, pharmaceuticals, and renewable energy products (i.e., the very industries that industrial policy fans in the White House and Congress now want to subsidize or protect).

Manufacturing jobs cannot justify a new U.S. industrial policy push either. Declines in manufacturing jobs are driven by secular trends shared by countries around the world, regardless of their industrial or labor policies. And as a 2013 Congressional Research Service report put it, “Although Congress has established a wide variety of tax preferences, direct subsidies, import restraints, and other federal programs with the goal of retaining or recapturing manufacturing jobs, only a small proportion of U.S. workers is now employed in factories.”

U.S. policy could in theory produce a onetime increase in overall manufacturing employment, but there is little reason to believe that such jobs would be sufficiently special or economically beneficial as to warrant government intervention, even assuming that such policies would be successful. For example, Cato’s Ryan Bourne showed in 2019 that U.S. manufacturing jobs are not significantly more stable or secure than jobs in other sectors, especially for lowskilled workers whose jobs have been disappearing for decades and are most exposed to automation and trade. Any additional increases in industrial productivity, moreover, would likely mean fewer jobs, a dynamic demonstrated by the last few years of increasing U.S. manufacturing jobs and sagging productivity.

The evidence that manufacturing provides “good jobs,” as President Biden and other politicians claim, is also thin. The manufacturing “wage premium” today is small if it exists at all. According to a December 2019 report by the Bureau of Labor Statistics, for example, by the end of 2018, “average hourly earnings of production and nonsupervisory workers in the total private sector had surpassed those of their counterparts in the relatively high‐​paying durable goods portion of manufacturing” (nondurables pay was even lower). Fortunately, middle‐​class compensation overall has not been stagnant, driven in large part by gains in services like warehousing and transportation. Median production and supervisory wages have increased by more than 30 percent since the early 1990s, and total personal compensation is up 61 percent.

American living standards cannot justify new U.S. industrial policies either. In terms of basic necessities like food, clothing, and home goods, Americans today are absurdly rich as compared to only a few decades ago. Cato’s Marian Tupy has shown that the average time that an unskilled American worker had to work to earn enough money to buy a long list of everyday items declined by 72 percent since the late 1970s, when manufacturing jobs were at their zenith. That means that for the same amount of work that allowed an unskilled worker to purchase one item in 1979, he or she could buy 3.56 items in 2019 on average. Tupy has found similarly impressive gains for food, helping to explain why food insecurity reached an all‐​time low before the pandemic hit.

Finally, there is little reason to believe that the industrial policy experiences of other countries, particular China, justify U.S. industrial policy. For one thing, most experts agree that differences in nations’ culture, economies, and political systems limit the extent to which perceived industrial policy successes can inform whether similar results are possible in the United States. In any event, the “successes” of countries like Japan, Taiwan, Singapore, and South Korea are routinely exaggerated, with studies showing that the nations’ impressive economic growth was, at best, mostly disconnected from industrial policy and, at worst, actually slowed by it. Meanwhile, any legitimate successes in these and other countries are more than offset by countless failures in Latin America, the UK, Europe, India, and — of course — the United States.

While China’s recent and troubling embrace of illiberalism and expansionism surely warrants criticism and attention, the view of Chinese industrial policy and China more broadly as urgent threats to the United States — one justifying a broad rejection of free markets and strong embrace of American industrial policy — is also misguided. China’s rapid growth is primarily owed to marketbased policy reforms (including World Trade Organization accession) following decades of self‐​imposed poverty, not industrial policy. Despite this “catch‐​up growth,” moreover, China still lags the United States in many important industries (e.g., semiconductors) and is struggling to advance.

Chinese industrial policy may have helped some other industries, perhaps even overtaking the United States in the process, but the cost of doing so was enormous, and those same policies have introduced distortions that could hamper future growth. China also faces several other challenges — an aging population, declining productivity, prioritization of moribund state‐​owned enterprises over private businesses and entrepreneurs, and increasing bureaucratization — that further undermine the all‐​too‐​common perception in the United States of China as an unstoppable economic juggernaut that, fueled by industrial policy, will inevitably overtake the United States.

In sum, industrial policy — properly defined — has an extensive and underwhelming history in the United States, featuring high costs (both seen and unseen), failed objectives, and political manipulation. Not every U.S. industrial policy effort has ended in disaster, but facts here and abroad demand that we rigorously question any new government efforts to boost “critical” industries and workers and thereby fix alleged market failures. Unfortunately, such skepticism is rarely applied.

The United States undoubtedly faces real economic and geopolitical challenges, but the solution lies not in copying China’s topdown economic planning on the grounds that the U.S. system is failing and that China is an inevitable economic power. Instead, American policymakers should lean into the things that made the United States a global leader to begin with: openness to foreign trade, workers, and investment; tax policies that avoid excessive burdens; flexible labor markets; stable monetary policy; and most notably, a lack of any grand industrial policy."