The pole dance is concluded. Jeff Bezos has chosen New York City and the Virginia suburbs of Washington DC for Amazon’s new headquarters. State and local governments will fork over more than $2 billion for the favor. Earnest policy folk are a bit shocked. The libertarian magazine "Reason" thinks Amazon would have picked these two rich and handsome cities anyway. New York and Virginia could have saved their money for worthier libertarian causes - cutting corporate taxes say - instead of flinging it solely at the wickedly dominant Amazon.
"Reason" also has an article sneering at the Taiwanese company Foxconn's plan to build a $10 billion advanced liquid crystal display (LCD) manufacturing plant in the township of Mount Pleasant, Racine County, Wisconsin. The magazine disparages as "a boondoggle" the financial incentives of $3-4 billion Wisconsin promised to Foxconn. It is upset by the "staggering" dimension of the "handout," amounting to "more than $230,000 per job created." It grows sarcastic that Foxconn may have to "import" Chinese staff because it is struggling to find American engineers. Elsewhere, it denounces the Foxconn project as exemplifying the Trump administration's "incoherent ideology of corporate populism," whatever that means.
“Reason” is right that the U.S. states shouldn’t offer handouts to attract investment. Most of Wisconsin’s financial incentives to Foxconn are tax exemptions linked to the company meeting targets for job creation and investment. Arguably the state loses nothing by not taxing a company that, without the exemption, would not have set up shop in the state in the first place. But that argument loses force when all the states compete for new investments, creating a prisoners’ dilemma or arms race. Take two states. Both would save money if neither offered incentives. But each thinks it could do a lot better if it attracted all the investment to itself, by offering incentives while the other does not. But when both offer incentives they end up worse off than if neither had: with no net gain in jobs or investment, just a transfer of tax revenue from states and their taxpayers to companies and their shareholders of up to $90 billion per year.
But there is still a case that, if not the states, the United States as a country should have a national industrial policy - should favor specific investments and economic activities, and promote specific directions for structural change because they would generate important net benefits for the country. There are both economic and national security reasons for a national industrial policy. Among economic arguments, the need for knowledge offers the strongest. A knowledge-based national industrial policy would, I think, support Foxconn, but not Amazon.
The political conditions for a bipartisan industrial policy may also be better after the mid-term elections. President Trump and congressional Democrats are less encumbered by libertarian economic ideology than many Republicans.
The economic case for a U.S. national industrial policy
Foxconn is an experiment. Along with products, jobs, wages, and profits, it will yield fruitful new knowledge - discoveries about how to rebuild an advanced electronics manufacturing industry in the United States. After decades in which that industry was crushed and outsourced to Asia, there is no textbook or formula for how to do this. The answers will be found only through learning by doing, by Foxconn's project managers puzzling out thousands of concrete problems they could not have foreseen, that are unique to the American environment. Creating an LCD supply chain or finding and motivating a skilled engineer is a very different problem in the U.S. than in Taiwan.
Foxconn’s discoveries will also profit other firms that want to invest in a potentially large new American advanced electronics manufacturing industry. That’s because knowledge is a public good. It is non-rivalrous – Foxconn’s knowledge will not deplete through being used in the way my eating a biscuit consumes the biscuit. Unlike the biscuit, knowledge remains available for others to use. Knowledge also tends to be non-excludable – Foxconn will struggle to stop much of its newly gained knowledge from spilling over to other companies for free, splashing along myriad channels, for example when its skilled workers go to work for competitors.
But Foxconn would likely not invest $10 billion in America in the first place if much of its hard-earned knowledge will spill over to its competitors for free, at least not without some public subsidy to offset the foreseeable losses. Then a new American industry that could have grown on the back of Foxconn would also fail to emerge.
That’s a market failure: a situation where the pursuit of private incentives leads to a less than optimal outcome for society. Market failures such as learning by doing, knowledge spillovers, and the public goods nature of knowledge are mainstream economics. Paul Romer won the 2018 Economics Nobel prize in part for showing how knowledge-based market failures depress economic growth. In a breakthrough 1986 paper, Romer analyzed how “long-run growth is driven primarily by the accumulation of knowledge by forward-looking, profit-maximizing agents.” Also, “the creation of new knowledge by one firm is assumed to have a positive external effect on the production possibilities of other firms because knowledge cannot be perfectly patented or kept secret.” And because private firms do not account for the benefit their knowledge accumulation bestows on other firms, there will be too little knowledge accumulated in aggregate. Growth will be less than socially optimal. An omniscient government could design a package of subsidies and taxes – an industrial policy – that incentivizes private firms to boost knowledge accumulation and growth. (Paul M. Romer. “Increasing Returns and Long-Run Growth.” The Journal of Political Economy, Vol. 94, No. 5. (Oct., 1986), pp. 1002-103.)
Knowledge-related market failures also help explain why international free trade does not necessarily benefit all countries - points worked out by such impeccably mainstream economists as Paul Krugman, Paul Samuelson, and William Baumol.
Unlike Foxconn, Amazon’s innovation - creating an online Sears Roebuck - is long since in the past. There will be few discoveries from replicating Amazon’s Seattle headquarters in New York and Washington D.C. There is no knowledge-based argument to subsidize Amazon. If anything, Amazon deserves tougher antitrust scrutiny on the dominant position it has secured in online retailing through being the first-mover in a platform industry with strong network effects.
The national security case
Michael Lind makes the national security case for an American national industrial policy. The United States is now in a Second Cold War for global predominance, primarily with China. Yet, “Under presidents of both parties, the Pentagon drew up war plans against China while the Commerce Department blessed the offshoring by U.S.-based multinationals of much of America’s industrial base to Chinese soil… America’s contradictory military and economic policies toward China have now collided… Sacrificing some American industries to keep smaller, dependent U.S. allies in the American bloc might have made sense in Cold War I. But with the perspective of history, we are likely to judge the Clinton, Bush 43, and Obama administrations harshly for their feeble responses to aggressive industrial competition on the part of China, a military and diplomatic rival.”
Lind continues: “In Cold War II even more than in Cold War I, the U.S. must play to its strengths, which are in technology and economics… Instead of letting state-capitalist nations such as China or profit-seeking multinationals restructure the U.S. economy to promote their own goals, the American republic needs a national industrial strategy of its own. To be precise, it needs to return to the time-tested and successful Hamiltonian industrial strategy of using whatever means are necessary — tariffs, subsidies, procurement, tax breaks, even overseas-development loans to countries that purchase U.S. manufactured exports — to ensure that strategic industries necessary to U.S. military power are introduced to America or remain here… And that means adequate and permanent production on U.S. soil, not just innovation in America and production elsewhere. Given a choice between leadership in manufacturing or innovation, any rational great power would choose manufacturing. It is much easier to steal foreign intellectual property than it is to build a factory and train a skilled labor force.”
What about government failure?
It is around this point (if not earlier) that libertarians, “free-market conservatives” and the Wall Street Journal will fall upon you crying “What about government failure?” Doesn’t government intervention always create more problems than the market failures it is supposed to fix? To which the only sensible answer is “No, not always. It depends…”
In “Normalizing Industrial Policy” Dani Rodrik summarizes the practical objections to industrial policy as twofold: “First there is the informational objection, which states that it is impossible for governments to identify with any degree of precision and certainty the relevant firms, sectors, or markets that are subject to market imperfections…This critique is often expressed by saying “governments cannot pick winners,” a highly effective conversation stopper. The implication is that in the absence of omniscience—that is, almost always—an activist government will miss its targets, support economic activities with no positive spillovers, and waste the economy’s resources. The second objection is that industrial policy is an invitation to corruption and rent-seeking. Once the government is in the business of providing support to firms, it becomes easy for the private sector to demand and extract benefits that distort competition and transfer rents to politically connected entities. Entrepreneurs and businessmen spend their time in the capital asking for favors, rather than looking for ways to expand markets and reduce costs.”
These are serious objections. But, says Rodrik, they could equally well be made in most other areas where market failures create a theoretical rationale for government intervention, such as macroeconomic stabilization, social insurance, infrastructure, education, and health. And yet we still judge that the benefits of government action in these areas outweigh the risks: the debate is typically not about whether but how government should act. Why should industrial policy be any different? Surely we can learn from our own and other countries’ experience to find some workable institutional solutions for an industrial policy that is, on balance, better than merely throwing up our hands in resignation and letting market failures have their way? Under the headings of ‘embeddedness,” “carrots and sticks,” and “autonomy,” Rodrik suggests practical ways in which governments can alleviate the problems that insufficient information and rent-seeking pose for industrial policy.
It’s the American Way
As Michael Lind suggests, the United States has a long history of government industrial policies that shaped economic development right from the Founding till today.
According to Berkeley professors Stephen Cohen and Brad DeLong:
“Repeatedly, government in the United States opened a new economic space, doing what was needed to enable and encourage entrepreneurs to rush into that space, innovate, expand it and, over time, reshape the economy. Each time, and there were many, this was done pragmatically. The choice of economic space seemed obvious, and the means—while powerful interests usually had a leg up—was never the bright idea of some smart economist or distinguished committee; it was never guided by ideology, whether pure or in the guise of theory. And each time in America’s long economic history—except for the most recent one, which was based on ideology rather than pragmatism—the results have been very positive indeed. (Stephen S. Cohen and J. Bradford DeLong. 2016. Concrete Economics: The Hamilton Approach to Economic Growth and Policy . Harvard Business Review Press.)
In the epoch from the Founding to the Civil War, the great American nationalist Alexander Hamilton and his successors used trade tariffs to promote America’s infant manufacturing industries. Military spending and government arsenals built up the high-tech skills of the day. Government actions enabled and subsidized infrastructure “improvements” such as canals. During and after the Civil War Lincoln and the Republican ascendancy put industrial development on overdrive with even higher tariffs, promotion of the railroads, the Homestead Act and creation of publicly funded land grant colleges. ‘Trust-buster’ Teddy Roosevelt made a necessary course correction with antitrust laws to sustain competition among America’s industrial giants. And while FDR was preoccupied with battling the Great Depression, “nevertheless, among [the New Deal’s] very many initiatives were several that clearly tried to open a new economic space for growth, such as the TVA and the huge program of dams in the dry West that opened vast areas for farming, industry, and even cities.”
After the Second World War Eisenhower and subsequent Cold War presidents reshaped the economy with massive highway and housing programs, financing of world-leading research universities, and development of new technologies through the huge defense budget. Even President Reagan deployed protectionist measures like “voluntary export restraints” and, intriguingly, agreed to a Defense Intelligence Agency proposal called “Project Socrates,” to use all the tools of statecraft to maintain American technological dominance.
In Cohen and DeLong’s telling, it was during the post-Cold War presidencies of the first Bush, Clinton, second Bush and Obama that American industrial policy went disastrously askew:
“Beginning in the 1980s, and continuing across a generation, the United States once again redesigned its economy. But this last time its choice of policy was not at all concrete. And it was not at all smart. For it was done very differently.
“For starters, the US government was not the only government targeting the shape of the US economy. On one side, the policies of East Asian governments—first Japan, then South Korea, and then, with quickly accelerating force and scale, China—pushed their economies onto a manufacturing-export development path. On the other, the United States accommodated their export-manufacturing push by pulling resources out of import-competing sectors and implementing a set of targeted policies to shift them elsewhere, into a new growth direction, toward what were supposed to be the higher-value industries of the future. It was ideology that told us these industries were out there… The invisible hand of economic magic was to pick up and realize what the stealth hand of politics had set in motion.
“The two teams, Asian and American, performed a kind of cosmetic surgery on the US economy—a body-sculpting. The American accommodation of the Asian export-manufacturing push—steel, shipbuilding, automobiles, machine tools, electronics—was sold as a liposuction, fat removal. It cut away a lot of muscle. Indeed, the weight of manufacturing in the economy dropped by 9 percent: from 21.2 percent of GDP in 1979 to 12.0 percent at the peak of the last business cycle in 2007…
“The Washington team performed the implant: deregulating both high and low finance; fueling real estate transaction processing; multiplying the share of economic activity devoted simply to the processing of health-insurance claims; and so forth. These sectors that were supposed to be the high-productivity, leading sectors of the economy increased by 5 percent of GDP…Today they account for a full one-fifth of the entire economy. This is pure economic bloat. Impure flab. Much of it, when all goes well, is close to a zero-sum activity: no net gain.”
Close to zero-sum at the best of times. And hugely negative when it’s not, as in the financial crisis.