Just as it was not an accident that the U.S. was unprepared for a global pandemic, it is not an accident that America is overly vulnerable to global supply chain disruption. The two phenomena are related: Four decades of purposeful neglect was engendered by an economics priesthood which any kind of national industrial strategy taboo.

The conventional economic wisdom was that it doesn’t matter what a nation produces (“Potato chips, computer chips, what’s the difference?”); that only companies, not nations, compete; that maximizing global, not national, economic welfare, matters most; that America should focus on its comparative advantage as revealed by the market; that market forces are adequate to ensure U.S. economic leadership; and that any attempts at prioritizing sectors would result in protectionism, having the government “pick winners and losers,” or some other variant of what they called crony capitalism. These positions were nonsensical before the onset of COVID-19. Today, they make even less sense.

As policy makers consider how to reduce U.S. supply chain vulnerabilities, it is important to distinguish between two kinds of sectors and products. Some globally traded sectors are not all that technologically sophisticated. If the United States lost all capacity in such sectors, and for some reason later needed to ramp up production, it easily could do so because the production “recipe” and skills for making them are not all that advanced. There is currently a shortage of medical masks for healthcare workers, but it is being solved by quickly ramping up production domestically, and going forward it could be addressed by increasing the Department of Health and Human Services’ Strategic National Stockpile.

But other sectors and products are more complicated and harder to simply order to ramp-up. America likely can ramp up ventilator production, in part because there are companies in the U.S. that can make them but also because sophisticated producers, like Ford and GM, have the technical capabilities to shift production from cars to ventilators. But when it comes to 5G equipment, for example, restoring production to the U.S. would be extremely difficult, not only because some of the components are not produced here but also because we lack a company with the technological capability of doing so. In other words, once capability in advanced production industries is lost, including in suppliers, skilled workers, and testing labs, getting it back is extremely difficult. These complex industries often involve a complex network that has developed over decades; what Seymour Melman used to call an “industrial ecosystem.”

When large numbers of trees are cut down, at some point if the surviving forest is too small, the whole ecosystem collapses. The ecosystem must be large enough to encompass all of the niches that make it work. The same applies in an industrial ecosystem: if the machine tool industry is allocated in one country, the textile industry in another, the steel industry in another, the manufacturing process becomes much more delicate and inefficient, depending on a greater range of variables to succeed. Crucially, the speed with which innovations can ricochet around the system decreases significantly. When Japanese auto manufacturers have access to Japanese machine tool makers, they can get at the machine tool advances way before GM and Ford, and their proximity becomes a competitive advantage.

These insights help to explain why the view that it is okay to offshore hardware while retaining software is not only fatuous but damaging to long-term economic welfare. It’s easy to steal intellectual property; not so easy to steal a manufacturing facility.

So, what does this imply for policy? For one thing, the coronavirus must not obscure the fact that the reindustrialization of America has to extend beyond addressing an immediate health emergency.

Additionally, businesses and policy makers must recognize that the issues in today’s economy extend well beyond thinking of labor as the most important cost variable.

This is liberating: it means that reindustrialization can take a multitude of forms: for example, 3D manufacturing is a universalizing process: one machine capable of producing a dizzying number of products that already compete with some products from global supply chains. A 3D-printing company called Formlabs is working to receive a Food and Drug Administration (FDA) exemption on 3D-printed swabs used in COVID-19 testing. Other companies are working on 3D-printed masks, face shields, ventilator valves and even testing booths.

Another important area is nanotechnology. The ability to engage in precision design and manufacture at a molecular scale presents humanity with a new frontier of materials and products that have capabilities beyond what we currently have. Products ranging from glass to suntan lotion are already vastly enhanced by nanotechnology production. Qualities of human-made materials such as fireproof, bulletproof, insulation, or space travel are in the midst of being revolutionized. Medicines can be deployed through nanotechnology at the molecular level, vastly improving their effectiveness.

But we can’t rely on the free market alone to regenerate our industrial base. Much as we did throughout most of our nation’s history, it is time for a sophisticated and capable national industrial strategy. That means first and foremost to stop flying blind. Today the government lacks the basic knowledge of what the U.S. can and cannot produce. To remedy this Congress should establish a unit within the National Institute of Standards and Technology (NIST) within the U.S. Department of Commerce, charged with understanding U.S. capabilities and gaps in advanced technology production sectors in the U.S.

Second, it means recognizing and welcoming the role that the state plays in helping to drive intelligent industrial policy. The state’s role must extend beyond passively protecting property rights and establishing a framework for the operation of efficient, rent-free markets. Congress needs to step up to the plate with policies and programs to incentivize companies to reshore critical industry production to the United States. For example, at one time many drug companies produced critical drugs and active pharmaceutical ingredients (APIs) in Puerto Rico because Section 936 of the tax code provided them with generous incentives to do so. But when that provision ended in 2006 not only did it devastate the Puerto Rico economy, it led to a rush of factories moving to China, many induced by generous government incentives over there.

Congress should also expand funding for the BioFabUSA center, one of 15 federally and industry funded Manufacturing USA centers which focuses on the production process for biotech drugs. Better technology for making drugs can reduce the costs of production, making it more economical to produce domestically.

Similar approaches should be adopted in other critical sectors. For example, companies that design but don’t manufacture semiconductors (fabless companies) should be able to receive a tax credit for having their chips made in the United States. Congress should also increase funding for the Semiconductor Technology Advanced Research Network (STARnet)—which is a collaboration of universities providing exploratory research on semiconductor system and design technology.

We also need to remember why companies moved production to China besides their government’s unrepentant innovation mercantilism: low wages. But the COVID-19 pandemic might induce a reconsideration of other inputs that add to costs, including the health risk to foreign chains, against investing in high-tech production closer to their markets.

We should also recognize that many companies that embraced the soft option of offshoring labor ultimately created grave production weaknesses in the U.S. economy and stifled innovation. This is because such actions created disincentives to move up the technology curve. As Melman recognized, however, more competent managers realized that higher labor costs could be offset by using and investing in more and better machinery, which in turn would lead to a virtuous cycle of production: higher profits, which can lead to higher wages, leading to better machinery and organization of work, and higher skills.

A key way to do that is to spur the development and use of advanced manufacturing technologies like robotics, “smart manufacturing systems,” and additive manufacturing. These technologies will underpin efforts not only to automate many functions but to increase production flexibility so factories can more easily switch to producing such items needed during crises as ventilators.

One thing we cannot do is return to a pre-globalized economy or retreat to a form of mindless protectionism. But we can and should not only reduce our dependence on China but build up America’s overall industrial base and capabilities. Even if we do that, we will still not be self-sufficient.  Given the complexity of the existing and emerging technology system, that would be impossible even for an economy as large as that of the United States. But one hopes that collectively, America and its allies can re-establish leadership in key areas. To do so, the United States needs not just a national industrial strategy, but an allied industrial strategy to ensure, as a group, allied, democratic nations still have the ability to produce innovative products at competitive prices in a set of key areas while leaving us far more able to cope with sudden disruptions in supply so that we’re less vulnerable in the future. As Congress considers a fourth economic relief package, it should ensure that at least some components in it work to increase the resiliency of the U.S. advanced manufacturing economy.

Dr. Robert D. Atkinson is President of the Information Technology and Innovation Foundation. His Twitter handle is @RobAtkinsonITIF

Marshall Auerback is a market analyst and a research associate at the Levy Institute for Economics at Bard College (www.levy.org). His Twitter handle is @Mauerback