The opportunity in renewed U.S. industrial policy

Joe Seydl, Senior Markets Economist
Jessica Matthews, Global Head of Sustainable Investing
Ian Schaeffer, Global Investment Strategist

Introduction: Behind the U.S.’s largest commitment to industrial planning since the Cold War

Several decades ago, economists and policymakers made a strong case for globalization: If countries focused on making the things they’re relatively better at, global trade would become more efficient, spurring faster economic growth for everyone.

That theory became reality. But it had two grim consequences—if not quite unforeseen, then unexpectedly severe: U.S. labor participation fell off a cliff, and many blue-collar workers in American factories, particularly men, lost their jobs. Depression and death rates climbed, especially in the American Heartland.

Recently, another cost of what we think of as hyper-globalization has become clear. U.S. national security has been compromised, as companies have sent overseas strategically critical technology, including electronic and green energy components and military hardware. The deterioration of U.S. industrial capacity can hardly be overstated: Services now comprise nearly 80% of the national economy.¹

The federal government is responding. Despite political polarization in Washington, lawmakers passed three major initiatives intended to raise the country’s industrial capacity. Together, the 2021 Infrastructure Investment and Jobs Act (IIJA), the 2022 Inflation Reduction Act (IRA) and the 2022 CHIPS and Science Act (CHIPS) represent the largest U.S. commitment to industrial policy since the Cold War.

Federal industrial policy has sometimes failed. Yet the government’s heavy hand has at other times created profound and lasting changes, particularly when it established a new system with self-reinforcing demand. Previous successes have relied exclusively on public funding. Today’s grand national experiment will see if public funds can be used to stimulate private capital holders so that, together, they can create a self-perpetuating market.

We think the new industrial policy, which seeks to decarbonize the U.S. economy and strengthen American industry, has the potential to reverse some of the devastation, while creating exciting, long-term investment opportunities.

Renewed U.S. industrial policy supports our view that in the current market cycle (and perhaps for several years after), “real economy” stocks should outperform the growth stocks that dominated the 2010s. We like real economy equities that tend to offer high free cash flow and dividend yields (and companies improving shareholder returns; for example, through dividend growth).

Among energy transition stocks, we favor manufacturers of clean energy technology over site developers and utilities, as we’ll explain. We also like investing in energy transition infrastructure, such as battery storage projects. Among corporate recipients of CHIPS Act benefits, we prefer semiconductor companies that have close ties to the U.S. military. Military spending, currently at a secular low relative to GDP, may rise in the years ahead.

Part 1: Grim consequences: The path to industrial policy today

Hyper-globalization’s impact on the American worker

Policymakers and economists understood from the start that globalization would lead to some job loss and economic hardship for certain communities. However, they thought economic gains would far outweigh losses and the bigger pie would compensate those affected, through worker retraining and other types of redistribution.

The overall efficiency of the global economy did improve—and the U.S. consumer benefited spectacularly. By one estimate, for every U.S. manufacturing job lost due to soaring trade with China in the 2000s, the U.S. consumer gained about USD100,000, or USD200 billion in total.²

However, there were two practical problems with the way globalization unfolded. The first problem was that globalization did not happen gradually, and its effects were not absorbed smoothly. In the 1970s and 1980s, manufacturing job losses due to globalization (and automation) were steady. Even in the 1990s when globalization’s pace was accelerating, factory payrolls held up. But in the 2000s, after China joined the World Trade Organization, globalization accelerated rapidly. The camel’s back broke, and manufacturing payrolls fell off a cliff.

A total of 5.7 million factory jobs were lost from 2000 to 2010—nearly 10 times more than during the previous 30 years, from 1970 to 2000. Redistribution and worker retraining can be effective if change is slow and predictable, but in the event of a shock like this, there is only so much worker support systems can do.³

Workers, especially men, had a hard time adjusting to the shock of massive manufacturing job losses. Social, health and economic pathologies have bloomed as a consequence.⁴

The share of working-age men not in the labor force rose from 5.7% at the start of the 1980s to 11.4% as of 2022. But the pain wasn’t smoothly distributed. It was concentrated in the Eastern Heartland, states east of the Mississippi River from Mississippi to Michigan (excluding the Atlantic Coast). That region’s mortality rates soared above the coasts’ and the Western Heartland’s.⁵

Had globalization not occurred as an abrupt shock but progressed more steadily and predictably, it is safe to say that less of these pathologies would have emerged.

How outsourcing opened national security vulnerabilities

Globalization’s impact on national security also didn’t exactly go according to plan. Its architects assumed that after the Soviet Union collapsed, the global economy would reap a “peace dividend” indefinitely. And so it largely did, in the decades after the fall of the Berlin Wall in 1989.

But war, and the threat of it, eventually reemerged. Russia annexed Crimea in 2014, then invaded Ukraine in 2022. China in 2021 tested a hypersonic missile system capable of evading U.S. defenses.⁶ Governments started to question globalization through a national security lens. More specifically, governments in the advanced economies began to realize GDPs based primarily on services are a vulnerability.

Emmanuel Todd, an influential French historian and sociologist, recently wrote, “Western power, measured on the basis of GDP, is fictitious.”⁷ We think that is an overstatement, but helpfully provocative. While the U.S.’s GDP may outstrip China’s and Russia’s, it has a low non-services share of GDP and fewer military personnel.

The aforementioned U.S. social and health pathologies also create a national security vulnerability. A 2020 Pentagon study found that more than three-quarters of Americans aged 18–24 were medically unfit for military service. Obesity was the most widespread health problem.⁸

Part 2: How industrial policy can succeed

What is industrial policy? The case of the U.S. interstate highway system

Industrial policy can be defined as “an effort by a government to change the sectoral structure of production toward sectors it believes offer greater prospects for accelerated growth [or for national security reasons] than would be generated by a typical process of evolution according to comparative advantage.”⁹ In other words, it is an intervention by the government to go against market forces.

Recent policy initiatives are intended to reduce the services share in GDP and raise the nation’s industrial capacity. Politically, there is now a bipartisan consensus that this is a necessity.¹⁰

The history of U.S. industrial policy includes numerous examples of failure. The Synthetic Fuels Corporation, set up in 1980, failed spectacularly in reducing dependence on foreign sources of oil. But done right, industrial policy can create profound change that reinforces itself over time.

The largest and one of the most successful examples in U.S. history is the Federal-Aid Highway Act, passed in 1956 to build the interstate highway system. It was designed to connect cities and towns, lift productivity and, as today, improve national security. Mindful of the Cold War, the network of highways would allow military resources to be rapidly mobilized.

The government played a huge upfront role. By the early 1960s, U.S. public spending on highways totaled 1.8% of GDP. But as more drivers took to the road in the late 1960s and 1970s, the Highway Trust Fund, set up to maintain the roads, grew. The fund derives its revenue from gasoline and diesel fuel sales taxes. Over time, its balance started to rise. By the 1980s, public spending on highways fell to around 1% of GDP, financed entirely by the system’s own revenues from the public.

The fund’s surplus eventually eroded after the 2008 financial crisis. People were driving less (fewer tolls collected) and cars’ fuel efficiency improved (generating less in gas taxes). Highway maintenance costs kept rising. Since 2008, the Treasury Department has plugged the fund’s shortfalls with fiscal transfers.

Whatever its later financial troubles, the government, playing a heavy hand upfront, set off a system that became self-financing with its own internal source of demand. And still today, the interstate highway system is profoundly important to U.S. transportation infrastructure: While accounting for only 1% of all road mileage, it accounts for 24% of all vehicle miles driven in rural America.¹¹

The U.S. government is attempting to do something similar today when it comes to decarbonizing the economy, but rather than relying entirely on public financing, it aims to provide a set of carrots that stimulate private sector investment.

The Inflation Reduction Act: Carrots for Players in the Energy Transition

The Inflation Reduction Act is a misnomer: IRA is an energy bill. It, and CHIPS, are designed to use public funds to stimulate private capital investment, using subsidies and tax credits. These enticements to private-sector partners taper off, and then fully phase out once renewables are cost-competitive without government support, and emissions targets (reducing electricity sector emissions by 75% by 2032¹²) are met.

IRA tax credits and subsidies are designed to push down the cost of renewable energy production and, in turn, a competitive power market should push down consumer energy prices over time, spurring a faster adoption of renewables.¹³ Economies of scale and technology learning curves have already resulted in sharp declines in the cost of wind and solar power. By 2030, Bloomberg New Energy Finance (NEF) projects that the cost of generating wind and solar power will be well below average wholesale power prices as a result of the IRA investment/production tax credits, which are expected to drive net costs (for solar) to just 15USD to 25USD per MWh (chart below).¹⁴ If that’s the case, then there should be substantial continued investment in wind and solar generation, constrained only by the speed with which additions can be integrated into the grid.¹⁵

The increased expected cost competitiveness of renewables resulting from the IRA already appears to be accelerating private-sector capital spending commitments. In the subsequent eight months since legislators signed the bill into law in August 2022, companies announced more than USD150 billion in private capital investment for U.S. utility-scale clean energy projects and manufacturing facilities—surpassing total investment in these projects commissioned from 2017–2021.¹⁶

How can industrial policy potentially impact investors? Changing profitability expectations

We think the energy transition that the IRA seeks to catalyze is likely to play out more visibly at the company level, as its incentives take hold and change profitability expectations.

The IRA offers manufacturers tax credits.¹⁷ That’s in part why the largest profit margin increases are likely to accrue to manufacturers. Single-stock analysts at J.P. Morgan Research are starting to incorporate these tax credits into their modeling of financial statements. Their initial conclusion: Expect big effects. They expect that by 2025, the IRA can double the profit margin for a representative renewables technology manufacturer versus the no-IRA baseline.¹⁸

Higher expected profitability is likely to drive up capex as firms begin to see more clearly their likely returns from newly expanded production. The end result: expanded renewables generation capacity relative to the pre-IRA baseline. Bloomberg NEF estimates that by 2030, there will be about 88,000 MW of additional utility-scale solar and wind capacity as a result of the IRA. That’s roughly the capacity of the Texas electrical grid in 2020.¹⁹

Project developers, however, may not experience spectacular profit gains. The IRA’s incentives for developers seek to remedy the geographic inequalities that grew during hyper-globalization. Under the IRA, project developers are entitled to multiple “stackable” investment tax credits, if their energy production is sited in a low-income area and/or an “energy community”²⁰—that is, in the Eastern Heartland areas where shuttered coal and brownfield sites stand to benefit.

Another reason manufacturers’ profit margins are likely to increase more than developers’ is that under the IRA, even some of the developer credits will likely accrue more to manufacturers because of the market structure for critical components. For developers to receive the “domestic content” credit, for example, they need to purchase components from U.S. manufacturers of renewables technology, and there are relatively few of them today. Being highly sought after can give these manufacturers a new premium in pricing power that they may pass on to their customers, likely eroding the tax credit for developers—at least until domestic manufacturing capacity expands.²¹

The potential macroeconomic effects of renewed U.S. industrial policy

We do not expect large effects on trend GDP growth over the next decade. These laws’ economic multipliers are expected to be mostly low. The IRA’s intent is to reduce greenhouse gas emissions, and to the extent this is achieved, it will be a non-GDP benefit for U.S. residents and the global population.

Nor is the CHIPS Act meant to raise GDP. The United States is a high-cost destination for semiconductor chip production. (An executive from Taiwanese chip maker TSMC has said in an earnings call that production in the United States costs four times more than in Taiwan.²²) It is highly inefficient to reshore the industry, especially considering semiconductor chips are among the most easily transportable goods in the global supply chain. Yet CHIPS may increase domestic semiconductor-chip security, serving as a hedge against any future supply disruptions—a benefit, but not to GDP.

For these reasons, we wouldn’t be surprised if the economic multipliers from the IRA and the CHIPS Act were close to zero. To be sure, these bills may shift the composition of investment demand, away from housing toward manufacturing (which could be problematic, given the chronic U.S. housing shortage).

On the other hand, the third new law, the Infrastructure Investment and Jobs Act of 2021, is expected to boost GDP through traditional productivity effects. But the addition is too small to meaningfully accelerate trend growth.²³

New laws are expected to have little effect on long-run inflation expectations, but may spur more short-run variability

These three new laws aren’t likely to change long-run expected inflation, which is anchored by monetary policy and the effect it has on the labor market (as the fastest pace of rate hikes in 40 years has reminded us). The laws may well change short-run inflation variability, however, by increasing resource nationalism globally; barriers may create unexpected pressure on commodity prices in the years ahead.

To see what that looks like, consider the inflation shock the pandemic brought about. Remarkably, it didn’t alter long-term inflation expectations, yet it created tremendous short-run variability (chart below). The energy transition may cause similar inflation dynamics in the future, although likely not as dramatic (COVID hit suddenly, while the energy transition will spread over years and decades).²⁴

Nevertheless, a more variable inflation outlook may mean the Treasury market’s current pricing of inflation risk premia is too low.²⁵

Conclusion: Investment implications

Renewed industrial policy supports our expectation that the current market cycle is likely to be a “real economy” cycle, after the “growth cycle” of a decade ago. Within this real economy thesis, we like equities that can potentially offer high free cash flow and dividend yields, including those companies shifting their focus toward shareholder returns (e.g., dividend and buyback growth). When it comes to the energy transition specifically, we favor the manufacturers of clean energy technology over both the developers and the utilities. (Many of the IRA’s subsidies for developers will likely accrue more to domestic manufacturers.)

When it comes to energy-transition commodities, we note that China currently dominates the supply chains for many critical minerals related to renewables, leading the world in production of aluminum refining and smelting (67% of global capacity), lithium and cobalt refining (80% and 66%, respectively) and graphite production and refining (about 80%)—along with many other critical minerals.²⁶ We also note that while the IRA offers U.S. consumers a tax credit for buying an electric vehicle, half the credit disappears if vehicles do not meet the requirements for critical minerals to be extracted, processed and/or recycled in North America, or in a country with which the United States has a free trade agreement.²⁷

In light of this, exposure to commodities necessary for the transition to renewables may offer attractive investment returns less correlated to a typical portfolio of stocks and bonds. Another favorable way of gaining exposure to the energy transition outside of stocks and bonds would be direct exposure to energy-transition infrastructure projects. Two attractive examples are battery storage and recycling.

When it comes to the CHIPS Act’s implications, we prefer U.S.-domiciled foundries and integrated device makers (IDMs) that have close ties to the U.S. military—in other words, investments in technologies that have so-called “dual use” capabilities. Military spending is currently secularly low relative to GDP, and we think there is a good chance it may rise in the years ahead, given how the Ukraine War is rapidly depleting NATO stockpiles,²⁸ and the rising U.S.-China military tensions in the South China Sea.

In sum, while the new industrial policy was designed to decarbonize the U.S. economy and reindustrialize the American Heartland after the devastation wrought by hyper-globalization, it also includes important incentives to stimulate private investment. We think it will usher in many interesting long-term investment opportunities.

¹World Bank. Data as of 2023.

²Measured for 2000–2007. The effect comes from lower prices (relative to the counterfactual) from greater trade with China. The price effect was found to be about two percentage points, so multiplying by nominal GDP of $10.3 trillion in 2000 yields a gain to U.S. consumers of about USD200 billion, or USD100,000 per manufacturing job lost due to offshoring the two million total jobs lost. Xavier Jaravel and Erick Sager, “What are the Price Effects of Trade? Evidence from the U.S. and Implications for Quantitative Trade Models,” Bureau of Labor Statistics Working Paper 506, September 2018.

³Also, the United States offers fewer support systems than to most other OECD countries. Lane Kenworthy, Social Democratic Capitalism, Oxford University Press, 2019.

⁴David H. Autor, David Dorn and Gordon H. Hanson, “The China Syndrome: Local Labor Market Effects of Import Competition in the United States,” American Economic Review 103, no. 6, October 2013; Adam Dean and Simeon Kimmel, “Free trade and opioid overdose death in the United States,” SSM—Population Health 8, August 2019.

⁵Benjamin A. Austin, Edward L. Glaeser and Lawrence H. Summers, “Jobs for the Heartland: Place-Based Policies in 21st Century America,” National Bureau of Economic Research Working Paper 24548, April 2018.

⁶The missiles are said to evade detection by flying at low altitudes over Antarctica. General Mark Milley, Chairman of the U.S. Joint Chiefs of Staff, said China’s test launch was “very close” to a “Sputnik moment.” David E. Sanger and William J. Broad, “China’s Weapon Tests Close to a ‘Sputnik Moment,’ U.S. General Says,” New York Times, October 27, 2021.

⁷“Western economic power, measured on the basis of GDP, is fictitious,” International Affairs, September 25, 2022; Emmanuel Todd, “Face à la Russie, l’Occident croyait voler à 10 000 m d’altitude. Il vole à 350 m,” Marianne, July 9, 2022.

⁸2020 Qualified Military Available Study, U.S. Department of Defense Office of People Analytics, 2022.

⁹Gary Clyde Hufbauer and Euijin Jung, Scoring 50 Years of U.S. Industrial Policy, 1970–2020, Peterson Institute for International Economics, 2021.

¹⁰According to William Galston, Senior Fellow at the Brookings Institute and former Deputy Assistant to President Clinton for Domestic Policy, “As recently as a decade ago, ‘industrial policy’ was anathema to elites in both parties. Along with many other economic issues, they have changed their stance on this matter, and they will enjoy public support if they choose to invest in the technologies that will define the economic and national security competition with China in coming decades.” Brookings Institute, March 2021.

¹¹Matthew Turner, Gilles Duranton and Geetika Nagpal, “Transportation Infrastructure in the U.S.,” NBER Working Paper 27254, May 2020/.

¹²The reduction goal is from 2022 levels.

¹³Prior to the IRA, the adoption of renewables was playing out slowly in the United States, rising from just under 10% of total electricity consumed in the early 2000s to 21% by 2022. Energy Information Administration, Monthly Energy Review.

¹⁴It should be noted that the chart compares expectations of costs in 2030 to today’s market determined 2030 forward power prices. The logic is that market-determined prices today should be based on current technologies, which are largely non-renewables, so effectively the chart is illustrating the cost-competitiveness increase for renewables relative to non-renewables as a result of the IRA.

¹⁵There are serious pitfalls when considering the so-called “levelized cost of energy” from renewables as the basis of comparison against non-renewable power sources. As Michael Cembalest, Chairman of Market and Investment Strategy at J.P. Morgan Asset & Wealth Management, has noted, the problem with LCOEs is that they don’t account for (a) the need for backup power and storage for when renewables are not generating, (b) how the value of electricity supplied changes throughout the day, and (c) the cost of overbuilding renewable capacity to meet demand in a deeply decarbonized system. Michael Cembalest, “Growing Pains: The Renewable Transition in Adolescence,” Eye on the Market, J.P. Morgan, March 28, 2023. For these reasons, the emissions reductions policymakers are assuming will be met as a result of the IRA may be too aggressive.

¹⁶Clean Energy Investing in America, American Clean Power, April 2023.

¹⁷The relevant credits are 45C and 45X; 45C is tax credit for companies to take against the cost of building factories to manufacture components for renewable energy projects; 45X is a tax credit for each renewables component manufactured. The two cannot be “stacked,” meaning the 45X credit is not available for components made at a factory that already benefited from the 45C credit. “Federal Tax Credits for Solar Manufacturers,” U.S. Department of Energy, Office of Energy Efficiency & Renewable Energy, October 2022.

¹⁸The mechanics of how IRA tax credits boost profitability for manufacturers is, most typically, via the direct pay option: Many or most manufacturers don’t have large tax liabilities to offset, so they opt for direct cash payments from the government (paid as part of the year-end tax filing process), which lowers their COGS (cost of goods sold) and thereby raises their margins. Most manufacturers seem to be opting for 45X (rather than 45C), which is an implicit bet that factory utilization and output will be high, translating into higher credit payments compared to 45C, which is based on the overall upfront investment of a new manufacturing site.

¹⁹Nusaiba Mizan, “Fact-Check: Does the Texas grid have 15% more power generation capacity than last year?” Austin American-Statesman, April 3, 2022.

²⁰Energy communities are areas where a coal-fired power plant has closed since 2010, or a coal mine has closed since 2000; or a metropolitan or non-metropolitan statistical area where 0.17% or more of direct employment, or at least 25% of local tax revenues, are related to extraction, processing, transport or storage of coal, oil or natural gas, and where unemployment is at or above the national average in the previous year.

²¹Some general information on those developer tax credits: Before the IRA, renewable energy production site developers for solar could receive a 26% federal tax credit. Now, the IRA’s full “stack” of credits could be as high as 70%. A developer could receive one 30% tax credit for facilities meeting certain labor standards, which rises to 40% for investments in facilities using U.S.-made “domestic content,” and which rises further to 50% if the facility is in an “energy” community, and finally to 60%–70% if the project site is located in a low-income area and/or meets low-income housing qualifications. The domestic content qualifications are satisfied if (1) 100% of the steel or iron used in a qualifying project is produced in the United States; and (2) 40% of the manufactured products used in constructing the project are produced in the United States. For the definition of a low-income community, see “Initial Guidance Establishing Program to Allocate Environmental Justice Solar and Wind Capacity Limitation Under Section 48(e),” Internal Revenue Service, February 13, 2023.

²²John Liu and Paul Mozur, “Inside Taiwanese Chip Giant, a U.S. Expansion Stokes Tensions,” New York Times, February 22, 2023.

²³The Congressional Budget Office estimated that in a deficit neutralized scenario, the IRA would increase GDP only on the order of 0.05–0.1ppts over the next decade.

²⁴It would be an altogether different story, though, if the Federal Reserve were to change its inflation target, as some think tanks are advocating. We are not expecting this, at least not in the near term, but it is something to continually monitor. Justin Bloesch, “A New Framework for Targeting Inflation: Aiming for a Range of 2 to 3.5 Percent,” Roosevelt Institute, November 17, 2022.

²⁵We say this as a secular (not cyclical) statement. If the U.S. economy were to slip into a recession in the near term, we would expect Treasury bonds to serve as an effective hedge against declines in risk assets.

²⁶Cullen Hendrix, “How to Avoid a New Cold War Over Critical Minerals,” Foreign Policy, November 22, 2022.

²⁷And the other half disappears if vehicles do not meet percentage requirements for battery components produced in North America. For the full details surrounding EV tax credit requirements, see “If I buy a new electric vehicle will I qualify for the full federal tax credit up to $7,500?,” Electrek.

²⁸“NATO chief says Ukraine’s ammunition use outstripping supply,” Associated Press, February 13, 2023.