Emphasizing how industrial policy can boost economic growth, Mariana Mazzucato and her colleagues serve our economic policy debate well. More and more scholars, advocates, and top elected officials—led by President Joe Biden—support activist government economic policies. They want to mobilize economic forces against challenges such as climate change and in the process create jobs, innovation, and growth. But today’s discourse around industrial policy largely leaves out labor.

Any industrial policy needs worker voice and power.

Mazzucato’s work—including her breakthrough book, The Entrepreneurial State: Debunking Public vs. Private Sector Myths—links public funding and activist policies to technological breakthroughs, in the process revealing that the private sector has received too much credit for what the government has actually done. But she and other progressive industrial policy advocates focus mainly on the supply side. On the demand side, we need sustained higher household income, driven by increased worker power.

Any industrial policy needs worker voice and power. After World War II, John Kenneth Galbraith theorized that a growing, innovative, postwar U.S. economy needed both economic and political “countervailing power.” He understood that unrestrained economic concentration hurts growth and undermines democratic politics. Unions and collective bargaining were central to balancing corporate power in the postwar era, as industrial relations created countervailing power.

Economists have known since Keynes that the main drag to long-term sustainable economic growth is a sluggish multiplier and low marginal propensity to consume, caused by unproductive capitalist spending. The Economic Policy Institute has documented  how the relentless and widening gap between productivity and wages (what on a graph resembles a big alligator jaw) strangles innovation and growth while simultaneously feeding inequality. By now we should know that without wage-led growth and increased worker voice and power, the demand for innovation will be weak and new industrial policy proposals will fail, both economically and politically.

This is why progressives should be wary of conservatives who embrace a certain type of industrial policy. For example, in 2019 Senator Marco Rubio became an unlikely advocate for activist economic policy. In a congressional report he stated that “the critical policy consideration. . . is not whether states should organize their economies, but how they should be organized.” But Rubio’s focus, like Trump’s, was on protecting strategic industries, especially military-related technologies. Indeed, despite his rhetoric, in practice Trump’s industrial policy primarily centered anti-Chinese rhetoric and unproductive tax cuts for corporations and the ultra-rich.

Policies that solely aim to shrink big firms won’t empower workers and, in fact, could actually stifle transformative innovation.

Biden’s industrial policy is more promising—and it includes some pro-labor initiatives that reflect his concern for worker protections and wages. For example, Biden supports improved industrial relations systems through the PRO Act, which would increase union membership and bind employers to bargaining in good faith. He also wants to boost wages and increase consumption, both of which should increase incentives for innovation. Moreover, his economic advisors include Jared Bernstein and Heather Boushey, both of whom advocate for using federal power to create a countervailing force against employer wage suppression. They also want to increase childcare, anti-poverty spending, and pay for care work. But these pro-worker policies aren’t explicitly linked to Biden’s other efforts in strategic sectors, trade policy, and antitrust policy.

There is now bipartisan support for strategic industrial action, especially focused against China. The Senate (with the House likely to follow) has already passed the United States Innovation and Competition Act (USICA), which expands the federal role in “strategic sectors” such as semiconductors and artificial intelligence, and empowers the president to use trade policy against China while strengthening “Buy American” procurement.

Biden is already implementing procurement policies and raising domestic content in products bought by the federal government. Both the industry and unions back this, as they push to fully fund the CHIPS for America Act, which would provide $52 billion to support domestic semiconductor manufacturing. When it comes to trade, Biden has continued Trump’s steel and aluminum tariffs (popular with unions); Secretary of Commerce Gina Raimondo says the tariffs have been “very effective” in preserving jobs. While these efforts might save jobs or raise wages, they will not enable labor to be the necessary countervailing force against employers. We need this if we are to achieve shared U.S. prosperity.

While some hope that Biden’s industrial policy could reconfigure antitrust policy and advance workers’ interests, we are skeptical. Though his appointment of Lina Khan to the FTC was promising—suggesting that antitrust policy will target corporate power rather than the classic antitrust standard of implementing competitive prices for goods and services—simply breaking up monopolies will not help raise workers’ living standards.

Of course, we should control harmful concentrations of economic power, but old- and new-style antitrust policy ultimately have the same goal—restoring competitive markets. Policies that solely aim to shrink big firms won’t empower workers and, in fact, could actually stifle transformative innovation.

At the end of the day, to sustain a progressive industrial policy we need labor’s political support; workers must see concrete benefits from these policies.

Instead, we contend that industrial policy should focus on well-regulated oligopolies. Monopolies crush potential competitors and small firms face brutal competition and downward pressure on prices, constricted information, and inadequate investment. Oligopolies—as some economic models recognize—are the real-world industrial form with the most incentive and capacity to innovate.

We see oligopolies as the Goldilocks of industrial organization, neither too small nor too big. However, oligopolies do generate excessive economic power that must be counterbalanced by a progressive industrial policy centering industrial relations. Without unions, big firms in the same industry topple over each other to slash wages and prices to get market share. But strong unions can discipline this race to the bottom. New antitrust frameworks would reign in oligopolies’ diversion of funds to executive pay, money that would be better spent on research and development, and help direct productivity gains to wages.

Well-regulated oligopolies can allow unions to take labor costs out of competition, stabilizing labor demand and costs over the business cycle. Moreover, oligopolies are more likely to keep workers on during downturns, especially if there are anti-layoff clauses, work and gain sharing agreements, more generous and flexible unemployment insurance, and policies like countercyclical and universal federal job guarantees.

With the right power-sharing arrangements and public policies, oligopolies such as Amazon and Starbucks could provide middle-class wages and working conditions for all their workers, something that small businesses—Aunt Bea’s sewing shop and Joe’s café on Main Street—have little ability to do.

Mazzucato’s work reminds us of the extent to which modern capitalism relies on industrial policy. But any successful strategy must center industrial relations that promotes unionization and shared economic returns, as well as activist policies advancing workers’ wages, health and retirement, family and childcare, and anti-discrimination. We need a strong government to regulate oligopolies while still encouraging their economic benefits, including innovation and prosperity.

After all, at the end of the day, to sustain a progressive industrial policy we need labor’s political support; workers must see concrete benefits from these policies. Without centering labor in industrial policy, both the economics and politics will fail.