In 1938, as the country sank into recession and national unemployment hit 19 percent, President Roosevelt announced that America had a monopoly problem. In a historic speech to Congress, Roosevelt warned that extreme consolidation was hampering the economy and threatening our democracy. “Among us today a concentration of private power without equal in history is growing,” Roosevelt said. “This concentration is seriously impairing the economic effectiveness of private enterprise as a way of providing employment for labor and capital, and as a way of assuring a more equitable distribution of income and earnings among the people of the nation as a whole.”

The speech signaled a new intellectual direction for his Administration, heralding what would come to drive the second major phase of the New Deal: anti-monopoly policy. Notably, the first phase of the New Deal had taken the exact opposite route. Its centerpiece—the National Industrial Recovery Act (NIRA)—had established industry-wide codes, enabling firms collectively to limit production, hike prices, and boost profits, the idea being that government-sanctioned cartels would thwart crushing deflation. Encouraging companies to collude rather than compete, the policy had effectively suspended antitrust laws. But a series of setbacks compelled the Administration to change course. In 1935 the Supreme Court declared the NIRA unconstitutional and would go on to weaken other key pieces of the New Deal. High unemployment, meanwhile, persisted, prompting calls for bolder reform. By deploring concentration, Roosevelt was admitting that centralization and corporatist planning had failed to revive America’s economy.

To achieve his new agenda, Roosevelt appointed Thurman Arnold to head up the antitrust division of the Justice Department. During Arnold’s tenure, the division grew from 18 employees to nearly 500 and increased its workload from around 70 cases and investigations in 1938 to more than 300 by 1940. In addition to targeting anticompetitive schemes and predatory conduct, Arnold forced dominant companies to open their patent vaults, empowering smaller firms and spurring innovation. This bolstered antitrust regime led industrial production to rise and unemployment to fall, aiding the nation’s recovery out of the Great Depression.

Some policymakers and politicians today are starting to realize that America once again has a monopoly problem. Excessive consolidation is now a defining feature of our economy across sectors, from hospitals and auto parts to eyeglasses and chicken slaughter. This lack of competition produces a variety of material harms. Evidence suggests it depresses wages and salaries, raises consumer costs, stunts investment, retards innovation, and renders supply chains and complex systems highly fragile. There’s sound reason to think concentrated control is holding America’s economy back, helping explain why post-crisis recovery has stayed sluggish even as corporate profits boom.

The stakes are also deeply political. By massing wealth in a few hands, economic concentration breeds concentration in our democracy; it allows our public fate to be steered by the private interests of a few. Dominant firms that occupy gatekeeper roles—Google, for example, or Amazon—hold sufficient power to direct the fate of other companies, deciding who floats or sinks.

Restoring anti-monopoly law could be one of the most significant policy acts of the next administration. Key to its success will be crafting the right philosophy and setting the right goals. Anti-monopoly policy has been used both to promote competition and to promote consolidation—even within a single decade, as the shift from Roosevelt’s first to second New Deal shows. If we are not careful, anti-monopoly efforts could actually further concentrate power in particular sectors and types of companies. But with a clear sense of history, and a focus on market structure rather than material outcomes, a new anti-monopoly push could lead to a renewal of the American economy and American democracy.

So far, much of the new interest in monopoly in America has focused on its effect on growth. In the 2016 Economic Report of the President, for example, the White House analyzes how competition contributes to innovation and productivity, stating, “Competition from new and existing firms plays an important role in fostering…growth.” A paper from the Council of Economic Advisers, meanwhile, also notes that competition “may lead to greater product variety, higher product quality, and greater innovation, which drives productivity growth and helps lift living standards.”

Evidence suggests that this concern is warranted: Excessive consolidation, indeed, saps growth in three important ways. First, it impedes new business creation and development. Dominant firms in concentrated markets can use their heft to stifle entrants through various methods, including predatory pricing and exclusionary deals. In some instances, a dominant firm need not even engage in explicit anticompetitive conduct outright to block new players; its market position and history of predation alone may dissuade potential entrants. Take, for example, the candy market, where Mars and Hershey’s control over 60 percent of national candy sales. Their size enables them to dish out huge sums for shelf space in grocery stores, making it effectively impossible for an independent producer to bring a new candy bar to mainstream markets.

Reflecting this trend, a host of studies now shows that business formation in the country has declined drastically over recent decades. In the first of these reports, which I co-authored for New America in 2012, we found that new business creation per capita fell by 50 percent between 1977 and 2011. Similarly, in 2014 the Brookings Institution documented that the firm entry rate—firms less than one year old as a share of all firms—fell by nearly half between 1978 and 2011. The economy is “engaged in a steady, secular decline in business dynamism,” the authors concluded, and this decline is “not isolated to a few regions” but “is a pervasive force evident in nearly all corners of the country.” In a second study, the authors reported a “robust” link between rising business consolidation and declining firm formation.

This dramatic fall-off in entrepreneurship is troubling in part because new businesses are a vital engine for new jobs. As the Kauffman Foundation reports, “New and young companies are the primary source of job creation in the American economy,” accounting for nearly all net new job creation and almost 20 percent of gross job creation. Precisely because startups play an outsized role in this way, their decline “has troubling implications for economic dynamism and growth if it is not reversed.”

Second, excessive consolidation may also have the effect of suppressing personal income and benefits. One result of consolidation is fewer jobs, as companies routinely lay off thousands of workers after merging. Another result is less competition for workers. Regional concentration among hospitals, for example, has led to collusive schemes that hold down nurses’ wages. In other cases, firms establish straight-up labor cartels, as Silicon Valley tech firms did when agreeing not to poach each other’s workers. The rise of platform bosses like Uber promises to make the situation only worse.

Again, the data reflect this dynamic. The vast majority of American workers have seen their hourly wages flatten or decline since 1979. Increasingly skewed distribution of labor income, meanwhile, has driven inequality to staggering levels. Facing job insecurity and stagnant wages, individuals are staying put in jobs rather than starting new ventures—a fact reflected not just in declining rates of new business formation, but also in lower rates of self-employed Americans.

And third, concentration of economic control undermines growth because dominant firms can hold back the pace of advancement. Today a handful of companies across sectors wield outsized control over key technologies—Monsanto over genetic traits, for example, or Intel over semiconductors. Many of these businesses have come to monopolize these tools primarily through rolling up competitors and their patents. While patents are vital for promoting innovation, they are also routinely abused, to weaken rivals as well as to stunt development by fencing off corporate estates. AT&T famously was found to be blocking a host of available advances, including automatic dialing and office switchboards—innovations that would have risked chipping away its dominance. Short of actively obstructing progress, firms may simply refuse to invest in it. Absent competition, companies face scant pressure to tinker and improve—potentially explaining why business investment remains low even as firms spend billions in stock buybacks or simply sit on piles of cash instead.

Compounding this hazard today is the rise of dominant platforms like Amazon and Google, which increasingly determine how buyers connect with sellers and producers connect with users. Given the network effects at play, a few companies have emerged as de facto rulers, serving as the railroads and roadways of the Internet economy. Flush with capital, dominant platforms routinely buy out companies that might threaten their empires. Since these firms—which also include Apple and Microsoft—vertically integrate across multiple lines of business, they are allowed to compete directly with many of the players that now depend on them, creating conflicts of interest. Amazon, for example, hosts millions of third-party merchants that sell through its platform, but also directly retails goods that compete with these merchants. The troves of data that platforms amass heighten the potential for abuse. Amazon, for instance, uses the information it collects on what third-party merchants are selling to boost sales of its own products. The future, in short, belongs not to the people devising better products and ideas, but to the giants that pick and choose among them to serve their own interests.

America’s monopoly problem today largely results from a successful campaign in the late 1970s and early 1980s to change the framework of antimonopoly law. Anti-monopoly laws were originally passed not as technical economic regulation but as political law, to preserve self-governing communities and individual sovereignty—to prevent, in the words of William Douglas, “the concentration in private hands of power so great that only a government of the people should have it.” While the vigor and approach of enforcement varied, the idea that anti-monopoly law should promote a variety of aims, including the dispersion of economic power and safeguarding local control, continued well into the 1970s.

In the 1970s and ’80s a group of legal and economic scholars associated chiefly with the University of Chicago upended the traditional approach. Driven largely by Robert Bork, this revolution declared that the only legitimate goal of antitrust is economic efficiency, measured in the form of “consumer welfare.” Some prominent liberals ratified this view. In his Economics and the Public Purpose, John Kenneth Galbraith concluded that centralized planning, rather than open markets, was the best way to stabilize industries and boost prosperity. By focusing exclusively on material ends, both the neoclassical school and its critics effectively embraced concentration over competition.

This philosophy—stamped into policy by Ronald Reagan and maintained by subsequent Democratic and Republican administrations—unleashed a torrent of mergers and acquisitions and resulted in the abandonment of cases against companies that abuse or unfairly obtain monopoly power. A conservative judiciary raised the bar for proving antitrust violations; conduct that was previously considered illegal was now hailed as generating efficiencies. The result is concentrations of power across the economy at levels unseen since the Gilded Age.

Reviving antitrust will require, foremost, restoring a key original purpose: the diffusion of economic and political power. It must reflect the understanding that our capacity to self-rule depends intimately on the distribution of economic control, and that it is our interests as citizens—not simply as consumers—that competition protects.

The tools for promoting competition vary. Some industrial activities, for example, can easily be organized into open markets; others require us to accept network monopolies and regulate their power instead.

This vision should be enacted through at least three practical steps. First, the antitrust agencies should strengthen merger enforcement by revising the merger guidelines, the principles they apply when reviewing horizontal and vertical deals. Strengthened guidelines would, for example, broaden the set of harms the agencies identify as posing injury to competition and reflect a more sophisticated understanding of how vertical mergers risk foreclosing rivals. They would also commit to blocking anti-competitive mergers outright, rather than seeking to fix them through regulating business conduct or requiring divestitures. Evidence shows these remedies have failed, enabling merging firms to hike prices, degrade quality, and block entrants instead. Seeking to police companies after they’ve whittled down the number of competitors is no substitute for real competition.

Second, the agencies should initiate monopolization cases in order to target dominant firms that have abused or unfairly acquired their monopoly power. The primary legal authority for this is Section 2 of the Sherman Act, the basis for the government’s break-up of Standard Oil, AT&T, Alcoa, and Microsoft (though, on appeal, the court reversed the order that Microsoft disband). Today, however, enforcers have all but abandoned Section 2 cases—the equivalent of shelving your shears when the shrubbery is at its fullest. Although unfavorable case law has made litigating these cases tougher, enforcers have unique investigative powers and resources to pursue cases beyond the reach of private parties. Since technology platforms and data markets challenge traditional theories, enforcers should advance creative arguments and test untested areas of law. Even court losses would serve a function, by publicly identifying the areas of law that need revising.

Third, current legal presumptions—which favor defendants—should be flipped. In recent decades, courts have raised the burden of proof for parties initiating antitrust cases, introducing stringent legal tests and requirements. In practice, this has made it all but impossible for plaintiffs to win certain types of cases, rendering entire areas of antitrust law—like predatory pricing—effectively defunct. Certain conduct by dominant firms in concentrated markets should instead carry a presumption of illegality. Though a mainstay of competition policy for decades, this structural approach was abandoned for “price theory,” which, as one scholar summarized, propagates that “[w]hat exists is ultimately the best guide to what should exist.”

Although antitrust has been ignored by policy elites for decades, there are signs that a reform movement is quickly emerging. Last spring, the Senate Judiciary Committee held a hearing where officials from both parties expressed concern that national competition policy has failed. In June, Senator Elizabeth Warren gave a major speech cautioning that excessive consolidation weakens our economy and endangers our democracy, and called on America to recover its foundational suspicion, the belief that “concentrated power anywhere was a threat to liberty everywhere.” And over the summer, Democrats included in the official party platform a commitment to strengthening antitrust law and enforcement—the first time anti-monopoly has been included in the platform since 1988.

The renewed interest in antitrust holds great promise, but it also carries risks. It is true that restoring competitive markets would boost growth and productivity. But it would be a grave mistake to reduce anti-monopoly to an exclusively material end like economic growth. That would suggest that the structure of our economy be secondary to the fruits it delivers. To recover anti-monopoly law, we must recall that it governs the distribution of ownership and control—irreducibly political, not material, outcomes. Pretending otherwise not only distorts the foundations of antitrust, but fundamentally precludes it.