It’s an image deeply embedded in American culture: management sitting on one side of the table, its representatives wearing suits and looking perhaps a bit standoffish; and labor sitting on the other, its members tailored less elegantly and looking insistent, even angry. The confrontation has a long history, but the terms on which the two sides now negotiate have changed dramatically from the days of strikebreakers and labor goons. Today, when a private-sector union and an employer negotiate their first contract, the union issues all the demands, offering in exchange only more of the labor peace that the employer presumably already enjoyed. The best that a company can achieve in this scenario is to placate the union at minimal cost. To a labor organizer, the question “What’s in it for the employer?” might seem bizarre—the whole point of organizing, after all, is to improve labor’s position against capital. But even if shifting power toward labor is desirable for organizers, browbeating management into making concessions may not be wise. At a minimum, the effectiveness of such an approach will depend on economic and social conditions.

Those conditions may have been favorable in 1935, when Congress passed the landmark National Labor Relations Act (NLRA)—guaranteeing the right of private-sector workers to unionize and bargain collectively for better pay and working conditions—but much has changed in 80-some years. The union movement’s strategy under the NLRA of trying to organize whole industries—which harmed economic dynamism, even as it penalized consumers through higher prices and discouraged new workers by limiting job opportunities—proved unsustainable. Further, employment regulations, by embedding traditional areas of bargaining directly into the standard law of the workplace, have undermined the entire premise of the NLRA. Meantime, the plunge in private-sector union membership from 36 percent in 1953 to less than 7 percent in 2015 has exposed unions’ minimal value to workers. The total employee share of national income—66.1 percent at the start of 2016—is actually higher today, with a much smaller union presence, than it was during the unions’ 1950s zenith (64.5 percent).

As its economic relevance has waned, Big Labor has mutated into a predominantly political force. Of $2.1 billion spent by the 30 largest federal-election donors since 1990, public- and private-sector unions accounted for more than $1 billion—and directed 97 percent of that largess to Democrats. Such one-way spending is at odds with union-member ideology, which aligns 26 percent “liberal,” 44 percent “moderate,” and 30 percent “conservative”—only slightly left of the general population.

Democrats prize union bosses’ Midas-like ability to transform the dollars and energy of a bipartisan workforce into homogenous left-wing support. Thus, their response to plummeting union membership isn’t to promote substantive reform that might make organizing more attractive to workers but instead to push for procedural changes to help unions bring more workers into the existing system and ensure that donations keep flowing. Republicans seem content to frustrate such efforts and watch the system continue to wither.

What a missed opportunity. Organized labor is neither inherently partisan nor inherently counterproductive economically. In theory, an arrangement by which workers “bargain collectively” and offer “mutual aid,” as the NLRA establishes is their right, can be a neutral or even positive part of a flourishing market economy. Other countries have implemented labor systems sharply different from—and more effective than—the American one. Even within the U.S., examples exist of organized labor’s potential to operate more constructively. A reformed legal framework for labor could help address several critical challenges, including the plight of less skilled workers struggling in the modern economy. It’s time for a new approach.

Union organizers locate the NLRA’s primary shortcoming in its failure to protect workers’ “free choice.” As they see it, the law’s requirement that they conduct an open campaign before any unionization vote within a firm invites the employer not only to inform workers about potential drawbacks of joining a union but also to intimidate them with illegal tactics that the law tries, but fails, to deter. Thus, say organizers, the appropriate remedy is to tilt the organizing process in their favor with rules that sideline employers—such as “card check,” which enables a union to show majority support in a workplace simply by collecting enough employee-signed pro-union cards, dispensing entirely with secret-ballot elections.

It’s pretty clear, though, that workers don’t like unions much these days. In a landmark 1994 survey of more than 2,400 nonmanagement workers, Harvard professor Richard Freeman primed respondents with various questions couched to make organizing seem attractive, and then asked those not already in a union whether they’d vote for one, if offered the chance. Only 32 percent said yes. When respondents had the option of joining a union or participating on a cooperative management-employee committee for discussing problems, union support fell to 23 percent; the committee concept proved more than twice as popular.

The mid-1990s through the beginning of the 2000s marked a high point for recent pro-union sentiment nationwide. In 1994, when the survey was conducted, 57 percent of Americans viewed labor unions favorably. That figure fell as low as 41 percent during the Great Recession and has climbed only a little since then. In 2010, constitutional amendments seeking to bar card-check campaigns won approval in four states. From 2012 to 2015, right-to-work laws weakening union power expanded into the former labor strongholds of Indiana, Wisconsin, and Michigan and now encompass the majority of states.

In a controversial 2013 essay, longtime organizer Rich Yeselson argues that efforts to reinvigorate the labor movement through new organizing campaigns are futile. “[S]mart union strategists,” he acknowledges, “can’t compensate for a mostly (though not entirely) uninterested working class.” Thomas Geoghegan, a veteran labor lawyer who sought the Democratic nomination for a Chicago congressional seat in 2009, admits that even in that political environment, he “lost [his] nerve” to campaign in favor of organized labor.

“The plunge in private-sector union membership to less than 7 percent has exposed unions’ minimal value to workers.”

“As we know, most Democrats just won’t say, outside of a union hall, ‘You ought to join a union,’” writes Geoghegan in his 2014 book on the future of organized labor. “Yes, of course I think they should say it, but they won’t. So let’s deal with it. Like it or not, the word ‘union’ brings up too many mixed feelings.”

Chattanooga, home to a Volkswagen manufacturing plant with more than 1,500 hourly workers, is the Waterloo of 1930s-style organized labor. VW decided to support the United Auto Workers (UAW) in what would have been the first successful campaign to organize employees of a foreign auto manufacturer in the American South. For two years, management coordinated public statements with the union and let organizers into its facilities to meet with workers. The majority of workers gave organizers signed cards signaling support. But when the secret ballots were counted in February 2014, the unionization drive lost, 712 to 626.

What workers understand, but labor organizers apparently do not (or will not), is that the NLRA’s model of hyper-adversarialism offers them much less upside than it once did and lots more downside, guaranteeing a conflict-ridden relationship with their employer and producing only short-term “gains” that, over time, reduce workers’ value and opportunities. Among the most telling findings in Freeman’s 1994 survey: given the choice of representation by an organization that “management cooperated with in discussing issues, but had no power to make decisions” or by one that “had more power, but management opposed,” 63 percent of workers preferred cooperation and just 22 percent an adversarial stance. These results held even among active union members.

What’s changed is the background against which labor law operates. In 1935, workplaces were largely unregulated, the Great Depression had pushed the working class to the brink, and widespread labor unrest was causing severe economic and social disruptions. The NLRA’s objective was not only to empower workers but also to end what Senator Robert Wagner, the law’s sponsor, called “a procession of bloody and costly strikes, which in some cases swelled almost to the magnitude of national emergencies.” Section 1 of the NLRA accordingly begins: “Experience has proved that protection by law of the right of employees to organize and bargain collectively safeguards commerce from injury, impairment, or interruption, and promotes the flow of commerce by removing certain recognized sources of industrial strife and unrest, by encouraging practices fundamental to the friendly adjustment of industrial disputes.”

Given the chaotic alternative, an employer might have seen real value in his employees organizing and bargaining peaceably under the new rules. But since 1935, regulations have made superfluous most of the basic terms and conditions of employment that were the basis for collective bargaining. Federal law now provides a safety net of benefits for the elderly, disabled, and unemployed; a 40-hour workweek, paid overtime, and a minimum wage; prohibitions on discrimination; workplace safety standards; personal leave; and a mandate that employers provide high-quality health insurance. Yet unions must still find something to deliver to their dues-paying members. If the political process dictates (via regulation) those practices of highest value to workers and it funds (via taxpayers) the basket of resources and protections that society seemingly wants citizens to enjoy, what’s left to bargain over?

Thus, the prevalence of destructive work rules, circuitous grievance procedures, and counterproductive seniority systems in collective-bargaining agreements. With most sensible restrictions already legally mandated, employers have two options: make wage concessions that will impose higher costs now; or make peripheral commitments that will be felt only later. Codifying new work rules tends to cost little at first, since the parties usually design them aware of the limits of current operations—but unaware of how market demands and production technologies might change over time. Forcing managers to retain incompetent workers costs little—until a critical mass of bad workers accumulates. Promising seniority doesn’t hurt, either—until layoffs become necessary and better workers get let go. Bestowing lavish pensions is painless, too—until they have to be funded. The longest-tenured and least effective workers are usually the union’s strongest supporters, which makes such provisions ideal from their perspective as well. The result: shortsighted agreements that best serve each side’s immediate interests but can have considerable long-term costs.

In 1936, after a summer that saw hundreds of deaths from poor working conditions in Michigan auto plants, GM workers in Flint launched their famous sit-down strike to achieve recognition for the UAW. Among their demands: permission to speak in the lunchroom. Seventy years later, GM had converted the site of the sit-down into a “rubber room” where laid-off union members earned salary and benefits for doing nothing, as part of an industry-wide “jobs bank” that cost nearly $1 billion per year.

As the declining value of collective bargaining eroded the NLRA’s foundation from one side, the growing burden of its costs caused significant damage from the other. Throughout the postwar boom, when oligopolies controlled large swathes of an American economy isolated from foreign competition, multiemployer negotiations between an industry’s major producers and a union representing all the industry’s workers were common. Collectively bargained terms that applied to all industry producers would maintain the competitive balance among them; if labor costs rose, they would raise prices together, preserving profit margins.

This translated into higher prices, lower output, and slower innovation for the economy as a whole—and weaker job growth, harming prospective employees. Union members were also consumers, of course, which meant that they were often negotiating against themselves—balancing their desire for higher wages and benefits against their desire for lower prices. If nonunion consumers were the deal’s real losers, well, then, that was all the more reason to join a union.

In recent decades, the collectively bargained concessions that unionized firms have adopted often put them at a disadvantage against foreign competition, as well as against new, nonunionized domestic rivals. Unionized firms, as a consequence, have sought to shift their capital toward plants, regions, and countries where they can operate free of union constraints. Even if an industry remained insulated from foreign competition, the cumulative effect of operating for decades in an environment that muffles competition, constrains output, and stalls innovation has tended to leave it smaller and less productive than it might otherwise be. Over time, the NLRA has burdened unionized facilities within a firm, unionized firms within an industry, and unionized industries within the economy.

The trend in manufacturing is illustrative. Union defenders cite a trade- and automation-driven employment collapse in that sector, a natural base for organized labor, as a no-fault explanation for the dwindling relevance of manufacturing unions. But this confuses cause and effect. Yes, unionized manufacturing employment plummeted from 7.8 million to 1.4 million from 1973 through 2015. But the drop accounted for more than 100 percent of the decline in manufacturing jobs during that period; nonunion manufacturing employment actually increased, from 12.3 million to 13.2 million.

The NLRA is clearly not accomplishing its goals. Yet rather than contemplate how organized labor might function better, union leaders and their political allies just keep the emphasis on getting more workers to join unions.

Collective action by workers under a different framework could help address some of America’s greatest challenges in ways that policymakers across the political spectrum should consider.

Call the new organizations “co-ops.” One potential path for this new kind of organization, taken by unions operating under the “Ghent system” of many European countries, is to establish relationships with workers and provide benefits to them outside the employment context. In Denmark and Sweden, for instance, unions administer an unemployment-insurance system that workers join, independent of their employer. Those countries don’t require workplace elections, good-faith bargaining by employers, or compulsory dues payments; yet a majority of workers are union members. While the scale of the American labor market may seem dissimilar, management of most social insurance—including unemployment, disability, and health—has devolved already to the state level. Co-ops can plausibly supplement or supplant state-level programs, with the relationship perhaps differing based on a state’s economic profile.

“In the Danish system,” explains SUNY Buffalo’s Matt Dimick, “the employer’s freedom to hire and fire is explicitly linked to the robust unemployment insurance guarantee in its vaunted program of ‘flexicurity.’ Hence, just as the official slogan of Swedish social democracy proclaimed in the 1950s and 1960s, Denmark’s model of flexicurity is described as providing ‘employment security, not job security.’” Remarkably, the American conservative Yuval Levin offers a prescription for twenty-first-century America that echoes the central tenet of mid-twentieth-century Swedish social democracy. Supports for workers, he writes, “will need to help make a diffuse labor market more secure, rather than trying to reverse its diffusion.” Co-ops could help do this.

Co-ops also could partner with employers to improve the job-readiness of new hires and offer job training for all workers. When business leaders complain that they can’t find enough qualified employees, the solution seems self-evident: if they need a better-trained workforce, perhaps they should invest more in training. But the economics of human capital are complex. If an employer improves a worker’s skills, the worker can demand to be compensated accordingly—or leave for another firm. The employer might design his training so that the worker’s skills are inapplicable elsewhere, but this leaves him far more vulnerable to an organizing campaign and strike; if his specially trained workers walk out, in other words, they can’t easily be replaced.

Employers thus would like to hire high-skilled workers whom they don’t pay to train, just as workers would like to acquire more skills and work in systems that value them accordingly. But how to get there? Co-ops could fund and conduct training programs, negotiate with employers to share costs and provide access to apprenticeships, and draw on government education programs as well. While workers would pay for a portion of their own training via co-op dues, they would benefit from substantial employer and government support—as well as, eventually, the full value of their improved skills.

Such schemes raise the prospect of twentieth-century corporatism, with established employers, unions, and government collaborating to drain a slush fund of other people’s taxes and dues—but in this case, the correct participants, sources of funds, and incentives are in place to minimize that risk. Placing worker-controlled and -funded co-ops at the center of training initiatives should keep them focused on the interests of workers. Even if co-ops get “captured” by self-interested leadership, that self-interest should point toward maximizing the skills, earnings, and, therefore, dues of as many workers as possible. For their part, employers will have to negotiate for access to the co-op pools that offer the highest-quality talent, and they should seek out those that can accomplish this at the lowest cost.

This approach would be superior to the nation’s largely ineffective current network of job-training efforts. Decision making would shift from government agencies and well-meaning nonprofits to employers and workers—who need each other, understand their requirements best, and will measure their return on investment by how many high-productivity jobs result, not by “enrollment” or “graduation,” as is typically the case with government and nonprofit plans. Co-op-led training would introduce a fair, insurance-like support mechanism, whereby dues from the currently employed help pay for the unemployed to learn new skills.

Examples of union-led training already exist. NABTU (North America’s Building Trades Unions), for instance, operates 1,600 joint labor-management training centers, funded by more than $1 billion in annual dues and employer contributions. The Las Vegas culinary union’s Culinary Academy, created in partnership with the Las Vegas hotels, is another example, providing basic job training to recruit new workers into the trade and then ongoing training to help them advance in their careers (though the union, one should add, is also known for its aggressive, oppositional tactics).

Least tangibly, but perhaps most importantly, co-ops could be a mediating institution that helps strengthen civil society for poorer and less educated Americans. Formal organization increases the chance that a workplace also becomes a nexus of community. It establishes the infrastructure for workers to build relationships, access resources, and provide mutual aid. It creates space for the “solidarity” that David and Amber Lapp of the Institute for Family Studies emphasize is critical to the social health of working-class individuals and communities. Pope John Paul II, they note, described workers’ associations as essential, “not only in negotiating contracts, but also as ‘places’ where workers can express themselves.” Such institutions can play a role in the broader community and offer a point of contact for those trying to enter the workforce. Policymakers generally agree on the importance of rebuilding mediating institutions and social capital but find few levers to pull on behalf of family and faith. Reforming organized labor, by contrast, is an appropriate and plausible government intervention.

Co-ops could establish collaborative relationships with management. From the workers’ perspective, the opportunity to raise concerns and suggestions from within the protection of a group has inherent value and is likely to lead—at least, sometimes—to positive change. From the employer’s perspective, such communication can provide useful feedback, which could improve the experiences of employees as well as customers. Further, regular interaction between the two sides—as equals committed to (and essential to) the success of the enterprise—should contribute to more autonomy and innovation on workers’ parts and to the emergence of more empathetic cultures within and beyond the workplace. That’s something most employees want, as Freeman’s survey showed.

Such arrangements exist in many countries as company-specific “works councils”—most notably, in Germany, where they’re present at almost 90 percent of firms with more than 500 workers and have significant authority not just to hold discussions but also to make operational decisions. They’re presently prohibited in the U.S. by section 8(a)(2) of the NLRA, which treats their formation, absent a formal union, as an unfair labor practice.

The failure to reform this provision offers another reminder that union and worker interests don’t invariably align. Unions rightly fear that companies and workers would embrace this alternative form of organized labor. (Volkswagen cooperated with the push to organize its Chattanooga plant chiefly because it wanted a works council.) Works councils, says former AFL-CIO president Lane Kirkland, are “sham organizations designed to prevent real worker empowerment.” Big Labor’s vocal endorsement of “free choice” does not, it seems, extend to choices other than joining Big Labor.

Co-ops could provide a market-based alternative to the government’s employment regulation. The need for such regulation stems largely from the premise that individual workers lack the leverage to protect their interests—an obsolete consideration when an organized workforce can bargain collectively. Why shouldn’t a co-op be allowed to accept wages below the federal minimum, perhaps for apprentice-level workers enrolled in job-training programs?

While a co-op, when negotiating with an employer, would never be forced to accept terms or conditions below the federal standard, it could have the option of doing so if the employer offered something compelling in return. Workers could substitute their own judgment for that of federal bureaucrats about what they value most. A retail chain and its workers, for instance, might both benefit from an hours-scheduling system that created certainty and eliminated “on-call” time, while lowering overtime pay below time-and-a-half. A regulatory adjustment, voluntarily agreed to by employer and co-op, could reduce employer costs and boost worker satisfaction.

Treating employment law not as ironclad but as a default rule from which the parties can agree to depart also gets rid of the problem of one-sided bargaining. Instead, workers in a given firm would have the choice between a federally regulated workplace and one overseen by a collective-bargaining agreement. Such bargaining could still be adversarial, in the way any business negotiation involves two sides seeking to maximize their own gains, but it would not be zero-sum; the outcomes could deliver significant benefits to both sides. Imagine the first day of negotiations opening with the employer dropping an inch-high stack of rules on the table and asking, “How much do you want if we can get rid of this?”

Co-ops could extend employment-like benefits and training functions to sectors of the economy with no traditional employer. Over the past decade, the U.S. added more than 9 million independent-contractor, on-call, and temporary jobs, while traditional employment actually fell. The NLRA delivers the worst of both worlds—pushing employers to avoid worker relationships that could trigger collective-bargaining obligations but offering little to the growing mass of less than fully employed workers that this dynamic helps create. Co-ops do the opposite, increasing the value of employment relationships and providing nontraditional workers with some of the infrastructure that they otherwise lack.

The “gig economy” introduces efficiency and dynamism into certain markets, while offering valuable flexibility to workers. But as George Mason University economist Tyler Cowen observes, that kind of labor-market flexibility implies that “more workers will have to teach and train themselves, whether for their current jobs or for a future job they might have later on. I submit many people cannot train themselves very well, even when the pecuniary returns from such training are fairly strongly positive.”

Given the burdens of current employment regulation and labor law, the transportation-network firm Uber is understandably reluctant to engage with its drivers as employees or to risk collective bargaining. But what if its drivers formed co-ops, independent of their relationship to Uber, to support one another and discuss with the firm arrangements like human-capital investment, which might benefit both sides?

Like the guilds of the past, co-ops could provide structure to otherwise free-floating careers and a valuable signaling mechanism to prospective employers. They would also give workers the scale and security to negotiate with buyers of labor in the marketplace. To take an ambitious example from Grand New Party, written by National Review’s Reihan Salam and the New York Times’s Ross Douthat two years before Uber’s founding:

In the long run, unions might evolve into enormous talent agencies with an economic stake in increasing the wages of members by, for example, taking some small cut of any salary increase and then reinvesting the money into providing workers with the resources to move from sluggish labor markets to booming labor markets. Whereas traditional unions have an interest in keeping workers in one place, new model unions would encourage them to be footloose, all while maintaining strong ties to their fellow workers.

Any proposed labor reform should be guided by a question: Will it facilitate such imaginative futures, or frustrate them?

Reflecting the lack of popularity of organized labor today, workers voted against unionizing at Volkswagen’s Chattanooga, Tennessee, production site in 2009. (NADJA BRANDT/BLOOMBERG/GETTY IMAGES)
Reflecting the lack of popularity of organized labor today, workers voted against unionizing at Volkswagen’s Chattanooga, Tennessee, production site in 2009. (NADJA BRANDT/BLOOMBERG/GETTY IMAGES)

The smoothest path to reform may be not to update or replace the NLRA but to leave it in place and go around it. A union whose members are satisfied can continue to represent them, and an organizer who believes that an NLRA union would win majority support in a workplace should be free to pursue such a campaign. But other options should be available, too. Over time, especially if co-ops proved their value, membership in NLRA unions would decline toward zero as capital departed unionized firms, remaining workers voted to change their organizational form, and few new organizing campaigns succeeded. If workers chose to give NLRA unions a fresh look, they could still do that. This is what real employee choice would look like.

Effective reform would have four elements. First, the NLRA must no longer have exclusive jurisdiction over relationships between employers and organizations of workers. Its definition of a covered “labor organization” must narrow from all organizations of employees whose purpose is “dealing with employers” to only those established for the purpose of using NLRA-defined rights and processes. The 8(a)(2) prohibition on nonunion collaboration between employers and workers must go. None of these changes affects the ability of a union to operate with its current model—to the extent that workers choose it.

Second, the government should formally recognize the existence of the “labor co-operative”: a nonprofit controlled by its dues-paying members for the purpose of advancing their employment and creating value, rather than merely reallocating it. Co-ops will be held to governance and financial standards appropriate to their potential roles and will be eligible to partner with government in delivering benefits. They will also have the capacity to earn recognition as the collective representative of employees in a given workplace, but their existence will not depend on such recognition.

Third, the government should recognize the “works council”—a group of worker-selected employees responsible for representing employee interests to management, including the ability to bargain collectively. An employer could support the creation and operation of such a council, but checks must exist on any possible employer domination of its decision making. For instance, employees might have recourse to a snap “vote of no confidence” in a council that they felt wasn’t fully representing their interests.

Fourth, a recognition process similar to that existing for NLRA unions should allow a co-op or works council to become the exclusive representative of employees for negotiating a collective-bargaining agreement with an employer—but that bargaining process will differ from the NLRA’s. Applicability of most employment regulation should be negotiable. Multiemployer bargaining should be banned. Any defined-benefit pensions should be set explicitly outside the protection of federal insurance—if workers want to make such a bet on employer solvency, they should do it on their own initiative. These changes improve the position of the employer by design, with the goal of creating a space for mutually beneficial agreements.

Co-ops could take several forms in such a framework. Some might be specific to a company, community, industry, occupation, or combination thereof. A works council might choose to form a co-op for its own employees and ultimately expand further into the community, or a nonprofit might incubate a number of co-ops within a city. An employer and community college might partner to get a co-op started. Benefits to scale and expertise-sharing will emerge, but the importance of local relationships and community engagement should place a high premium on local control. Some of these hypothetical arrangements will undoubtedly prove impractical, while many more alternatives not yet conceived will emerge if employers and workers have the freedom to innovate outside the NLRA’s sclerotic confines.

Unlike with unions, membership and dues payment would always be voluntary, meaning that the path to success would entail effective provision of services, not an accumulation of political power and victories in contentious workplace votes. Organizations would have to behave more like employee-owned service providers than the national machines of Big Labor. The act of “organizing” would look more like selling gym memberships than sidewalk picketing.

Social safety-net reforms could allow properly constituted co-ops to receive funding for programs like unemployment insurance, job training, and community college. Current proposals like wage insurance and subsidized apprenticeships should consider what role co-ops might play in delivery. Where employers do not provide health insurance, co-ops could serve as an alternative before workers shop for individual plans.

Last, co-ops must be kept from redirecting funds toward political advocacy. “There is nothing in the nature of unionization that requires the bundling of economic and political functions,” says Harvard professor (and former SEIU assistant general counsel) Benjamin Sachs. “Bundling is instead an artifact of history and, more to the point, of law.” After all, unlike unions, co-ops will potentially receive funding not only from member dues but also from employers, nonprofits, and the government. The restriction on advocacy is consistent with that applied to 501(c)(3) organizations (including those engaged in public-policy debates), which does not raise First Amendment concerns. Co-ops will be free to create, and raise funds for, separate political action committees (PACs), as many nonprofits do today. But their worker-representation and political functions would remain separate.

It’s not crazy to suggest that organized labor focus on delivering economic benefits to workers instead of campaign contributions to politicians or to imagine that workers might benefit more from collaboration than from conflict with their employers in the modern economy. Workers themselves seem to believe this. So perhaps it is conceivable that policymakers committed to a robust civil society, competitive markets, widely shared prosperity, and a more effective safety net could give organized labor—in a new form, that is—a fresh look.

Top Photo: The famous sit-down strike in 1936 at General Motors’ Fisher Body Plant in Flint, Michigan (SHELDON DICK/GETTY IMAGES)


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next