Unionization in the United States has passed two remarkable – and seemingly contradictory – milestones.

First, unions barely exist as a national economic force. The share of workers represented by trade unions slipped below 10 percent in 2024. Strip out public-sector workers, and the rate for the rest of the workforce was a mere 5.9 percent. That figure represents quite a fall: union representation is less now than it was in 1934, the year before the right to organize was enshrined into law by the National Labor Relations Act at the dawn of FDR’s New Deal. It’s as if the right to unionize has been eradicated. All that’s left, then, is to write a respectful eulogy.

Or not. Reports of the death of organized labor may be premature. Gallup polling shows that 71 percent of Americans approve of unions, a rate not seen in seven decades. A majority agree that the decline in unionization is bad for the country – and the samples polled supported striking auto workers, actors and writers in 2023 by overwhelming numbers.

It may be no coincidence that this surge in approval for unions has coincided with a cratering of approval for big business, which Gallup shows has dropped from 48 percent to 25 percent since 2001. The nonprofit, nonpartisan American National Election Studies organization, which has been collecting data since 1964, asks respondents to gauge their opinion of people and groups on a 0-100 scale. Between 1964 and 2012, labor unions and big business moved together in public perception even as they oscillated on the scale between 45 and 60. But after 2012, pro-union sentiment jumped more than 10 points and big business sentiment slipped by 5 points, creating the first gap between them in the survey’s history.

So which is it: are today’s unions the last gasp of a relic or a phoenix on the rise? Answering that question requires moving past the traditional framing of unions as one side of the boss-versus-worker struggle and viewing them through the lens of a rapidly changing economy. The rise and fall of organized labor, taking place over decades, has given economists the opportunity to develop more nuanced views of the role of unions.

Unions’ economic story just doesn’t fit a one-dimensional, “pick your hero” sort of choice. This isn’t about taking sides, pledging solidarity with the AFL-CIO or the conservative American Legislative Exchange Council. Rather it makes clear that organized labor affects the economy and society in complex ways. At this moment in history, one potential constructive role stands out: the United States has slipped into ever deeper economic inequality, to the point that the market-driven distribution of income and wealth – not to mention lack of generational mobility – has become as toxic to society as it was during the Great Depression. Unions, I believe, could play a major part in reversing the trend.

Inside The Sausage Factory

Economics, one could argue, is a discipline with twin North Stars: price and power. Prices are the mechanisms that reconcile market supply and demand. There aren’t enough goods and services – or the labor and capital to produce them – for everyone to have everything they want all of the time. Prices are the means of setting priorities. This “market clearing” function explains why, however much analysts are skeptical about the fairness and efficiency of real-world markets, few would abandon them entirely for alternative means of allocation.

Power, for its part, is the mechanism that determines the outcome of allocation through bargaining rather than markets. As an example of these twin pillars of price and power at work, consider what determines how much a worker earns. In the price world, the wage is the market-clearing price of labor – that is, the price that equates workers’ willingness to provide labor with its value at the margin to the employer (aka workers’ marginal productivity). In the power world, workers earn what they are able to bargain for, given their outside options (temporarily withholding labor or working elsewhere) and the information they have going into negotiation (how much their employer pays for similar positions, for example). And clearly, wage-setting is a product of both.

The classic argument against unions is that they use their power to bargain collectively to establish anti-competitive labor cartels that pull wages above marginal productivity. This effect can generate a net loss in the value of output since some unemployed workers would willingly sell their labor for less than the wage determined by bargaining.

Unionized workers certainly earn more than their non-union counterparts, but establishing that this results in lower labor demand or net loss for the economy is extremely difficult. The predicted negative effects of unions from classic theory is at least partly offset by the positive spillover of unions in practice. For example, unions decrease wasteful, productivity-reducing labor turnover within a firm. And they help ensure that legal workplace quality standards are met, increasing safety for workers and decreasing enforcement costs for the government. Union experts have asserted that, behind closed doors, employers often acknowledge that a good relationship with a union boosts morale, aids in communication and fosters teamwork.

Yet, a key part of the evolution in economic thinking about unions is not just understanding the externalities of unions in sustaining solid relationships between labor and management on market outcomes but viewing labor’s bargaining power as a force that can offset the anti-competitive impact of management’s wage-setting power. Econ 101 models assume that, in the absence of unions, labor markets are competitive. But over the past quarter century, economists have come around to the reality of “monopsony” in the labor market – that a lack of effective competition among employers allows them (and gives them strong financial incentives) to pay less than the value of their workers’ marginal productivity.

In many cases this lack of competitiveness among employers is created by mundane, real-world factors such as poor public transportation that limits the distance workers can commute and thus limits the size of the job market. In others, this power can manifest in employers’ insistence that workers sign noncompete contracts barring them from working in the same industry for months or years after they quit. But in a surprising number of cases it follows from illegal collusion, as employers form “no poaching” compacts or agree to follow fixed wage caps negotiated behind closed doors.

Where it was once not easy to find academic economists who believed the social benefits of labor organization exceeded the costs in terms of reduced productivity, the worm has turned. Mainstream economists now acknowledge that, in light of the distribution of power in the labor market, unions can be a corrective force, a source of countervailing power that pushes wages toward the efficient level. And the rise and fall of unions in the 20th century offers a historical proving ground for this view.

Lessons from the Golden Middle

Arguably the natural dynamic of the U.S. economy – the distribution of income and wealth determined by technology, market forces and accumulation of productive assets – is deepening inequality, with the richest gaining an ever-greater share. That seemed to be the case in the three decades before World War II (which is as far back as the data can reliably take us). In these decades, the top 10 percent of households went from garnering 43 percent of all pre-tax income to over half. And a good chunk of redistribution came at the expense of the bottom half of households, who went from taking 15 percent of pre-tax income to 13.5 percent. Much the same pattern can be seen since 1980, 45 years in which the top 10 percent saw their income share rise from 33 percent to nearly 50 percent while the bottom half fell from 20 percent to 13 percent.

But note the disconnect. In the middle years, the 40 years between 1940 and 1980, the dynamic of distribution reversed. During this era – dubbed America’s “Golden Middle” – the income share of the bottom half of workers surged while the very top did no better than hold steady. These were years in which blue-collar workers could look forward to owning their own homes, to regular vacations, to sending their kids to college – and sometimes on the wage of a single worker. Thereafter, of course, the music stopped, with inequality creeping back to the point that the U.S. is more unequal by this bottom-50 to top-10 metric than it was before World War II.

To be sure, this measure focuses on pretax, pre-transfer-payment share of total income, essentially measuring the inequality produced by labor, capital and financial markets before the government has stepped in with taxes and household subsidies to change distribution. It’s the inequality the economy generates, not necessarily the inequality that individuals experience. Nevertheless it tells an important story – really many stories – about how markets perform in the absence of countervailing power exercised by government and organized labor.

Economists and politicians have offered varying narratives of what caused the era of the Golden Middle. Some cite the reallocation of the vast resources mobilized for war back to civilian enterprises. Some cite the huge investments made by the federal government in infrastructure, technology, housing, consumer finance and education. Some include the economies of scale gained from the globalization of markets. But the narrative can’t be complete without examining how and why most workers got a bigger share of the rapidly expanding pie.

Recent research has leveraged historical data on union membership from Gallup to tease out the causal relationship (as opposed to simple statistical correlation) between union membership and economic inequality. They found that unions decrease inequality, and in fact were a source of the decline in inequality in the first decade of the Golden Middle. It is also likely that high rates of unionization helped sustain the Golden Middle for four decades and that falling unionization helped bring about its end, but that was beyond the scope of their analysis.

The conclusion that unions decreased inequality is in one sense obvious and in another surprising. It has long been established that being in a union is associated with significantly higher wages – but it’s unclear if those high wages raised or lowered the wages of non-union workers. Even at their peak, unions never covered more than a third of workers in the U.S. In the price world discussed above, one would expect unions to reduce the wages of other workers (by reducing the total demand for labor), and in the power world, to increase them (by redistributing bargaining power to workers). Both phenomena can be real. But the research suggests that the effect on worker-power dominated.

In terms of inequality, the economy today carries echoes of the economy of the late- 1930s. The pre-tax, pre-transfer income share of the bottom half of households was just 13.4 percent in 2023, the lowest on record apart from the worst years of the Great Depression (1932-34) and the feeble labor market following the Great Recession (2014-16). Going forward, unions might serve as an effective counter to this inequality. But getting from here to there would require a fundamental rethinking of how labor organization functions.

The Once and Future Union

It is a rule of thumb in policymaking that trying to re-create the virtues of the past by retracing our steps is futile. But it is still useful to understand what transpired. The blossoming of unions in the 1940s was made possible by the National Labor Relations Act of 1935 and accelerated by the needs of manufacturing during World War II that opened the door to organization. Neither are reproducible conditions – the war for obvious reasons, the protection bestowed by the NLRA for insidious ones.

Assaults on the right against labor organization efforts can only be characterized as relentless. The legal power of the NLRA was first curtailed by the Taft-Hartley Act of 1947, which, among other provisions, gave the states the discretion to pass their own “right to work” laws that allow non-union workers to free-ride on the bargains negotiated by unions. The latest blow was the peremptory firing of two members of the National Labor Relations Board who had been appointed by President Biden. The legality of those terminations is still to be determined. But at the moment, the board lacks enough members to rule, implying that it effectively has no legal authority, and by extension, neither does the NLRA.

The equally important story here is that employers have been tireless in their efforts to stop unionization in recent decades. “Union avoidance” is a $400-million-plus industry that dispatches consultants to organizing shops to persuade workers to oppose unionization. Amazon spent $14 million on union avoidance in a single year. By the same token, union busting, a miscellaneous set of mostly unlawful tactics to keep unions from gaining legal recognition, is rampant and brazen. A 2019 study found that employers were charged with violating federal law in 41 percent of all organizing campaigns.

Even where workers do vote for unionization, employers have found legal means to stall the contract process for years. The first Starbucks to unionize was in Buffalo in 2021 and more than 500 stores followed – but there is still no collective bargaining agreement. Same with the Amazon warehouse in Staten Island that voted to unionize in 2022.

Union advocates have long fought for the PRO Act, a bill that would reestablish protections for the right to organize that have been chipped away over the years. Given the legal precarity of unions, propping up labor rights with this or similar measures is a bare minimum step toward reintegrating unions into the economy. But there’s no getting around that this is an uphill battle: employers have decades of experience making end runs around – or simply ignoring – labor law.

Hence, an irony. There are solid reasons to believe that a reinvigorated union movement could make a substantive difference in reversing income inequality. But the opposition to unions is so fierce and so entrenched that no amount of evidence or economic justification will bring forth compromise. Power must tip toward workers, but empowerment of workers will be fought unflinchingly.

Sectoral bargaining, I would argue, offers a narrow path forward. Unionization in the U.S. has largely been enterprise-based, where individual worksites are unionized and governed by collective bargaining agreements. With sectoral bargaining, entire industries are organized within a region. Sectoral agreements are by nature broad, meaning they have more coverage and less teeth. Each industry’s agreement is dictated by a board that writes standards to establish the floor for wages and working conditions. All workers are covered – individuals are not required to join the union to share in the benefits.

Sectoral bargaining does offer numerous advantages, though. For example, sectoral agreements cover all workers, regardless of whether they are legally defined as employees or independent contractors. Employees like this, because it prevents employers from undercutting employee wage and benefit standards by shifting to non-employee contracts. If all workers are covered, it can greatly reduce if not eliminate the incentive to cut costs by shedding fully protected workers. Contractors like this because they have limited ability to bargain for protection as individuals. Employers like this because it allows for the use of flexible, spot-worker contracts without employee or legal objections.

Sectoral agreements also create specific industry- region floors for pay and benefits. While low-paying, low-quality employers may recoil at this, most firms would welcome the chance to obviate the need to compete on the product or service market with firms who undercut the labor market.

Enterprise-based unions, it should be noted, are not rendered obsolete in a system that promotes sectoral organization. Workers still have the right to organize their worksites. However, the negotiating burden on any one union falls considerably, as much of their wage and benefit “asks” are handled at the industry level. Enterprise-based unions can thus focus on enterprise-specific concerns, like working conditions, holiday hours, communication and investment – the functions of unions that employers will at least privately admit can serve them as well as workers.

Because sectoral agreements are the standard in Europe, it is easy to lapse into the thinking that they are alien to Americans. But sectoral agreements are in many ways more fitting to American labor markets. Indeed, in the early days after the Fair Labor Standards Act of 1938, industry “tripartites” of union, business and public representatives set wages for specific industries. The tripartites were both effective and popular, but were abandoned after the Taft-Hartley Act of 1947.

Conservatives are inclined to blame big unions like the American Federation of Labor that opposed government involvement in labor. At the time, liberals blamed Southern Democrats who were wary of empowering Black workers. Neither explanation carries much water today.

With sectoral bargaining, the tripartite board system of worker, employer and public representation is tasked with setting both wage and working conditions. The framework allows for considerable room for creativity in designing the governing standards of the tripartite boards, which might prove critical for matching the institutions to the sectoral circumstances.

An overly political public representative would reduce stability; the position is better thought of as a place for an informed, professional mediator. Unless the framework includes a clear, mandated timeline for reaching bargaining agreements, employers could simply stall indefinitely to prevent agreement from happening – as they often do now with union contracts. Sectoral boards would also need to be coordinated across regions to prevent a competitive race to the bottom in terms of wages and working standards.

All this isn’t pie-in-the-sky. The fast food wage board in California, the rideshare driver wage established in Massachusetts, and the new workforce standards board for nursing home workers in Minnesota are each examples of sectoral agreements established in just the past year. It is not that sectoral agreements could be effective policy – they already are effective policy.

The Economy We Want

The economic, social and political consequences of deepening income inequality compound over time as the majority of Americans are left behind. The traditional framing of unions dictates that there are two sides to the fight – that one is either with workers or their employers. But the economic evidence and history of the past century belie that choice. If there are sides to take, they’re being for or against income inequality at the extremes we’ve reached. Unions – whether they are enterprise-based or sectoral-based – should be a means to an end, not the object of an ideological battle resurrected from the 19th century.