Labor Day 2019 comes in the midst of a crisis for the American labor movement.
In the mid-1950s, a third of Americans belonged to a labor union. Now, only 10.5 percent do, including a minuscule 6.4 percent of private sector workers; public sector unions took a hard blow in 2018 when the Supreme Court cut off a major source of revenue. The decline of union membership explains as much as a third of the increase in inequality in the US, has reduced voter turnout among low-income workers, and has weakened labor’s ability to check corporate influence in DC and state capitals.
Other countries have experienced union decline too. But this kind of dramatic erosion hasn’t been the global norm. Most European countries still have far greater levels of union coverage than the US, and as of 2013, more than two-thirds of workers in Denmark, Sweden, and Finland were union members. In France and Austria, a minority of workers are in unions, but 98 percent are covered by collective bargaining contracts.
So a growing number of labor law experts and even presidential candidates (including Sen. Bernie Sanders, former Rep. Beto O’Rourke, and South Bend, Indiana, Mayor Pete Buttigieg) have settled on a new approach to revive the US labor movement. Major labor leaders like SEIU President Mary Kay Henry are embracing this strategy too, and shifting their organizations in turn.
Henry has dubbed this “unions for all” — an entire industry’s workers, that is, not just in one company. This approach involves moving beyond the traditional form of union organizing that you know from movies like Norma Rae, where unions organize workplace-by-workplace and fight tough battles for recognition, to an approach closer to that used in Europe or Australia.
This approach is called “sectoral bargaining,” and it could change the way work is done in the United States.
“In 2016, we had the most pro-labor president since the 1960s, the most pro-labor secretary of labor since [FDR’s Secretary] Frances Perkins, an economy with shrinking unemployment and rising wages — and yet we lost a quarter-million union members in the United States,” says David Rolf, president of SEIU 775, a local union representing home care workers in Washington and Montana, told me in 2017. “We need to be trying everything.”
And “everything” goes beyond sectoral bargaining alone, to encompass policies like:
- Works councils, which are committees elected by workers in their workplaces meant to serve as a vehicle to register concerns and resolve disputes with management, even in workplaces that are not union-organized.
- Codetermination, a system in which workers have the ability to elect members to the company’s corporate board, giving them a voice in the company’s high-level decision-making.
- Union-administered unemployment insurance, which gives workers a reason to join unions and pay dues even if their specific workplace isn’t organized with a given union.
The common thread is moving beyond workplace-by-workplace bargaining to a system where union-like protections are the norm, not the exception.
How the US organizes unions, and how Europe does it differently
Before we get into all these exciting ideas percolating among labor thinkers and organizers about how to change the way American unions and labor relations work, let’s first review how unions in America do work.
The model of unionizing that dominates American labor, familiar from movies like Norma Rae, has been in place since the 1935 National Labor Relations Act. For a union to be formed, at least 30 percent of workers in a workplace petition for a union election. The National Labor Relations Board sets a time and place for the election to be held. If a majority of workers vote to be represented, then they’re all unionized. Sometimes, as happened at Vox Media, companies will voluntarily recognize a union for which a majority of employees have expressed support.
Because unionization happens in individual companies and workplaces, the system is known as “enterprise-level” bargaining. And if you’re covered by an enterprise-level union contract, the system works pretty well. Unionized workers in the US enjoy significantly higher wages and better benefits than nonunion workers, and have greater recourse if they’re being mistreated by their employer.
The problem is that as unions shrink, fewer and fewer people get those benefits — and that’s partly due to the structure of the enterprise-level bargaining system itself. “It creates all these perverse incentives for employers to oppose workers trying to join the union,” David Madland, a labor expert at the Center for American Progress, has told me.
Unions get higher pay for their members by demanding money that would otherwise go to shareholders and executives, and so the latter have every reason to fight union drives.
But according to Princeton economist Henry Farber and Harvard sociologist Bruce Western, an even bigger reason for the decline of unions than corporate resistance to organizing drives is that unionized companies in the US have added fewer jobs over time than their nonunion counterparts.
The slower growth has a few causes: Unions were most successful in now-stagnating or shrinking industries like manufacturing and transportation; investors are less willing to put money into firms where unions capture some of their profits; and unions increase labor costs for employers, who respond by hiring fewer workers. Western and Farber found that unionized firms’ slower growth accounted for most of the decline in union membership between the 1970s and ’90s.
But workers in most European countries, and some other rich countries outside the US, have figured out an ingenious way around this. Unions there bargain not at the company level but at the sector level — negotiating for all workers in an entire industry rather than just one company or workplace.
In Sweden, for example, bargaining takes place at three levels: nationally for all industries, between the national union confederations and an association representing all employers; nationally, for specific industries, between the relevant unions and employers; and locally among individual companies. For the vast majority of workers, wages are set at a combination of the three levels, with only very few having deals set primarily at the company level.
Because every company covered by the national deals has to abide by the same pay and benefit rules regardless of how many of their employees are union members, those companies have less incentive to discourage union membership among their workforce. Firms with more union members don’t have any competitive disadvantage relative to firms with fewer: They’re all paying the same wages and offering the same benefits. And employment growth doesn’t necessarily vary among firms based on how many workers are in unions, so there’s no reason for union membership to decay as firms with more union members do worse.
What sectoral bargaining would look like in the US
Allowing a union of fast food workers to reach an agreement with restaurant owners that gets uniform benefits for everyone might seem impossible in the US. And at the national level, it probably is. But the recent victorious fight for a $15 minimum wage in New York offers a path to sectoral bargaining at the state level.
Organizers there achieved a $15 minimum wage for fast food workers by convening a wage board. Wage boards have the authority to mandate pay scales and benefits for whole industries, after consultation with businesses and unions. That’s an awful lot like how European countries implement sectoral bargaining.
The New York wage hike was a partial victory; in 2016, the state legislature agreed to set a $15 minimum across industries, but in doing so, it stripped the labor commissioner of the power to use wage boards to raise minimums for specific occupations in the future.
But New Jersey, Colorado, and California still have laws like this on the books; Arizona does too, but only for minors. California’s works through an entity called the Industrial Welfare Commission, which still has wage orders setting industry-specific minimums on the books.
The IWC has been defunded since 2004 and does not convene currently, but there’s nothing stopping California’s progressive majority in the legislature from refunding it and spurring it to adopt more modern wage minimums than the ones left in effect 15 years ago. New Jersey’s law actually requires a wage board to be empaneled if at least 50 workers in a given occupation petition for one. In all three states, wage board recommendations that are approved by state authorities and go through public review have the force of law. (For more, read a recent law review article by the University of Michigan’s Kate Andrias, a vocal wage board advocate, laying out the specific laws in each state.)
Labor researchers like the Center for American Progress’s Madland have also been developing nationwide proposals for wage boards. In Madland’s proposal, the US labor secretary would convene separate national boards for each industry, composed of five employer representatives in the industry, five worker representatives, and a Labor Department representative. Each one would convene every few years to set minimum wages for occupations within the industry.
“The panels would be required to hold at least one public hearing in each region of the country, where priority would be given to the worker organization representing the most workers in the industry—with similar priority given to the most representative employer organization,” Madland writes.
States would also be allowed to maintain their own boards, provided the standards they set are higher than federal standards. This is similar to how the federal and state minimum wages work currently.
Madland and Andrias hardly alone in urging a move to sectoral bargaining. University at Buffalo law professor Matthew Dimick helped popularize the idea even before Fight for $15 took off in a paper called “Productive Unionism.” In another report released by the Center for American Progress in fall 2016, Madland called for “transforming unions from individual firm-level bargaining units into organizations or structures … that negotiate for higher wages and benefits across an entire industry or sector.” Columbia law professor Mark Barenberg wrote a report for the Roosevelt Institute in 2015 urging the same.
And in a way, sectoral bargaining is a natural extension of “alt-labor” approaches that have become popular in the labor movement in the past decade, which put less emphasis on traditional workplace organizing and more on building other groups to represent workers’ interests — like “worker centers,” which provide services to low-wage, often immigrant workers in cities and advocate for policy changes on their behalf. Those groups can push for policy changes, like minimum wage hikes, that effectively set a new labor standard for a whole industry.
“The increasing attention on sectoral bargaining is new, but it’s also part of a broader trend of experimentation that has been going on for many years, as people worried about the decline in union membership look for better ways for organizing groups to grow both their membership pools and their revenue streams,” Shayna Strom, a senior fellow at the Century Foundation and Obama administration veteran, told me.
“Sectoral bargaining is certainly getting more attention in legal academic and labor law policy debates,” Benjamin Sachs, a professor at Harvard Law School and former practicing labor lawyer, says. “The way I would think about it is that there’s an existential panic about what will happen to the labor movement. That’s not new, it’s just getting worse. … If we need unions for economic and political equality as I think we do, we have to do something to stop that downward spiral.”
In Scandinavia, unions run unemployment insurance — and the effects are huge
While sectoral bargaining could offer US unions a way out of the abyss, it has its limitations. Ninety-eight percent of French workers may be covered by some kind of bargained contract, but only 7.7 percent of French people are in a union, an even smaller share than in the US. The unions negotiate deals covering the vast majority of workers, but because those workers are covered whether or not they join, there’s little incentive to sign up and pay dues.
The low membership means unions can’t always negotiate the best deals — a significant share of sectoral contracts in France specify minimum wages lower than France’s legal minimum, meaning they have no practical effect. It also hampers unions’ ability to know what workers really want, makes them reliant on the government to “extend” deals, and weakens the unions’ financial standing because few members are around to pay dues. And it’s left the French unions less able to resist reform, like President Emmanuel Macron’s laws meant to move the country away from sectoral bargaining and toward US-style enterprise bargaining.
“Sectoral bargaining creates a free-rider problem even bigger than our current free-rider problem at the enterprise level, because all workers benefit from the higher wages that are negotiated,” Madland says. “So you have a strong disincentive to pay dues.”
But there’s a surprisingly simple plan to get around this, proposed by Dimick, the professor at the University at Buffalo School of Law. Unions could run the unemployment insurance system using subsidies from the government. That, known as the “Ghent system” after the Belgian town where it originated, is a key part of how Sweden, Denmark, Finland, and Belgium have achieved the highest union membership rates in the developed world.
The system emerged almost by accident. “Back before there was any unemployment insurance, unions just did it on their own as a mutual aid function,” Dimick says. When the depression hit and unions lacked the funds to keep paying out benefits, “State governments came to their rescue by subsidizing them. It was an easy fix to the problem of unemployment rather than enacting wholesale government insurance.”
The result was that many countries were left with totally voluntary unemployment systems. In the US, unemployment insurance is funded through taxes on employers jointly administered by the federal government and states. Participation is mandatory.
In other countries, you were required to actively walk into a union office and sign up in order to receive benefits if you lost your job. That put workers in close contact with unions and encouraged them to join; in some countries, union members are also given discounts on unemployment insurance. It’s quite rare for people to sign up for unemployment benefits but not join the union administering them.
Over time, a number of countries, like Norway and France, junked this system in favor of mandatory unemployment insurance. But countries that kept it, like Denmark and Finland, have seen extremely high union membership as a result. Their unions have also been able to do sectoral bargaining with less reliance on government; Nordic governments don’t “extend” contracts as happens in France, as unions can just cut deals themselves using their huge membership as leverage.
And the Ghent system really seems to be what made the difference. A simple comparison between Sweden, where unions run unemployment insurance, and Norway, which abandoned this system, shows that in Sweden, union membership rates kept growing for most of the 20th century.
In Norway, they lagged behind. Since Norwegian unions stopped administering unemployment, Western, the Harvard sociologist, writes, “Swedish union density has persistently exceeded the Norwegian by 20 to 30 percentage points.” Oxford political scientist Bo Rothstein, similarly, has found that adopting a Ghent system leads an additional 20 percent of the workforce to join a union.
That number implies that if the US could turn unemployment insurance over to unions, they could see membership triple from 10 percent to 30 percent, a change that would dramatically transform American politics.
There’s no way that could happen at the federal level with Republicans in charge. That’s where Dimick’s cleverest idea comes in: He thinks that progressive states like California could adopt Ghent systems all by themselves.
The Social Security Act gives states some autonomy in setting up unemployment insurance systems, and Dimick argues that a Ghent system would be an acceptable way for states to implement it. Everyone would be able to collect unemployment insurance, but those who don’t sign up for a union would get fewer unemployment benefits than those who do join.
Most labor scholars and activists I talked to were enthusiastic about the idea of letting unions administer unemployment at the state level. “Independent of union density considerations, there are reasons to believe such an approach would be valuable,” Andrias, of the University of Michigan, says. “As countless union training programs demonstrate, worker organizations can run extremely effective programs to benefit working people.”
The big limitation here, Harvard Law’s Sachs points out, is that the Department of Labor would have to sign off on whatever agency a state wants to have administer its unemployment insurance program. That’s unlikely in the Trump years. But states could also raise their own funds and set up a system without relying on federal unemployment insurance funds.
The Century Foundation’s Strom notes that both a US Ghent approach and sectoral bargaining would need some kind of avenue for workers to voice concerns about their work.
“Sectoral bargaining or a Ghent-like model seem like promising ways to grow the union movement, but because the whole point is to move beyond bargaining at the enterprise level, the focus would no longer be on workers’ struggles for respect at their specific workplaces,” she says. “I think sectoral bargaining would need to be paired with some kind of mechanism for workers to have more voice in the workplace — otherwise we would be losing out on one of the important roles that unions play today.”
Labor is throwing everything at the wall and seeing what sticks
Rolf, the Seattle SEIU president, endorsed the idea of a Ghent system in a paper he wrote for the Aspen Institute. But in partnership with the pro-labor tech billionaire Nick Hanauer, he’s proposed something even more ambitious: “Shared Security Accounts,” a system in which employers would pay for workers to have vacation, sick leave, health insurance, and 401(k) matching benefits that are portable and travel with the worker to whatever job they take.
Rolf sees this plan, which is designed to respond to the rise of quasi-employers like Uber and TaskRabbit, as a way the Ghent system could come to America — if the security accounts were administered directly by unions.
“We intended to replace the permanently employment/firm-based benefits framework with a new framework that’s more worker-centric,” Rolf told me in 2017. As the plan developed, “we took more pains to spell out the idea of a workers’ organization being at the center.”
The benefits Rolf envisions don’t stop with unemployment insurance but also include health, retirement, paid leave, and more. But that could make the system more attractive to workers, not less, and lead more people to enter the orbit of labor unions.
As labor continues to lose membership, sectoral bargaining and the Ghent system are far from the only solutions leaders are considering.
Janice Fine, a political scientist at Rutgers, has proposed something else unions could do: enforce labor laws. Inspectors from the government are often short-staffed and underfunded, and many undocumented immigrants in low-wage jobs are understandably skeptical of cooperating with government authorities of any kind. So Fine has proposed that state and local governments — even the federal government if it wants — contract out to unions and other worker organizations to keep an eye on employers and report abuses.
“There are so many examples of business and the state working together,” Fine told me. “Professional associations set standards all the time. What makes this idea so shocking to people is not that it’s never been done. It’s that it’s worker organizations.”
Baruch College sociologist Hector Cordero-Guzmán suggests that job training (perhaps subsidized by the government) could also be a hugely fruitful activity for unions and other worker organizations, including “alt-labor” groups like worker centers.
Employers often complain that it’s not worth their while to train workers, since they could just be poached by rival companies. Outsourcing the job to unions helps solve that problem. “That employer collective action poaching problem is well-known,” Cordero-Guzmán notes. “When I saw a tweet about Ivanka Trump meeting with Germans about job training, I thought, did they tell her how the training is run through unions? They work with employers to learn what the training needs are.”
In a sense, labor law experts and activists are proposing throwing everything at the wall to see what sticks. The prevailing impression from talking to them is that the total demise of private sector unions in the US is too close at hand to do anything other than try absolutely everything.
“It’s so frustrating to hear all these single-bullet theories,” Rolf says. “Someone says we need card check, someone says striking needs to be a civil right. The reality is that the hour is too late for single-bullet theories. What if we’re wrong?”
Still, it’s remarkable how many different people have all converged on something like a European approach to unionism. As Madland put it, “This feels like an idea whose time has come.”