The distribution of income in the United States is decidedly more uneven than in countries across the European Union. But the underlying cause of that inequality is not so much a lower level of redistribution by the US government, but instead the much higher level of US income inequality before taxes and transfer payments. Indeed it is precisely because the US distribution of income is comparatively so unequal that the US tax system is more progressive than other countries–that is, those with high incomes in the US pay a large share of taxes than those with high incomes in other countries, because the high incomes in the US are so much higher than in other countries. The US starts with such a high level of income inequality that it can redistribute more than other high-income countries in the form of taxes and transfer payments, while still ending up with greater inequality.

In a post last week, I referred to these comparisons of inequality and redistribution between the US and nations of Europe as the issue of “predistibution” vs. redistribution. Here, I want to set aside the issue of redistribution and instead look at “predistribution.” That is, I want to discuss a few examples of labor market institutions in European countries that make the distribution of pre-tax, pre-transfer income more equal in other countries–and in particular, institutions that boost the share of income going in the form of wages to the bottom of the income distribution.

The Fall 2022 issue of the Journal of Economic Perspectives (where I work as Managing Editor) takes up this topic in a four-paper symposium on “Labor Market Institutions.” Here are a few thoughts:

Claus Thustrup Kreiner and Michael Svarer discuss “Danish Flexicurity: Rights and Duties.” They describe the basic setting in this way:

The Danish labor market model has come to be known as “flexicurity.” A stated strategy underlying this approach is the so-called “right and duty” principle (in Danish, “ret og pligt”). Unemployed individuals have a right to receive income support and to receive public assistance in getting back into work. But it is also their duty to search actively for jobs, to take on appropriate work, and to participate in active labor market policies. Correspondingly, society has a right to make demands of recipients of income support, but also a duty to help improve their job prospects.

In Denmark, national wages guideline are negotiated for each occupation by representatives of labor and business, with final wages being set in local negotiations at the firm level. About two-thirds of Danish workers belong to a union, and about five-sixth of Danish workers are in jobs where the union negotiations determine their pay. However, Denmark has never had a minimum wage. Moreover, it is easy for employers to hire and fire workers in Denmark, and rates of job turnover are similar to the US.

Denmark has much longer and more generous unemployment benefits than the United States. But in order to receive these benefits, all unemployed workers in Denmark are assigned a caseworker who monitors job search activities and provides support, including setting up connections to government-supported job training, along with short-term practice jobs or subsidized jobs. If the unemployed person doesn’t participate actively, then the generous unemployment benefits can be cut off and the person instead must fall back on the less-generous social support system. About one-eighth of the unemployed are cut off in this way, at least for a time.

Making this system of “active labor market participation” work is costly. Denmark spends about 2% of GDP on this system of caseworkers and job training and subsidized training. The equivalent in the US would be spending about $450 billion a year, with perhaps 100,000 or more job caseworkers, to run such a system. Denmark’s is a small economy operating with high levels of automation and international trade, but with relatively high wages and low unemployment rate for low-income workers. But low-wage workers in Denmark need to develop and improve their skills–because otherwise firms paying the union-negotiated wages can and will fire them.

As an alternative model, Simon Jäger, Shakked Noy, and Benjamin Schoefer discuss “The German Model of Industrial Relations: Balancing Flexibility and Collective Action.” As the authors describe, the German approach is built on a basic tension:

The German labor market is shaped by large-scale collective bargaining agreements containing schedules of minimum requirements for wages, hours, working conditions, entitlements, and promotion criteria for workers in different industries, regions, and occupations, and with different levels of skill and experience. These agreements, typically negotiated at the industry-region level, have broad coverage and create significant standardization in wages and working conditions … At the same time, the collective bargaining system in Germany allows for an unusual degree of decentralization and flexibility in wage-setting relative to the more rigid bargaining systems of many of its European neighbors—and even makes it relatively easy for employers to avoid coverage altogether.

This combination of national guidelines but local flexibility means a really large number of collective bargaining agreements at any one time: apparently 82,000 of them in 2021. About 15% of German workers actually belong to a union, but slightly more than half of all German workers have their wages set by a collective bargaining agreement signed by their employer. In the US, workers vote on whether to unionize; in Germany, individual firms decide whether to join an employer association and participate in union negotiations. The authors point out that this employer flexibility is a defining feature of Germany’s labor institutions. The authors write:

Why do German employers ever join employer associations, thereby restricting their wage-setting discretion? First, membership in an association guarantees employers access to peaceful, coordinated, and widely legitimate mechanisms of dispute resolution through sectoral bargaining. In fact, active collective bargaining agreements preclude unions from strikes pertaining to any matters regulated in the pertinent collective bargaining agreement (Friedenspflicht). Second, membership brings various side benefits, including access to strike insurance, legal advice, lobbying support, and professional networking. Third, employers—especially large ones—may face pressure to join from workers and sectoral unions. Tesla’s 2022 expansion into Germany provides an illustrative example. During the first half of 2022, a new Tesla factory near Berlin has faced several complaints over its wage policies. In particular, wages are low relative to nearby manufacturing firms covered by sectoral agreements (Raymunt 2022), and Tesla has begun raising the wages offered to new hires in an effort to increase recruitment, which has introduced a wage gap between new recruits and identically qualified incumbents (Der Spiegel 2022). Discontented workers have appealed to the local IG Metall (manufacturing union) branch, which has begun publicly agitating for Tesla to enter collective negotiations.

But in Germany, the share of firms that choose to be part of unions and the share workers covered by collective bargaining has been declining, and even within the union framework, the amount of flexibility in local agreements has been rising. The overall belief seems to be that the less-flexible German labor markets of several decades ago had high wages but also high unemployment, while today’s more-flexible markets have lower unemployment, but also lower wage growth and higher inequality. The authors write:

[T]he increasing flexibility of the German system means that Germany is no longer a poster child for strong sectoral bargaining. Bargaining coverage in Germany is middling, and decreasing. The flexibility to which Germany’s strong macroeconomic performance is often attributed involves the omission of large segments of the labor market from bargaining coverage. Germany is now starting to face many of the challenges that its historically more rigid industrial relations system used to suppress: significant increases in earnings inequality, the spread of precarious work, and the gradual expansion of a low-wage sector that is now larger than the OECD average (though still 25 percent smaller than in the United States).

For yet another perspective, Manudeep Bhuller, Karl Ove Moene, Magne Mogstad, and Ola L. Vestad offer an overview of Facts and Fantasies about Wage Setting and Collective Bargaining,” with some discussion of labor market institutions in Norway. They emphasize that intro-level textbook models often assume that wages are determined in a negotiation between a firm and a worker, but in many countries, most wages are in fact determined by a collective bargaining process, not a series of individual bargains. However, these collective processes can differ considerably. They write:

Even economies with the same share of unionized workers (“union density”) or with the same share of workers
whose terms of employment are covered by a collective agreement (“bargaining coverage”) can negotiate their wages rather differently. Countries like Germany, Sweden, and Norway typically have export-led pattern bargaining, in which unions in the metalworking sector and the chemical sector set the path for wage increases in private and public services. Other countries such as Israel, France, and Portugal have much less coordination across types of workers. Such differences in so-called horizontal coordination are important for how centralized the wage
setting is and for how centralization works. Equally important is the level of so-called vertical coordination, reflecting whether wage bargaining takes place at level of the firm, the industry, or the nation. As we shall see, there is also a wide variation across otherwise comparable countries in terms of vertical coordination. …

More than 100 years ago, when the United Mine Workers of America teamed up with the National Progressive Unions of Miners and Mine Laborers, basically every organized miner in the United States became a member of the same union organization. With all substitute workers organized under the same union leadership, the leadership could safely be more militant in their wage demands. When the American Railway Union almost at the same time became an industrial union, organizing all the crafts that worked within the US railroad system, it expanded by organizing workers who were each other’s complements. Consequently, the leadership of the union had to be more careful in its wage demands, as lower activities caused by higher wages to some workers would threaten the employment and wages of many other members of the same industrial union. These two examples illustrate a simple and forceful principle, what we call the Hawk-Dove divide: Coordinating substitutes induces militancy, coordinating complements induces acquiescence.

They argue that across a range of countries, labor negotiations have tended to become more decentralized, both horizontally and vertically. They also discuss what they call “tiered” labor agreements in Norway, where an overarching collective bargaining agreement sets wage floors at the level of lower-productivity firms in a given industry but then allows higher-productivity firms in that industry to pay more.

What does all of this mean for the United States, and in particular for workers with below-average levels of wages and income? What if the US wanted to focus some of its policy attention on “predistribution,” which would raise wages of lower-income workers? After all, a country like Denmark with less than 6 million people and about twice as large as Massachusetts, may not be a workable model for the US economy. In the JEP issue, Suresh Naidu asks, “Is There Any Future for a US Labor Movement?”

Naidu points out that unions remain popular in US survey data, with 70% approval rates. He also emphasizes that US labor law follows a “one establishment at a time” philosophy. He writes:

The National Labor Relations Act protects collective action at the “bargaining unit” level, which is a mix of job categories and geographic establishment. The premise of the law is that the establishment-level bargaining unit is the natural level at which workers share interests, and implicitly, that the barriers to collective action at that level are relatively easy to overcome. However, a hostile legal regime and transformations in employment have invalidated the presumptions on which establishment-level bargaining was built. As a result, the NLRA is as much a legal graveyard as it is a sanctuary for American unions. In the United States, any worker that wants a union cannot just join one, but instead needs to persuade 50 percent of their coworkers …

Naidu points out that in countries where the labor movement has more success than in the US tend to have one or both of two features. One is sectoral bargaining, so that multiple firms and their employees across a sector participate in the bargaining agreement. In this situation, individual firms don’t feel the same need to push back against unions, because they know that their competitors in that sector will have (roughly) the same labor contract. Another is “Ghent-style” unions that take on the responsibility for administering other programs, like health care, training, or retirement.

Supporters of stronger US labor movement have a longer wish-list, of course. For example, they often support legalizing sympathy strikes, where if one firm is arguing with a union, then all the unions for suppliers or transportation needs of that firm might refuse to deal with the firm as well. There could be more aggressive enforcement against unfair labor practices.

Naidu also emphasizes the difficulties of organizing unions in a modern American workplace, with a wide range of workers with different jobs and different skills, different levels of commitment to the company, across different departments, perhaps working different shifts. Old-time US unions of some decades ago were both workplace and social organizations, which tended to reinforce each other. “Social capital” connected to the workplace may have diminished. Naidu writes:

One tentative hypothesis is a decline of social capital created at work as a part of a general decline in social capital, particularly among low-education workers. While convincing evidence of this hypothesis must wait, some suggestive evidence can be found in the General Social Survey data on the share of friends who are coworkers, for the group of private-sector workers with a high school education or less declined from 21.5 to 16.4 percent between 1986 to 2002, while it increased from 17.8 to 19.2 percent for those with more than high school education. The Social Capital Project (2017) published by the Joint Economic Committee writes, based on data from the American Time-Use Survey: “Between the mid-1970s (1975–1976) and 2012, the average amount of time Americans between the ages of 25 and 54 spent with their coworkers outside the workplace fell from about two-and-a-half hours per week to just under one hour.” Union decline might be seen as yet another form of associational life that has declined for all the same reasons other forms have declined. In this sense, the decline of unions may be as akin to the decline of churches as the decline of heavy manufacturing.

Naidu offers this overall judgement:

Rapid increases in union density are like wildfires (or pandemic waves), and I have little confidence in predictions about whether worker organizations will grow, or even persist, in the twenty-first century. If they do, I suspect they will be very different from the labor organizations of the twentieth century. These new organizations, possibly incubated inside or alongside existing labor unions, will depend on government in new and multiple ways, deploy collective action at multiple scales for both economic and political goals, and use and bargain over technology in ways that are hard for any middle-aged academic to anticipate.

For example, he points out that while it may seem difficult to organize gig workers, from the standpoint of the traditional US model of a labor union, when a group of workers is connected by a platform, that platform can also be a target for labor organization. Areas like health care, elder-care, day-care, counseling, education, job training and mental health all offer possible targets for future labor movements.

My own sense is that labor movements of the future need to be increasingly self-conscious about what they are providing to both workers and employers. As Naidu points out, many US unions have become closely identified with all the causes and candidate of Democratic politics, which is nice for the Democrats, but also means that union causes become a hostage to party politics, and will have a hard time succeeding in locations with a substantial number of Republicans. Unions which can make the day-to-day case that they are helping workers build skills, receive benefits, and providing a useful flow of feedback about workplace issues will be more appealing than unions that surface every couple of years to fight a grudge match against employers, and then vanish until the next grudge match. I don’t think the enormous US economy can just import labor market institutions from other countries; after all, those institutions emerged in different places from different historical context. But surely there is something for the US to learn from other high-income nations about how to predistribute higher wages to the bottom half of its highly unequal income distribution.