Today we hear a lot about the problems of the American middle class: stagnating wages, vanishing pensions, high unemployment, rising income inequality, and the decline of steady work. And we hear a lot of proposals to address these problems by raising the minimum wage, investing in more training, initiating a new stimulus program, and so on. While the problems are real and the proposals are valuable, the issue is more fundamental. As Nick Hanauer and David Rolf have it [“Shared Security, Shared Growth,” Issue #37]—and as I have argued in my work—the problem is that over the past three decades, the entire context of the work experience has changed profoundly. As a result, the social-welfare mechanisms we have devised—institutions that helped create the largest middle class in history—no longer work.

In the past, most middle-class Americans could get jobs that were stable and secure and that served as a platform upon which social-welfare programs such as health insurance, pensions, unemployment insurance, vacations, medical leave, and workers’ compensation were built. This was the social contract of the late twentieth century. Today, jobs tend to be precarious and do not offer the protections and security they did in the past. Workers perform work for multiple entities, sometimes as part-time or transitory employees, sometimes as independent contractors, and sometimes as “partners” in the sharing economy. As a result, working people have variable and stagnant incomes, little job security, no social insurance, and hence, no economic security. Hanauer and Rolf argue that without widespread economic security, there can be no middle class with the purchasing power necessary to sustain a robust and innovative economy.

Hanauer and Rolf call for a “middle-out” economic program to replace the trickle-down economic policies of recent decades. The heart of their program is a “Shared Security System,” in which all employment would carry with it prorated, portable, and universal benefits. Under their proposal, all employers of an individual—whether the employee works part time or full time—would be required to deposit a percentage of the wage into the worker’s Shared Security Account. Over time, the worker’s account would accumulate funds to provide sick leave, paid vacation leave, 401(k) pension contributions, and health insurance premiums. It would also provide social insurance, including unemployment compensation, workers’ compensation, and paid maternity, paternity, family, and medical leave. In addition, employers would be required to pay a minimum wage of $15 per hour and time-and-a-half overtime for those earning up to $69,000. The program would also mandate pay equity between men and women, and fair notice on scheduling.

The Shared Security program is creative and bold. It responds to the problems of atypical and mobile workers, whose multiple, changeable, and intermittent jobs make them ineligible for many social-welfare programs. It is structured to mimic Social Security, in that payroll contributions are paid by whichever entity is paying a worker, and they accrue to the worker’s account even as he or she changes jobs or moves in and out of the labor market.

The proposal echoes some policies that are being implemented elsewhere. For example, the Netherlands requires employers to provide prorated benefits to part-time workers. These benefits include not only social insurance but also paid vacation and sick leave. The European Union’s Part-Time Work Directive requires similar measures in all EU countries. In addition, most European countries require employers to pay for parental leave for all their employees, no matter how many hours they work.

The Shared Security proposals go far, but not quite far enough, to address the problems of the middle class. There are three respects in which the proposals need supplementing: a clear classification of who is an employee and who is not; a plan to reverse the decades-long union decline; and a new benefit to help retrain and sustain the increasingly transitory workforce during the inevitable periods of transition.

The Shared Security plan requires “employers” to make payments into a fund to benefit individual “workers,” but it does not help determine who is an employer and who is a worker. In today’s labor market, this is not a simple matter. Hanauer and Rolf define an employer as “whatever entity is paying the worker,” but their formulation avoids the hard questions in today’s world of quasi-workers, dependent contractors, sporadic gigs, and people in an infinite variety of other atypical work relationships.

The authors describe a hypothetical worker named Zoe, who has a part-time job, drives for Uber, rents out her apartment on Airbnb, and does occasional landscaping work, first via TaskRabbit, then by contracting directly with her clients. They state that all of these activities—even her landscaping business—would enable her to accrue contributions to her fund. In doing so, they assume that Uber, Airbnb, and TaskRabbit are “employers,” for purposes of the mandatory contributions. But that precise issue is hotly contested. Uber claims that it is simply a platform that brings together service providers and service users, akin to eBay. The California Labor Commission recently rejected Uber’s position and held that an Uber driver is an “employee” and entitled to the protection of California’s labor laws because, it found, Uber exerts significant control over drivers’ work performance and dismisses drivers who do not measure up to its performance standards. Even if this ruling is upheld on appeal, it is not clear that it would apply to Airbnb, which exercises relatively little control over those listing on the site. And it is even less clear how TaskRabbit should be classified. These types of service platforms evade traditional categories and require new theorizing about what we mean by a worker entitled to employment benefits and social protection.

Even more problematic is the claim that Zoe’s private landscaping clients are her “employers.” Most people would agree that landscapers are truly independent contractors, as are other people who operate their own businesses and provide services. We may pay a hairdresser, plumber, car mechanic, or lawn-care provider for their services, but rarely do we consider these service providers to be our employees. Would the Shared Security System require homeowners to pay contributions each time they hire a plumber to fix a leaking faucet? Would car owners pay contributions each time they get an oil change? Is someone who buys a necklace at a craft fair the “employer” of the jewelry maker? Indeed, are people who sell an item on eBay to be treated as “employees” of eBay?

Not all service providers are “employees” and not all users of services are “employers.” One of the biggest challenges for labor policy is to determine which service providers are and are not “employees.” As appealing as the Shared Security program is, it bases contributions on employment status, thus making it necessary to devise a definition of who is an employee and who is an employer that is appropriate to today’s complicated labor market.

Another shortcoming of the Shared Security System is that it does not address the decline of unions. Because unions have dropped to less than 7 percent of the private-sector workforce, it is easy to treat them as irrelevant. Yet we cannot fix the problems of the middle class without talking about unions. In fact, since the 1980s, the incomes of the middle three income quintiles in the United States have declined almost in lockstep with unions. This trend is not a mere coincidence. One of the most important determinants of income distribution is the presence or absence of unions. Wages are largely a product of bargaining. Skill scarcity and labor shortages play a role as well, but the role they play is primarily achieved by influencing the bargaining power of individuals and groups of individuals. Thus, to talk about income distribution, we need to talk about bargaining power, and unions are one of the most important mechanisms through which workers can improve and exert bargaining power.

Over the past 30 years employers have been successful at destroying unions. Many employers that had previously been union-tolerant reversed their position in the face of intense global competition. New human-resource theories counseled firms to seek flexibility and treat all resources, including human resources, as variable costs that are expendable when market conditions shift. Union-avoidance techniques became aggressive, widespread, sophisticated, and effective.

In addition, unions themselves have been slow to respond to the changing workplace. Since the 1930s, unions have been workplace-centered, and have protected people in long-term employment relationships. But the workplace has been fundamentally transformed in the past three decades. As a result, workers have different concerns than they once did. Workers, especially younger workers, are no longer expecting to have a lifetime job with a single enterprise. Hence, to them, traditional union goals such as negotiated seniority, job-bidding rules, and just-cause protection are not as important as they were in the past. Many workers ask themselves, “I’m only going to be here a year or two, so why should I take risks involved in organizing a union? I might get sacked or get a reputation as a troublemaker.” And if there is a union in their workplace, some ask, “Why should I pay dues to support a union? Why should I get involved in my union when I could spend my evenings and weekends with my family or my friends? What can a union do for me?” Moreover, traditional union practices have little to offer temporary workers, intermittent workers, on-call workers, project workers, or other atypical workers.

Despite these problems, employee organizations are necessary if we are to create shared prosperity. Thus it is necessary to look at unions from the vantage point of the individual worker and ask, What types of problems do workers encounter today, and how can unions help address them? In some places, unions are changing their traditional practices and finding ways to relate to atypical and transient workers. In other places, new types of workers’ organizations are springing up. Unions have reinvented themselves in the past, and today they need to do so again.

Most people today will experience discontinuities in their careers. They will move in and out of the labor market, with occasional bouts of unemployment, entrepreneurial activities, and time off for care work. In addition, people will need training and retraining at various times during their lives because skills today become quickly obsolete and need to be upgraded periodically. Many are also moving into and out of retirement. One of the most important challenges for social policy is to provide for people during transitions. The Shared Security proposals address the problem tangentially by calling for portable benefits that follow workers as they move from job to job. However, these prescriptions are incomplete. They do not address the inevitable gaps between jobs, nor do they address the need for lifetime learning.

The problem of transitions was addressed by a group of distinguished labor relations experts convened by the European Commission in 1999 to address the implications of the changing nature of work. In 2000, the group issued its report, containing a number of suggestions for changes in the institutions regulating work, of which the most visionary proposal was for the creation of a mechanism to facilitate worker mobility and to enable workers to weather transitions. Under the proposal, an individual would accumulate credits, termed “social drawing rights,” on the basis of time spent at work. The credits are like a hypothetical time bank in the worker’s name that she can use for specific purposes—to obtain training, to work in the family sphere, to perform charitable or public-service work. The amount in the hypothetical account increases each month the worker is in paid employment, and it remains in her account and is available to draw on throughout her life. As legal scholar Alain Supiot writes, “They are drawing rights as they can be brought into effect on two conditions: establishment of sufficient reserve and the decision by the holder to make use of that reserve. They are social drawing rights as they are social both in the way they are established . . . and in their aims (social usefulness).” The concept of the social drawing right is that it is a right that the individual can choose to invoke to navigate career transitions, thereby giving her flexibility and security in an era of uncertainty.

Hanauer and Rolf’s proposal for portable vacation benefits is similar to the concept of social drawing rights, but their proposal is not sufficiently generous, nor is it targeted to assisting with career transitions. They propose that each worker have a right to 20 days of vacation and sick leave combined each year, paid out of the worker’s Shared Security Account. The time off could be used for recreation or illness. The European proposal would require employers to give workers paid time off for vacations, illness, and parental leave, in addition to time workers can accrue for the explicit purpose of career transitions. It is envisioned that the time off for career enhancement would be measured in months, not in days. Australia has a variant of the social drawing rights idea with a benefit called “long service leave” that can give workers several weeks, even months, of time off.

In the United States, we have precedents for the concept of paid time off with reemployment rights under certain circumstances. There are well-established precedents for paid leaves for military service, jury duty, union business, and other socially valuable activities. Some occupations also offer periodic sabbatical leaves. These programs all reflect and acknowledge the importance of subsidized time away from the workplace to accommodate socially useful activities or facilitate a greater contribution to the workplace. They could serve as the basis for developing a more generalized concept of career transition leave or, to use more familiar parlance, a workplace sabbatical. A workplace sabbatical would be a right, accrued by time spent in the labor force, to paid leave for the purpose of retooling, retraining, and repositioning oneself in the labor market.

Hanauer and Rolf’s analysis is incisive, and the Shared Security proposals are exciting. But their proposals need to be supplemented in ways that are consistent with the analysis.

We need to rethink the concept of “employee,” and invent new forms of employee organizations that are responsive to the problems workers encounter today. And we need to provide for people during times of transition in and out of the labor market. With those addenda, the Shared Security System presents a political agenda that progressives should rally around and support.