There are widely circulating but misleading stories about how the employment relationship is changing.

One view is the “end of careers” narrative exemplified by books like Free Agent Nation, The Boundaryless Career, and Escape from Cubicle Nation.¹ These books glorify what they describe as a new world of talent mobility, but they have much in common with a more academic, and less glorifying, literature that also argues that traditional work arrangements are dead. As an example, the labor historian Louis Hyman entitled a chapter of his book on the temp industry “We Are All Temps Now:”² But these views are wrong. The laws and regulations and legal opinions—both federal and state—that sustain work arrangements are extensive and indeed expanding.

The misunderstanding about standard work goes hand in hand with a near obsession with “gig jobs;” where income is earned assignment by assignment often transacted via digital platforms. Car owners making money by transporting passengers and delivering takeout food orders are examples of gig workers, but so are computer coders and copy editors who register with websites such as Upwork and Mechanical Turk. A recent CNN headline blared, “The Rise of Gig Workers is Changing the Face of the US EconomY:”³ A New York Times headline warned that “The Gig Economy is Coming for Your Job:”⁴ Yet the facts do not support this. The US Bureau of Labor Statistics estimated in 2017 that those earning their pay through electronically mediated jobs comprised just 1 percent of the workforce, and in a 2021 survey by Pew, only 2 percent of Americans said they were currently working in either ride-sharing or delivery jobs.⁵

In my own survey, only 9 percent of workers in 2022 were ride-share or delivery drivers or obtained their jobs via online apps. Strikingly, Pew also found that 68 percent of gig workers view that activity as a side job, not their main job. For them, the claim that gig work is displacing standard work is just wrong.

There is, not surprisingly, mythmaking on the other side of the equation as well. Two decades ago, the strategy consulting firm McKinsey described employers as embroiled in a “war for talent;” and this rhetoric persists as CEO after CEO intones that “people are our most important asset:”⁶ The problem is that their behavior does not support these claims. Employers have typically reduced the amount of train ing provided to workers and embraced what I describe as a “disposable workers, not employees” human resource strat egy. Other evidence along these lines is the lack of broad adoption of “high-commitment” management approaches and work systems. This style of managing proceeds from the belief that performance improves when employees are more highly committed to an organization and its purpose. Obtaining commitment is a matter of providing training, giving employees more voice in decision-making, and empowering them to act on their own understanding of how to get tasks done. For at least half a century, management experts have pushed these ideas, as have advocates of “high road” employment practices. The reception has been disappointing, to say the least.

Yet another myth to resist is the belief that the main story about the job market is that technology—robots or artificial intelligence—will eliminate work. Type the phrase “will a robot take my job?” into a search engine and it will return an endless list of claims. Similar arguments are emerging with respect to artificial intelligence. The robot apocalypse never arrived and it is too soon to reach any judg ment about the impact of AI.

I discuss the debates regarding the overall impact of AI on employment levels and also discuss the issue of greatest concern (How will AI affect the mix of jobs between standard and disposable work) in my new book, Disposable Workers, arguing that it is likely that AI will push firms toward an even greater reliance on disposable workers. This said, while technological change is certainly of interest, how work is organized and how managers view workforces are not primarily about technology. Tech nology is a side issue. The real issue is how employers are reconfiguring their relationship with the people who do the work.

EMPLOYEES AND DISPOSABLE WORKERS

To understand what it means to be a disposable worker we need to first understand what it means to be a standard employee. There are two components: being eligible for the benefits and protections that the law mandates for employees, and being connected in a meaningful way to an organization’s career system (as opposed to being seen as disposable and interchangeable).

The starting point is legal but far from being just boilerplate.

If someone is a standard employee, it means their organization withholds their taxes and provides them with a W-2 at tax time, and that they are protected by an extensive list of legally required benefits and “process” requirements. Some of these we can put a dollar value on, such as the Affordable Care Act’s requirements regarding health insurance, and benefits such as Social Security and unemployment insurance contributions. Others are process-related, such as protec tions from age, gender, and race discrimination, or state court rulings regarding unfair dismissal.

These and other features of standard employment add up to what legal scholar Cynthia Estlund has termed “a fortress of rights and benefits:” But these are just the table stakes of what it means to be an employee. Beyond legal and process requirements, being a standard employee means that the entity you work for cares about you and wants to retain you. You are not just a body that is interchangeable with another. A succinct way of stating this is that there is an implicit contract between an employer and its employees, and a set of norms that emerge relating to mutual commitment.

The implicit contract means that the employer invests in train ing, that there are career paths, that the worker has access to the organization’s human resource system and rules and procedures, and that there is an element of mutual commitment between the worker and the organization. All of these features are less precise than the legal requirements for employee status and they are implemented in differing degrees in different organizations. But they add up to the worker’s being seen as a valued member of the organization, rather than someone who can be easily disposed of and replaced.

From the worker perspective, employee status brings the obvious advantages of reasonably stable and secure work and opportunities for skill development. Of course being an employee is not paradise: among other things, organizations are shot full of bias against women and people of color, and the struggles for a fair wage, predictable hours, and fair treatment have never gone away. Though being an employee is far from perfect, on balance it certainly appears—as evidenced by the long political struggle to achieve it—that employee status is a desirable arrangement from the perspective of the workforce and that its imperfections could be addressed without undermining the basic architecture of the system.

Having employees also provides important advantages for organizations. They benefit because their labor supply is more predictable, since they control the career ladders within their organizations and can reduce turnover via backloaded pay and benefits. High turnover is disruptive because it impedes skill development and also makes control difficult, since short-term employees do not master key systems and procedures.

This said, while having employees can be a plus for organizations, it also constrains them. A powerful driver to avoid having too many standard-job employees is that the “fortress” is costly, but beyond sheer cost, managers often find the prerogatives that go along with standard employment inflexible and constraining. Managing employees who are attached to job ladders and career systems means accepting cer tain rigidities that impede the flexibility and agility managers want to achieve. For example, when workers remain stably attached to an employer, expectations arise regarding wage relationships and the employer’s ability to adjust staffing levels.

The bottom line is that having employees is costly and limiting. But the work needs to get done, and this leads to the idea of dispos able workers. “Disposable workers, not employees;” is shorthand for having their cake (obtaining people to do the work) and eating it too (avoiding the responsibilities entailed by standard employee status). My central argument is that this strategy is common, already constituting a surprisingly large fraction of the job market, and will become increasingly widespread. An additional nuance is that some people prefer to be disposable workers, and while employers are the main driver of the change, these evolving preferences complicate both the assessment of and the responses to what is happening.

DISPOSABLE WORKERS

Employers obtain disposable workers through several strategies and work relationships. Some of these are well-defined whereas others are hidden and more subtle but important.

Contracting is widespread and one of the most visible manifesta tions of disposable workers. Contractors are employed by a staffing agency but are assigned to work at another location. Consider the example of building service workers and travel nurses. Note that clean ing buildings is low-paid work and becomes even lower-paid when the work is contracted out; by contrast, travel nurses are highly paid.

The New York Times recently reported that Google’s 130,000 or so contractors outnumbered its 120,000 or so standard employees.⁷ A Washington Post article about National Public Radio was headlined “At NPR, an Army of Temps Faces a Workplace of Anxiety and Inse curity:”⁸ The Wall Street Journal reported that Facebook employed just under one hundred thousand contractors.⁹ And in Boston, the Museum of Fine Arts recently stirred up controversy when it sought to lay off its security staff and replace them with contractors.

Freelancers lack any employer and typically work for multiple cli ents. The mention of freelancing might cause many to picture the prototypical software coder hopping from one client’s project to another. But freelancers are far more varied. Some work for individuals-for example, clearing their gutters, caring for an elderly parent, or editing memoirs. But while these account for more than half the people officially classified as freelancers, they are not relevant to the discussion here because they are not working in organizational settings. My focus is on what I term organizational freelancers—freelancers hired by enterprises to perform tasks that could be handled by standard employees.

Marginal employees is a category introduced in my book. Think of a marginal job as an attenuated or denatured version of a standard job. These workers are employees in the legal sense, not externalized in the way that contractors and freelancers are. At the same time, they are disconnected from the organization’s career pathways and human resource systems and are treated as disposable. Some marginal workers hold part-time jobs, and many are low-wage service workers. The defining feature of these jobs is that their holders are not treated as full members of the organization, as their employers avoid career commitments and expect high attrition in their positions. Indeed, turnover is built into these jobs. They are marginal.

A reasonable objection to the foregoing might be that, in the American labor market, every worker is disposable. The legal regime governing employment is termed “employment at will;” which means that there are very few constraints on layoffs. There are, of course, exceptions to this-including federal judges, tenured fac ulty, and unionized employees for the terms of their contracts—but employment for the vast majority of American workers is always at some degree of risk.

Yet there is no question that some jobs are disposable in a deeper sense than employment at will. The fact that, say, an underperforming employee can easily be fired or that a decline in product demand leads to large-scale layoffs does not mean that the company, when it initially filled the positions, thought of those workers as disposable employees whose contributions could just as easily be matched by any number of other job-seekers. Marginal employees are disconnected from their organizations from the start and investments are never made with an eye to their growth and advancement. This is obviously also true of contractors and freelancers, who do not receive W-2 tax forms from the organizations they assist and are not treated as full-fledged colleagues. But it is also true of marginal employees for whom career progress in the organization is unlikely and continued employment is always uncertain.

THE KIDS ARE ALRIGHT

One thread in the preference discussion is the claim, peddled fre quently in popular discussions but also in some more serious venues, that young people these days do not want “real jobs” but instead are opting to be free agents—that is, to be disposable workers. This is a version of the free-agent nation, boundaryless career, and pack your own parachute rhetoric, but is focused on the argument that the pref erences of young people have shifted in radical ways.

The data do not support this. In surveys conducted by the Bureau of Labor Statistics across the years, the median number of years that workers aged twenty-five to thirty-four had spent with their current employer was 3.0 in 1983, 2.8 in 1996, 3.0 in 2014, 2.8 in 2016, 2.8 in 2018, and 2.7 in 2024.¹⁰ If there is a surge of free-agent behavior, it is not doing much to budge this high-level statistic.

Researchers at Pew also analyzed tenure patterns for millennials in detail and concluded that, to quote the title of their report, “Millen nials aren’t job-hopping any faster than Generation X did:” Notably, they added that job-hopping falls as education levels increase, which is inconsistent with the free-agent rhetoric.¹¹ Finally, Gallup in 2016 surveyed millennials (those born between 1980 and 1996) and asked what they were looking for in work. They did appear to differ from older generations in their preferences for work-family balance and for learning on the job, but their preferences regarding what they wanted out of work were all oriented toward organizational life and not toward free agency.¹²

These conclusions are supported by two National Academy of Sci ence Committees that examined the claim that generational differences in attitudes toward work are significant. In 2020, the committee was highly critical of the quality of evidence behind claims of generational differences and concluded that the few studies with plausible research designs “have found little evidence for generational differences in work values ... .” This finding replicated an earlier 2002 report commissioned by the military that concluded that “the notion of distinct generations with clear differences is not supported by the research:”¹³

LOOKING FORWARD

If you’re interested in diving deeper into how to protect workers, Chapter 10 of Disposable Workers contains a vision for intervention and a menu of strategies: persuasion, advocacy, unionization, legal reform, and pres suring clients who use contractors and freelancers. Not every strategy is relevant for every component of the disposable workforce. The debate about who is an employee is not relevant for marginal workers for whom the issue is instead how to raise the floor for their work. Meanwhile, defining an employee is central to the discussion of gig workers and some other freelancers and, related to this, enforcement is central to the challenges of deliberate misclassification. Unions and the making the National Labor Relations Act effective are strategies not likely to be helpful for freelancers but are central to contractors and marginal workers.

But it has to be understood that writing down these ideas is much like a football coach before a game chalking X’s and O’s on a blackboard in a quiet, controlled locker room. The plays look great in the abstract, but then the players go onto the field and the other team shows up and disrupts everything. It will be hard getting all this done.

It is important to understand that these complexities do not obvi ate the need for new policies.

The first difficulty lies in the heterogeneity of the phenomena. It is clear that, for many people, being a freelancer is desirable. Therefore, any policy initiative aimed at entirely eliminating this choice or mak ing it harder to obtain this status would be misguided. But for other freelancers, such as ride-share drivers, being classified as a freelancer instead of an employee is not a good thing. Yet even here the situation is complicated: about two-thirds of ride-share drivers are part-time and most have health insurance through another job or a spouse. Accordingly, these folks may not be strong advocates for change.¹⁴ One can make the same point about heterogeneity among contractors, where, for example, travel nurses are happy with their situation, but in non-union settings being a contractor for a staffing firm providing building cleaners or security guards is problematic, to say the least.

It is important to understand that these complexities do not obvi ate the need for new policies. For example, in the case of ride-share drivers, the fact that most are part-time and have health insurance from other sources does not mean that the job should not be given standard employee status. After all, as we have seen, nothing prevents a firm from hiring part-time employees and rewarding them differen tially than full-time ones—or, for that matter, offering flexibility with respect to hours or even days of work. The doomsaying by ride-share companies is not credible. But the heterogeneity does complicate our assessment of the problem and considerably complicates the politics of policy reform. The failure to take diverse preferences into account is a weakness in the literature and in policy debates.

The second challenge facing reform is the “simple” one of authority and venue. There are multiple federal agencies with somewhat different definitions of what constitutes employment, and while states can adopt their own standards, they are preempted by federal law when it comes to the law regarding collective bargaining. Federal policy clearly has the advantage of resources, and it also has the advantage of reach. Many of the state and local initiatives have been on the east or west coast, while policymaking around gig workers and contractors and marginal workers is nonexistent across much of the country.

On top of all the foregoing, politics are tough at all levels. Con sider, at the federal level, efforts to expand regulatory interpretation of the Fair Labor Standards Act to classify some freelancers, notably gig workers, as employees. Not only did the Trump administration undo the Obama administration’s efforts, but when President Biden nominated Julie Su as Labor Department Secretary, she was blocked by the Senate on the grounds that she was too aggressive on classification questions. The same thing happened when David Weil was nominated to be administrator of the Wage and Hours division. And it is notable that some Democrats joined Republicans in this.

A similar battle has played out on the state level. In California, the ride-share industry launched a petition campaign to block clas sification of gig workers as employees and spent $224 million in a winning effort.¹⁵ In Massachusetts, Uber and Lyft waged two campaigns to block efforts along the same lines and spent $51 million and succeeded.¹⁶ Even Minnesota Governor Tim Walz, with a strong pro-union record, blocked efforts to reclassify ride-share drivers as employees.¹⁷

Labor law reform is another arena in which politics make it very hard to move forward. There is a broad consensus that the National Labor Relations Act is not delivering a level playing field for unions due to difficulties in blocking or punishing employer tactics that violate the law. At least since the 1995 Dunlop Commission on the Future of Worker- Management Relations, there have been multiple commission reports, lobbying campaigns, and legislative drafts all aimed at updating the legislation. This consensus has added up to nothing: every effort to improve, even to tweak, the National Labor Relations Act has been blocked by unified and fierce employer oppo sition. And this opposition has prevailed regardless of which political party controls the White House or Congress. Indeed, the Washington Post, in the context of reporting on President Biden’s support of a strike by dock workers, commented that this was “a rare aberration in the long-standing bipartisan tradition of siding more closely with management:”18

Nonetheless there is some good news. There is a history of innova tions at the state and local level setting the stage for national progress, and indeed, this is just what happened with the New Deal—when, as an example, Wisconsin established an unemployment insurance pro gram and Roosevelt followed those footsteps. Other examples of creative state and local innovation include standard-setting for ride-share drivers, employmentstandard enforcement mechanisms that partner with community organizations, and sectoral bargaining among unions, employers, and state government representatives sitting at the same table. Of course, all these efforts are small relative to the challenge. Nonetheless, models for progress are being proposed, implemented, and tested, and this sug gests that, when the overall political environment is more welcoming, employees, unions, and advocates might find ways to put boundaries around disposable-workers strategies. Over time, those boundaries could be tightened and the circumstances of people who find them selves in disposable worker status improved. The tools are there; “all” that is needed is the willingness to pick them up and use them.

Adapted from Disposable Workers, published by Harvard University Press. Copyright © 2026 by Paul Osterman. All rights reserved.

About the Author

Paul Osterman is Professor Emeritus of Human Resources and Management at the MIT Sloan School of Management. His numerous books include Good Jobs America, Who Will Care for Us? and The Truth about Middle Managers. Paul Osterman lives in Boston, MA.