One strange thing about the American unemployment insurance (UI) system — which provides weekly payments to jobless people who meet certain criteria — is that it’s not insurance against being unemployed. More accurately, it’s insurance against losing a job “through no fault of your own,” which makes UI more like “getting laid off insurance.”
Aside from a few exceptions in some states for things like escaping domestic violence or hostile workplaces, voluntarily leaving your job disqualifies you from receiving unemployment benefits. Allowing people who quit to receive those payments would be “contrary to one of the fundamental tenets of the UI program. The idea is that we want to incentivize people to work,” said Doug Holmes, president of Strategic Services on Unemployment & Workers’ Compensation (UWC), an association that has represented the interests of businesses in matters of UI reform since 1933.
So the point of the American UI system is not to make it easier to quit a job. But a few economists are now beginning to ask: Should it be?
Why I reported this story
A few weeks ago, I was reporting a broader story on the status of unemployment reform now that the dust is settling from the pandemic’s temporary UI boosts. Buried in my list of questions was one that I worried was a little naive. Sheepishly at first, I began asking experts: How ridiculous would it be, really, to extend unemployment benefits to people who quit? Turns out, there’s a pretty serious (and growing) case to be made.
A safety net program that would encourage more Americans to quit their jobs has generally been seen as a bad thing. Quitting takes someone out of the labor market, sapping energy away from the economy. Giving quitters unemployment benefits is like paying them to stay out, bankrolled by taxes that fall on businesses. The result can look a little strange: Businesses could effectively be paying people not to work.
But that story assumes that people who quit stay unemployed. A different story is starting to rack up empirical evidence: Boosting UI generosity doesn’t affect overall employment rates one way or the other. Instead of loafing around in subsidized unemployment, more generous benefits can support people to quit their jobs in search of better ones, which benefits workers through higher wages and better job satisfaction, and the economy through enhanced productivity as people find better uses for their skills.
Put simply, more quitting can be good for the economy. If UI made it easier for more workers to quit their jobs, people would still look for work and the economy could be better off overall. The real losers would be lousy jobs, which would struggle to retain workers with a greater cushion to quit and go looking elsewhere.
Even if one still believes in the core logic of a free market, inadequate unemployment insurance can break it
“It’s pretty novel,” Michele Evermore, the former deputy director for policy in the Department of Labor’s Office of Unemployment Insurance Modernization, said about the idea. “There is a really solid old guard that protects the system as is, but it’s starting to feel like a crumbling retaining wall in some ways at this point.”
Late last year, the question of whether UI should cover quits finally made its way into the bastion of American unemployment research. The Federal Reserve Bank of Minneapolis published one of the first papers to test how extending UI benefits to people who quit would impact the American economy. On squarely economic grounds, they concluded that “giving nothing to quitters is unlikely to be the optimal policy.” That is, their model predicts that giving benefits to people who choose to leave their jobs would grow the economy.
Extending UI to people who quit has hardly been discussed in the 89 years since the program was launched during the Great Depression. There’s now a growing consensus that America’s UI system has been broken for decades, cannot handle another recession, and is in urgent need of a major overhaul. There’s even something of a consensus around what needs fixing.
A major barrier facing lasting reform has been that most people do not care about improving the unemployment system long enough to build the kind of political momentum that gets laws through Congress. Even during major economic downturns, it’s rare to see unemployment rise above 10 percent. If UI were to become a true safety net for all unemployment, covering any American who wants to quit their job rather than mostly people who get laid off, the political base of potential beneficiaries would balloon from a transient and difficult-to-engage minority to, well, how many people in the US do you think are unhappily employed?
Among the many frayed threads that make up America’s safety net, unemployment insurance is one of the most neglected. Reconsidering the airtight conventional wisdom around excluding quitters could turn the program into an engine that nudges the entire economy toward higher wages, better jobs, and the kinds of freedom that capitalism is supposed to create.
Capitalism runs on the freedom to quit — in theory
In Scottish philosopher Adam Smith’s 1776 book, The Wealth of Nations, Father Economics himself describes a society “where there was perfect liberty” as one “where every man [sic] was perfectly free both to choose what occupation he thought proper, and to change it as often as he thought proper.”
Nearly 250 years later, millions of Americans are still a ways off from that kind of liberty. Without financial support, quitting in search of better work just isn’t always a viable option, especially for the more than one in 10 US households that have zero wealth to fall back on.
But the idea is still baked deep into the justifying logic of market economies. If people don’t like something, they’re free to choose otherwise. Capitalism, in the words of economist Milton Friedman and his wife Rose, is the “freedom to choose.”
This was the beauty of capitalism, particularly as seminal economists like Friedrich Hayek saw it. Without a central planner dictating prices or doling out jobs, markets process information produced by the huge variety of voluntary transactions made across the entire economy. Whether you’re buying a carton of organic eggs for $8, accepting a new job, or quitting a crummy one, you’re sending a signal to the market about your preferences. Markets, in Hayek’s view, aggregate these signals across a huge diversity of contexts and coordinate them into a sort of “spontaneous order” that efficiently organizes the economy.
In theory, working a job and buying that carton of eggs are both voluntary transactions. If you don’t like your job, you’re as free to find another as you are to choose a different, cheaper carton of eggs. In practice, especially for lower-wage workers who face relentless economic pressure and lots of debt, adding a job search on top of full-time work just isn’t feasible.
As a result, people trapped in jobs aren’t able to send signals to the labor market that their work sucks and leaves them too drained to find something better. Let this kind of labor market evolve over the course of decades or centuries and you can wind up with an economy full of jobs that make too many people miserable. Without enough freedom to quit, the core logic aligning labor markets with people’s preferences is flying partially blind.
Even if your sole focus is economic efficiency, the economy is much better off when people who quit receive unemployment benefits, too
One way to see the problem is by adopting an idea developed by the economist Albert Hirschman in his 1970 book, Exit, Voice, and Loyalty. He describes “exit” — literally leaving a situation — as the default mode of expressing dissatisfaction in modern economics. His framework didn’t focus on labor markets but has since been used to do just that, scrutinizing how the practical realities of unemployment insurance are out of sync with the theoretical assumptions that workers have reasonable freedom and support to exit bad jobs.
To put this all a bit more bluntly, even if one still believes in the core logic of a free market, inadequate unemployment insurance can break it.
Hirschman noted that even though exit is central to the smooth functioning of a market economy, Americans have a curious tendency to sometimes just ignore it altogether. The labor market is one such situation. The virtue of work — any work — has been held so high that the real freedom to exit lousy work has been ignored, despite that being one of the central mechanisms by which labor markets would adapt to offer better jobs.
Unemployment research has always assumed that quitting is bad for the economy. Now, that idea is under review.
The unemployment insurance system was established during the Great Depression as part of the Social Security Act in 1935, when the unemployment rate was about 20 percent; helping those workers who still had jobs quit wasn’t exactly a policy priority. About half of American workers were excluded from coverage, including agricultural and domestic workers (many of whom were Black).
While the share of American workers covered by UI has shot up above 90 percent (though that figure doesn’t count the self-employed or gig economy workers), the actual share of unemployed people who receive benefits has been steadily declining. Being “covered” only means that you’ll get UI benefits if you meet the additional eligibility criteria. A covered worker who quits their job almost always loses their right to benefits.
In 2022, more than half of all unemployed people didn’t apply for benefits because they didn’t think they were eligible, while the share of the unemployed who wind up receiving benefits continues trending downward.
Although UI has gradually expanded, continuing to exclude people who quit without “good cause” has remained a bipartisan consensus. It guards against what economists call “moral hazard,” or the idea that more generous UI benefits make recipients less interested in getting a job, subsidizing what economists would call “unproductive leisure.”
Many papers through the ’80s and ’90s found that, for example, increasing UI benefits by 10 percent can extend unemployment spells by as much as nearly 9 percent. And this was evidence of a problem: If a government program induces people to quit and stay unemployed for longer than otherwise, it isn’t working.
That belief held for decades. Then, in 1999, economists Daron Acemoglu and Robert Shimer made the case that sometimes a UI program that leads to more quitting and longer unemployment is actually a good thing. In a paper titled “Efficient Unemployment Insurance,” they argued that by reducing the risk of job searching, UI fuels economic growth by improving the job matching quality between workers and employers.
By reducing risks, UI allows workers to look for better jobs, which sends stronger signals to the market about the kinds of jobs workers want. In turn, companies begin investing in more productive jobs with higher wages, challenging the conventional wisdom about UI and economic dynamism. In the years since, further research has bolstered the idea, finding that longer unemployment spells raise overall economic welfare and more generous UI programs can lead to higher-wage, higher-quality jobs.
These papers challenged the idea that a UI program leading to more quits or longer unemployment spells is necessarily bad for the economy but they didn’t question the consensus on excluding quits from coverage. That went relatively unchallenged until late last year.
So should Americans who quit get unemployment insurance?
If more generous UI really can play a role in boosting economic growth — and covering people who quit is a form of increasing the program’s generosity — it’s entirely unclear how much UI is optimal. If inadequate UI breaks the logic of free markets, how much is adequate? Since the question has never been taken seriously in the US, there’s been little by way of actual research.
That changed late last year. The surge of quits during the pandemic and the expansion of unemployment insurance created a unique dataset that caught the attention of economists Zhifeng Cai and Jonathan Heathcote. Evidence from the temporary UI boosts challenged the conventional wisdom around moral hazard. After an extra $600 was added to weekly UI checks, along with a major expansion to who is eligible for the benefits, studies found no connection between the boosted UI and laziness or joblessness (echoing findings around unconditional cash transfers more broadly, where giving people cash doesn’t undermine their desire to work).
When 26 states, largely Republican, pulled out of the federal expansions and rolled their UI programs back to their pre-pandemic levels, the same thing — nothing — happened. Turning down the generosity of UI payments didn’t spur people back to work. Looking abroad, Cai and Heathcote found that on a list of 36 OECD countries, more than half extend some sort of UI coverage to quits (though they all come with a waiting period between quitting and collecting benefits that deters doing so too flippantly).
Of course, the pandemic was a really weird time. So in a staff report for the Federal Reserve Bank of Minneapolis published in October 2023, Cai and Heathcote did as economists do: They took the unusual dataset of quits and built a model of the US economy to put alternative UI policies to the test.
Their results: Even if your sole focus is economic efficiency, the economy is much better off when people who quit receive unemployment benefits, too.
Economists have historically held equality and efficiency at odds with each other, with higher UI benefits seen as an equality booster that trades off against economic efficiency. But Cai explained in an interview with Vox that “if you give nothing to people who quit, it’s actually not an efficient choice, because there are too few people quitting. Our point is that even from an efficiency perspective, you still want to have some UI going to quitters.”
Specifically, their model found that extending UI to cover quits did indeed raise the job matching quality that Acemoglu and Shimer theorized about more than two decades earlier, translating into higher wages and economic growth overall.
That said, their model does find that to maximize welfare across the economy, you’d want to give lower UI payments to people who quit. For people who get fired, the optimal UI amount would be 48.5 percent of their lost wages. For those who quit, it would be 29.5 percent. Both figures are well within today’s average UI replacement rates, which fluctuate between 30 and 50 percent depending on the state. And it’s worth pointing out that even the most generous states cap maximum UI benefits so people who lose high-paying jobs don’t get really high UI benefits, too.
But using those split rates only gets you a very small welfare gain, and it requires the infrastructure to collect and verify information on who gets fired and who quits to ensure people get the right UI amounts. Bureaucracy imposes its own costs, so Cai and Heathcote also report that a universal UI payment set at 38.4 percent of wages — the same amount whether one quits or gets fired — would still deliver most of the welfare gains while avoiding the need to have more bureaucracy.
Cai explained that introducing UI for quitters raises welfare in the economy by up to 1.5 percent. “I know that sounds small,” he said, “but for many other policies that economists evaluate, the welfare gains are like 0.1 percent. So when we got 1 percent or above, we were actually pretty surprised by that. It’s a large welfare gain.”
Starting with a job seekers allowance
If you go looking for research or proposals around extending UI to include quitting, there’s virtually nothing. In recent years, though, the idea has started showing up in UI reform proposals by way of a different program: a job seekers allowance (JSA).
In 2016, the Center for American Progress, the Georgetown Center on Poverty and Inequality, and the National Employment Law Project came together on a proposal to modernize the UI system. The first section focused on things like re-employment services, raising benefit amounts, and financing reform to rescue the program from chronic underfunding. The second proposes establishing a new JSA to cover people who aren’t eligible for traditional UI.
The idea was simple enough: a payment of about $170 per week, for up to 13 weeks, for people who deserve UI coverage but don’t qualify. The stated intent was to cover groups like former students, self-employed people, and independent contractors. Although the report doesn’t emphasize the point, it would also cover anyone who voluntarily quits.
In total, the authors estimate the JSA would have cost roughly $10.9 billion per year (in 2016). Together, the JSA and their proposed modernization reforms would about double the number of unemployed Americans who qualify for some sort of UI or JSA benefit.
The idea of a JSA has now made its way into the bill that gets re-introduced essentially every year by Sens. Ron Wyden (D–OR) and Michael Bennet (D–CO). Their 2023 version of the JSA was largely an echo of the 2016 proposal, except the payment was bumped up to $250 per week and indexed to inflation, and the benefit duration was doubled to 26 weeks.
While the name suggests the JSA would go to people actively looking for work, the proposal only requires that recipients are at least 19, unemployed, ineligible for traditional UI, and don’t have a household income greater than the Social Security taxable wage base, which was $168,600 in 2024. The responsibility for enforcing things like job-search requirements would be left flexible and up to states’ discretion.
That said, the JSA doesn’t quite fit into the bipartisan consensus around UI reform. Most of the rest of the bill does, though. “If you look through all the various reforms that have been proposed over the years, it all goes back to the 1980 and 1996 bipartisan commissions. There’s nothing in Wyden-Bennett that hasn’t been recommended on a bipartisan basis for decades,” said Evermore.
When I asked Holmes about the JSA, he was skeptical. “I could see where there might be a need for a safety net for people that don’t qualify for UI, but you’ve got to pay attention to how it’s funded,” he said.
UI benefits are funded through taxes paid by employers that increase as companies lay off more workers — a situation that should also be reformed. Any rise in UI generosity threatens to put more of a strain on an already cash-strapped system. That, in part, is reason to consider a separate program like the JSA, which could be financed entirely by the federal government, rather than placing additional costs on state UI programs (elsewhere on the list of reforms to consider is just federalizing UI altogether).
Even so, adequate funding seems to be a lesser concern than disrupting UI in general. “We’ve got a program that has worked for going on 90 years,” said Holmes, “and we’re gonna overlay that with something new?”
Holmes doesn’t think that covering quitting through UI would lead to a major change in behavior, just that it would grate against the ethos of the program. On this, I’m not so sure. I don’t know how many Americans are working crappy jobs that they really don’t like, but I think if they all knew they could quit and receive some financial support, a not-insignificant number would. Others point out that if you’re going to do a program that gives cash to people who quit their jobs, why not avoid the incentive to quit altogether by giving that cash to everyone via a guaranteed income. That way, you help out low-wage workers who don’t quit, too (at the cost of a more expensive program).
This is all the more reason to proceed cautiously. We shouldn’t overhaul a major design aspect of a nearly century-old social program on the basis of a couple of papers. That would be a bit hasty. But given their findings, together with the pandemic’s successful experiments in more generous UI, it would also be a bit strange to keep ignoring the idea altogether.
On paper, low unemployment is a signal that the economy is doing well, but low-wage work can really suck. Making it easier to quit, to “exit” a situation we don’t like, can help unearth some of the signals of dissatisfaction that our statistics miss and our markets are blind to. More quitting, it turns out, could be just what capitalism needs.