Fault LinesHow the Job Market Will Shape the Next GenerationYoung people are already discontented. It’s not going to get better anytime soon.By Jay Caspian KangAugust 25, 2026Illustration by George WylesolSave this storySave this storySave this storySave this storyYou’re reading Fault Lines, Jay Caspian Kang’s weekly column on politics and the media.Youth sentiment is a notoriously hard thing to track. Polling is a fraught enough enterprise without having to deal with flaky kids, occasionally impenetrable survey questions, and the real possibility that words such as “confidence” or “satisfied” might not mean the same thing from one generation to the next.What I’ve come to believe is that the only way you can genuinely track the attitudes of young people is through annual surveys, which, from year to year, sometimes show bright trend lines. For example, last year forty-three per cent of Americans between the ages of fifteen and thirty-four who responded to a Gallup survey felt positive about their local job market. In contrast, sixty-four per cent of respondents above the age of fifty-five felt that it was a “good time to find a job.” This is a flip from earlier years—until 2024, the younger group felt more hopeful about the job market, and that seemed to make sense, as it’s typically harder to find a job when you’re approaching retirement age.But there has been a downward spike in youth confidence in the past few years. In 2023, seventy per cent of fifteen-to-thirty-four-year-olds said that it was a good time to find a job. The drop since then, according to Gallup, is “similar to the decline between 2007 and 2009 near the time of the global financial crash.”Similar flips can be found all over the polling about young people and their job prospects—and in some of the raw economic data, too. As I mentioned in a column a few weeks ago, the current unemployment rate for recent college graduates is higher than the rate for all workers, and that’s also a flip. This reversal is the result of a trend that started in 2021, after at least thirty years during which the numbers ran in the other direction. If you zoom in on the time line, there’s evidence that these recent shifts have particularly affected the group I call Subaru Socialists: middle-class, college-educated young people. In the fourth quarter of 2024, according to the same updating Gallup survey, forty-nine per cent of college graduates expressed confidence in the job market, compared to forty-four percent of respondents who had less than a college degree. The following year, those numbers flipped dramatically, to nineteen and thirty-five per cent, respectively.What accounts for such a stark change? After all, as I also noted in an earlier column, the unemployment numbers for college graduates grew only mildly worse. Did the start of a second Donald Trump Presidency stoke despair in these voters, who strongly prefer the Democratic Party? Was it all the stories that they were hearing about artificial intelligence?No single one of these trends tells us much about how the future will go, but, together, they do show that something unusual is happening to young people, something different from what their parents and grandparents went through. Many people can feel the difference: in 2020, according to a Pew study, forty-two percent of Americans thought that today’s youth would be better off financially than their parents. In the spring of 2024, that number had fallen to twenty-six per cent.Here’s my read on what has happened in the past five or so years. Although young college graduates still experience relatively low unemployment and make decent wages, they have clear and sound reasons to feel more financially precarious than one might expect. The most obvious and long-standing reason is housing: many young people measure their economic maturity through homeownership, and buying a house is becoming nearly impossible in many major metropolitan areas. The second reason is that entry-level white-collar jobs appear vulnerable to A.I., especially in fields where A.I. is already being used to automate tasks once done by humans. A.I. does not yet account for much job loss, but its arrival has magnified negative sentiments and has helped to create a collective despair, especially among the young and educated.I ended my previous series, on the future of the university, with a series of predictions, which I hoped would be clarifying, even if many of them ended up being wrong. This time, I have five predictions about what the future will look like for young people and the job market.A.I. won’t replace all entry-level white-collar jobs, but it will cause an extended period of pain, during which an increasing number of employers, wowed by new advances in technology, decide to quietly stop hiring many recent college graduates. This will accelerate the already widening unemployment gap between young, college-educated people and everyone else. If unemployment for this group goes from, say, five per cent to ten per cent, their confidence in the job market, already rapidly trending downward, will dive toward a true bottom.What will happen then? Possibly a deepening and widening of populist politics, as both the far right and the far left win over more of the allegedly indolent youth. It seems quite possible that establishment politicians—whatever that means these days—will pivot to the left on key issues, such as military aid to Israel and health care, in an effort to ward off a truly disruptive populist revolt. But we are currently in our sixth year of a Trump Presidency; normalcy certainly is not guaranteed in this country.A.I. job loss will be extremely uneven across industries. It seems perfectly possible, for example, that law firms will decide that, even though an A.I. bot can do most of the work of a junior associate, it’s in their best interests to charge clients for the time of someone who went to Stanford Law School as opposed to a few million OpenAI tokens.For now, the industries that will embrace A.I. are the ones that believe in it. That might seem painfully obvious, but, for at least the next decade, there will be enough resistance to a total robot takeover to delay the clanker revolution. People like to go to work. They like to interact with humans. Many workers are older and have grown used to certain ways of doing business. That won’t all go away just because Claude is standing outside your office waiting for word to kick you out.The split between the A.I. believers and the nonbelievers will persist for a while. Some A.I. evangelists believe that, as the technology continues to improve, more people will come to realize the limitless potential of having a robot do their work, and these robots will produce enough wealth to usher in an era of robot communism, in which everyone gets a stipend that can be used to purchase robot goods and robot services. Or something like that. But many people don’t actually want that world, and do want young people to find meaningful and steady work even if the robots can do it.The media—and social-media—environment is going to get worse. This is the least important of these predictions, but the college-educated job crunch will also affect the tone and content of journalism, and not for the better. Despite Elon Musk’s efforts to turn the entire written internet into Stormfront, the conversation on Twitter and Bluesky still runs through the media and the academy. And the escalating nastiness on these sites comes in part, I believe, from the precarity increasingly felt by workers in what we’ll loosely and somewhat artlessly call “knowledge fields.” These professions weren’t doing particularly well before A.I., and employees in these fields are deeply worried now. The actual ability of A.I. to replace such jobs almost doesn’t matter, because nearly every corporation or university will have someone on staff ready to convince a C-suite executive or a chancellor that, if they do not replace some of their fallible and whiny employees with robots, they will die a quick and humiliating death.Relatedly, the media’s coverage of itself and the academy will become even more annoying, until it consumes much of the industry—soon, every other story will be about a woke professor who spent his department’s budget on the racial dynamics of Labubus or whatever. Perhaps that’s a slight exaggeration, but I do think that the humans left in the industry will only get more self-obsessed and cloistered as the shutters hit the ground.A sense of hopelessness and disconnection will spread, even if the job market rights itself. The Great Recession, as I discussed in my series about the future of college, cut birth rates and delayed the traditional markers of adulthood, and, so far, there has been little to no recovery. Even if the A.I. takeover of entry-level jobs gets sequestered to tech, finance, insurance, and a handful of other fields, I would expect to see an even further delay in family formation, with more young people dropping out of the workforce entirely.It’s worth looking at what happened in Japan in the nineteen-nineties, during its so-called ice age for young people entering the workforce. A lack of opportunity, a prolonged recession, and financial and social instability led to scarring effects that followed them into middle age. Much like today’s situation in the United States, over-all unemployment numbers for young people at the time weren’t horrible in Japan when compared to the rest of the world, but young college graduates were particularly affected, and this, in turn, contributed to delayed adulthood and the phenomenon known as hikikomori, in which disconnected young people, primarily men, withdraw fully from society.Japan sought to address this, in 2003, with the Independence and Challenge Plan for Young People, which opened job cafés, vocational-training programs, and something called Wakamono Jiritsu Juku, a network of camps that took NEETs (Not in Education, Employment, or Training) and put them through, essentially, rehab for deeply unemployed and antisocial youth. The successes of this program are hard to measure, and I am not suggesting that we should build Betty Ford Centers for social-media addicts and gooners in America. But Japan’s experience suggests what can happen when a bunch of young people who did everything right to prepare for adulthood find the fruits of adulthood deferred, en masse. Their disaffection can create a culture of hopelessness and emotional stasis that lasts well past any economic recovery. ?