Today, though, just over half of women in their early 40s with a high-school degree or less education are married, compared to three-quarters of women with a bachelor’s degree; in the 1970s, there was barely a difference. The marriage gap for men has changed less over the years, but there the trend lines have flipped too: Twenty-five percent of men with high-school degrees or less education have never married, compared to 23 percent of men with bachelor’s degrees and 14 percent of those with advanced degrees. Meanwhile, divorce rates have continued to rise among the less educated, while staying more or less steady for college graduates in recent decades.

Plummeting rates of marriage and rising rates of out-of-wedlock births among the less educated have been linked to growing levels of income inequality. More generally, these numbers are causes for concern, since—even though marriage is hardly a cure-all—children living in married households tend to do better on a wide range of behavioral and academic measures compared to kids raised by single parents or, for that matter, the kids of parents who live together but are unmarried.

Whether this can be attributed to marriage itself is a contentious question among researchers, since some studies suggest that what really drives these disparities is simply that those who are likeliest to marry differ from those who don’t, notably in terms of earnings. (Other studies, however, find better outcomes for the kids of married parents regardless of the advantages those households tend to have.) Regardless, it is clear that having married parents usually means a child will get more in the way of time, money, and guidance from their parents.

What’s more, the U.S.’s relatively meager safety net makes the cost of being unemployed even steeper than it is in other industrialized countries—which prompts many Americans to view the decision to stay married with a jobless partner in more transactional, economic terms. And this isn’t only because of the financial ramifications of losing a job, but, in a country that puts such a premium on individual achievement, the emotional and psychological consequences as well. Even when it comes to private matters of love and lifestyle, the broader social structure—the state of the economy, the availability of good jobs, and so on—matters a great deal.

* * *

Earlier this year, the economists David Autor, David Dorn, and Gordon Hanson analyzed labor markets during the 1990s and 2000s—a period when America’s manufacturing sector was losing jobs, as companies steadily moved production overseas or automated it with computers and robots. Because the manufacturing sector has historically paid high wages to people with little education, the disappearance of these sorts of jobs has been devastating to working-class families, especially the men among them, who still outnumber women on assembly lines.

In doing research for a book about workers’ experiences of being unemployed for long periods, I saw how people who once had good jobs became, over time, “unmarriageable.” I talked to many people without jobs, men in particular, who said that dating, much less marrying or moving in with someone, was no longer a viable option: Who would take a chance on them if they couldn’t provide anything?

And for those already in serious relationships, the loss of a job can be devastating in its own way. One man I met, a 51-year-old who used to work at a car plant in Detroit, had been unemployed on and off for three years. (As is standard in sociology, my interviewees were promised confidentiality.) Over that period, his marriage fell apart. “I’ve got no money and now she’s got a job,” he told me. “All credibility is out the tubes when you can’t pay the bills.” The reason his wife started cheating on him and eventually left him, he said, was that “a man came up with money.”

The theory that a lack of job opportunities makes marriageable men harder to find was first posed by the sociologist William Julius Wilson in regard to a specific population: poor, city-dwelling African Americans. (Disclosure: Wilson was my advisor in graduate school.) In later decades of the last century, rates of crime, joblessness, poverty, and single parenthood soared in cities across the country. Many conservatives blamed these trends on a “culture of poverty” that perpetuated indolence, apathy, and instant gratification across generations. Some, such as the political scientist Charles Murray, argued that federal assistance programs made these communities dependent on outside help and discouraged marriage.

Many liberals criticized these “cultural” explanations, pointing out that, among other things, the inflation-adjusted value of welfare and other benefits had been falling over this period—which meant overly generous government aid was unlikely to be the culprit. In a 1987 book, Wilson put forward a compelling alternative explanation: Low-income black men were not marrying because they could no longer find good jobs. Manufacturers had fled cities, taking with them the jobs that workers with less in the way of education—disproportionately, in this case, African Americans—had relied on to support their families. The result was predictable. When work disappeared, people coped as best they could, but many families and communities frayed.

The predicament of today’s working class is no longer just about the decline in manufacturing jobs. A study last year by the sociologists Andrew Cherlin, David Ribar, and Suzumi Yasutake found that in places with relatively large disparities in earnings, parents were more likely to have at least one child outside of marriage. Part of the reason, the researchers concluded, was that these highly unequal areas had little in the way of jobs that paid well and that high-school graduates could get—not just factory jobs, but also lower-level office and sales jobs. What have replaced jobs like that are, for the most part, low-wage service jobs as janitors, restaurant workers, and the like. “The kinds of jobs a man could hold for a career have diminished,” the sociologists wrote, “and more of the remaining jobs have a temporary ‘stopgap’ character—casual, short-term, and not part of a career strategy.” The result: As many men’s jobs have disappeared or worsened in quality, women see those men as a riskier investment.

At the same time, they are not necessarily postponing when they have kids. As the sociologists Kathryn Edin and Maria Kefalas have found in interviews with low-income mothers, many see having children as an essential part of life, and one that they aren’t willing to put off until they’re older, when the probability of complications in pregnancy can increase. For mothers-to-be from more financially stable backgrounds, the calculation is different: They often wait longer to have children, since their career prospects and earnings are likely to improve during the period when they might otherwise have been raising a child. For less-educated women, such an improvement is much rarer.

How do these findings square with those of Autor, Dorn, and Hanson? The authors of the fracking study suggest that the disappearance of good jobs could well have played a crucial role in an initial turn away from marriage, as well as childbirth within marriage. But what had taken over since then, they speculate, was a new set of social expectations: Over several decades, Americans have come to view marriage as less of a necessity, and more of an ideal, and this shift has continued into recent years. Now that singlehood and out-of-wedlock childbirth have shed a degree of social stigma, the theory suggests, an increase in men’s incomes won’t revive norms that have already faded away.

This model may seem to focus unduly on men’s economic prospects, compared to women’s, but that’s actually the point. Americans still on the whole expect men to provide, meaning their worth as partners is more closely tied to their income. In fact, what seems to be decisive in Autor, Dorn, and Hanson’s study is not really whether men’s incomes go up or down, but whether they go up or down relative to women’s. For instance, when competition from China chipped away at jobs in female-dominated manufacturing sectors, such as the leather-goods industry, marriage rates actually increased. As women’s wages fell compared to men’s, the economists argue, marriage was more likely to lead to economic security, and single motherhood became less attractive.

But even if expectations around gender and earnings remain firmly in place, they are clearly changing, likely in response to the reality that, nowadays, women are the primary breadwinner in four out of 10 families. I spoke to a 54-year-old former factory worker in Mount Clemens, Michigan, who told me that her husband’s resentment about the frequent temporary layoffs (which came during slow periods at her plant) eventually spilled over into vicious fights over money. “Anytime I got laid off, he got pissed,” she said. The two later separated. In today’s economy, when oftentimes both partners must pitch in their wages to make ends meet, it’s increasingly hard to see how anyone in the working class has the luxury of sticking with someone without a job—male or female.

* * *

Does it really have to be this way? Must a job—or a lack of a job—shape one’s romantic and family life? When I was doing research for my book, I talked to both Americans and Canadians affected by the retreat of manufacturing jobs, many of whom were separated by just a quick drive across the border between Michigan and Ontario. I was surprised, though, that unemployment appeared to be more toxic to the romantic relationships of the Americans I talked to, who were more likely to go through a separation or divorce following a layoff than my Canadian interviewees were.

To some extent, this reflects cultural differences. As Andrew Cherlin, a sociologist whose research was cited above, noted in his 2010 book The Marriage-Go-Round, Americans tend to place great importance on both marriage and personal autonomy, which is reflected in their very high marriage and divorcerates (higher than in other advanced industrialized countries, including Canada). An intensely individualistic worldview, when applied to relationships, may make someone more willing to end them when their partner doesn’t have a good job; the can-do, competitive values that America rightly celebrates can, when taken to extremes, make relationships seem to be as much about self-advancement as about unconditional love and acceptance.

At the other end of the earnings spectrum, this view of relationships leads well-educated people to search for partners who, on some level, will set them (and their children) up to be financially better off. Increasingly, this means that well educated people marry other well educated people—something that has always been the case, but not to this degree.

All that said, the difference I detected in the durability of Americans’ and Canadians’ relationships following the loss of one partner’s job may also have to do with how the two countries’ social policies shape residents’ views on the stakes of being employed. Of course, some researchers believe that a strong safety net may actually discourage people from getting married in the first place. They point to the fact that in European countries with expansive government programs, there tend to be lower rates of marriage and childbirth within marriage. But it’s unclear whether the explanation is different values, or different policies. In many European countries, for example, cohabiting relationships are often long-term and stable, such that they look much like marriages. In the U.S. that tends not to be the case, which suggests that attitudes about live-in relationships, like views on marriage, diverge across the Atlantic.

He and his wife talked things over, and he decided to get help. A local support program for people out of work—an “action center” funded by the government and staffed by some of his former coworkers at the plant—provided him with a support network of peers who understood his situation. The center also lobbied his former employer to extend his remaining health-insurance coverage so that he could pay for his therapy. (Even under Canada’s single-payer system, not all health-care costs are covered by the government.) He said he emerged from that experience with a stronger marriage and a stronger relationship with his daughter. “Before, we didn’t have that openness, that communication,” he said.

The Canadian safety net later helped him in other ways. He took remedial courses to get his high-school degree and then trained to become an addiction counselor; the government paid all his tuition, which included a job placement at the end of the program. Even when his public unemployment benefits ended, he continued to receive income through a special program for laid-off workers like him who had worked at least seven out of the previous 10 years. The fact that he could still bring home a check every other week, he said, made him feel less ashamed about not working. “Everything is moving in the right direction,” he told me at the time. For that he credited his family, his own motivation, and the government’s help.

Canada has a robust set of policies that help struggling families, especially those with just one earner. For example, Canadian parents receive “baby bonuses,” monthly tax-free cash benefits for each child under the age of 18, which were greatly expanded for lower-income households last year. (America’s federal government offers a child tax credit, but it helps only those who have done a certain amount of paid work that year, and jobless workers and low-income families who don’t pay much in the way of federal income taxes receive less or none of it.) Canadians with modest incomes also receive quarterly, tax-free payments to offset the costs of various sales taxes. Policies like these make having two full-time incomes less crucial in keeping a Canadian household financially afloat. They may also make the relationships in that household less transactional—that is, less dominated by a calculus that tallies what one partner does for another.

Confronted, like the United States, with global economic realities such as free trade and automation, some countries have built or strengthened safety nets to give their residents a measure of financial stability. There’s a reason American family relationships have been shaped so much by labor markets. It’s not a matter of destiny, but policy.