Economic inequality is the looming issue of the 2016 campaign. The presidential debates have focused on personalities and inflammatory topics like immigration, but the underlying cause of voter frustration is the lack of wage growth.

Bernie Sanders has been the only candidate to focus on inequality, and so far no major policy proposal has emerged on either side to address the issue. But if there’s a single policy idea that could galvanize voters, it could be the call to hike the federal minimum wage, now a paltry $7.25 an hour.

Sanders backs a drastic leap to $15; Hillary Clinton would go to $12. The top GOP contenders oppose any increase at all. Sen. Marco Rubio, for instance, has called the prospect of raising the minimum wage “a disaster.”

Polls, however, overwhelmingly support an increase. Fourteen states hiked their minimums on January 1 (California and Massachusetts led the pack at $10 an hour). Meanwhile, the federal minimum hasn’t moved since 2009. Adjusted for inflation, it is lower than in 1968. A half-century is a long time to wait for a raise.

Under present law, people who work full-time can still fall short of the basic necessities of life. A fulltime mother of two who worked 40-hour-weeks at the federal minimum would gross $15,080 a year–well below the poverty line.

Opponents say hiking the minimum wage would be fruitless and perhaps cost jobs. But while their arguments shouldn’t be taken lightly On balance, though, the critics are wrong. There is a strong economic and moral case for a slow and steady increase.

The reason free-market arguments against the minimum wage fail—at least for moderate increases—is that the labor market does not behave according to the classic supply-and-demand model taught in Economics 101.

If labor markets were simply guided by the invisible hand, unemployment would not exist. In tough times, firms would cut wages rather than lay off workers. This is how the market in hamburgers works. McDonald’s doesn’t stop selling Big Macs in a recession; it lowers their price.

But employers do not cut the price of labor. As Arthur Okun, the economist and Lyndon Johnson adviser, explained, labor markets are governed by an “invisible handshake”—an implicit contract between firms and employees. Firms are reluctant to cut wages owing to an unspoken obligation to their workers. Even if they could find and retrain new workers (a costly business) willing to accept less, they are fearful of upsetting morale. Conversely, most workers do not scour the want ads every day for a job that might pay more, because switching jobs exacts a cost in time and effort.

What does this have to do with minimum wages? Plenty. Perhaps the most persuasive-seeming argument against an increase is that many people who earn the minimum are not actually poor: They are teenagers from middle-class homes.

The flaw in this argument is that hiking the minimum wage has a surprising side effect across the bottom half of the wage scale. Even if teens are the largest group getting the minimum wage, low-wage adult workers benefit when the wage is increased, because firms feel a responsibility to increase pay for workers further up the scale.

This result is inconsistent with classical theory, according to which firms would never raise salaries unless forced to. But it is documented by the newer school of behavioral economics. According to a 1986 paper by eventual Nobel Prize winner Daniel Kahneman, co-authored with Jack Knetsch and Richard Thaler, workers care about more than merely maximizing wealth. Notions of fairness keep them from working at too low a wage, even though they will obviously be poorer if they don’t work at all. Economists call this individual minimum the “reservation wage.” (The reservation wage differs across job categories; a bricklayer will accept a lower wage than a doctor.)

Kahneman, Knetsch, and Thaler showed that people’s notion of fairness is influenced by a “neutral reference point.” For most employees, the usual reference points are the wages of other workers at the same firm. Think about it: If wages in the outside world rose faster than yours, you probably would stay on the job, at least for a while. But if you were the only person at your firm passed up for a raise, you’d likely look elsewhere.

Not coincidentally, when the government mandates a hike for a firm’s lowest-paid workers, employers (governed by that invisible handshake) feel pressure to boost other workers’ wages as well. According to more recent research by another trio of economists, Armin Falk, Ernst Fehr, and Christian Zehnder, after a hike in the minimum, “a substantial share” of higher-paid workers also demand raises. “The minimum wage leads to a kind of ratchet effect in workers’ perception of what constitutes a fair wage,” the authors wrote. In a converse illustration, after a hike in the minimum wage, low-paid workers who do not get a raise tend to work less hard. They suddenly feel underpaid.

A frequent argument against the minimum wage is that the earned income tax credit, which provides tax refunds to low-wage workers, is a better vehicle. But neither policy is wholly adequate. Short of a radical increase, the EITC should be a complement to the minimum wage—not a substitute.

What about the claim that a higher minimum will cost jobs? Alan Krueger and David Card chilled that theory in a now-famous 1994 study, “Minimum Wages and Employment,” which found that New Jersey actually gained restaurant jobs after hiking the state’s minimum relative to neighboring Pennsylvania, which did not raise its minimum. This experiment has been replicated, with similar results in subsequent studies.

The fact that job losses do not occur is less surprising when you look away from the textbook and think about the real world. The typical business owner does not use a spreadsheet to calculate whether the incremental output of a new employee will be worth $7.25 an hour (or $8, or $9). Rather, they hire staff when they think it will enable them to serve more customers. A business that is humming will not sack a worker just because the minimum wage is lifted by a buck.

There’s an important caveat: This detour from strict supply and demand can work only for moderate increases. At some wage (pick a number: $15 an hour, $20), the arithmetic becomes overwhelming. A few cities, such as Seattle, recently have imposed Sanders-level wage hikes; early returns suggest that Seattle has indeed suffered restaurant job losses.

Therefore, most economists favor a moderate increase—which is also favored by common sense. “There are positive effects on morale, positive effects on retention—more people want to work,” says Krueger, who was President Obama’s economic adviser from 2011 to 2013. In effect, a minimum wage nudges employers to adopt a pay scale where they, also, are better off. “It’s not what [classical] economics tells us,” Krueger says. “But companies care about their social hierarchies.”