Both sides of the political aisle are misleadingly bleak about the state of the American worker.

A national Pew survey conducted in April of this year found that 56 percent of Americans believe that “life in America today” is worse than it was “50 years ago for people like you.” In a separate April survey, the exact same percentage rated their financial situation as either “only fair” or “poor.” Even as progressive Democrats and populist Republicans wage an apocalyptic culture war, they largely agree that the economy is failing the American worker and that globalization has hollowed out the middle class.

However, these perceptions of long-run decline are at odds with the facts. Certainly, Americans’ consternation about higher grocery bills and gas prices is valid, as are mounting concerns about housing costs, but it doesn’t paint a full picture of how financial circumstances have changed in just a couple of generations. Earnings trends over the past 50 years show that not only are middle-class earnings up significantly — by $23,000 among men working year-round and $34,000 among women — earnings growth among young men has been stronger during the past 35 years than in the previous 15. These findings belie many popular accounts of what ails the economy today.

It’s easy to draw the wrong conclusion about earnings trends. Economic measurement requires analysts to make numerous methodological choices. Different decisions produce disparate answers to the question of how the American economy is doing, and, in turn, those answers suggest very different kinds of policies. Indeed, my estimates of the change in earnings for the typical man from 1973 to 2023 range from a decline of 18 percent to an increase of 47 percent.

We can’t just pick the estimate that accords with our impressions — we all have biases, live in our own social bubbles, and get information from the sources we seek out. Getting earnings trends right requires nationally representative data and careful attention to issues such as how to account for the rising cost of living and how to value fringe benefits. It necessitates thinking hard about earnings that are taxed away — sometimes without workers’ awareness, in the case of corporate income taxes that reduce not only profits but also employee pay. And the analyst must consider how to address people without any earnings or who work only part-time or only part of the year. Some of them have voluntarily chosen their work status, while others would like to work more than they do.

Taking all of this into account, the typical male worker in 2023 made $55,000, not including fringe benefits and before his individual taxes were deducted. If we are interested in 50-year trends, it is more informative to look just at year-round workers. (This trend is less positive than the trend excluding all nonworkers but more positive than the trend including all of them as below the median.) That raises the 2023 median to $62,000. Accounting for fringe benefits lifts the amount to $70,000, and adding back the earnings taken away by employer taxes leaves the median at $74,000. That is 46 percent higher than it was in 1973.

The story is even more striking for women. The typical woman working year-round makes 121 percent more than in 1973 — an increase from $28,000 to $61,000 in today’s dollars.

Younger men and women saw smaller gains than their older counterparts. Without adding in fringe benefits or income taken away by employer taxes, earnings rose by 40 percent among men and 116 percent among women, but among adults in their late 20s, the increases were only 21 percent for men and 60 percent for women. But that’s still significant growth for young women, and men’s pay growth lagged mostly because their earnings fell between 1973 and 1989. If their earnings had risen as fast in this earlier period as they did after 1989, the increase among young men by 2023 would have been 39 percent.

Social media is rife with unsourced charts depicting stagnant wages and internet memes lamenting the insecurity of “late-stage capitalism.” These posts are as likely to be shared by Tucker Carlson acolytes as by followers of Alexandria Ocasio-Cortez. But the best evidence on earnings trends presents two insurmountable problems for populists who blame supposedly stagnant living standards on economic developments and policy choices over recent decades.

First, nothing about the overall 50-year earnings trends for men or women indicates people had it better in the early 1970s. Earnings are essentially at all-time highs for both. And second, while the worst earnings trend over the past 50 years was for young men, the period that makes this long-term trend look unusually bad ended before the “China Shock” occurred, before immigration’s rapid acceleration in the 1990s, before the Great Financial Crisis, and before income concentration in the hands of “the top one percent” approached its peak.

The most powerful factions on the left and right today — both sides of the populist coin — start with the belief that we are in stagnation or decline, blame the usual targets, pronounce their favored policies, and then search for statistics to make their case. But for anyone who cares about workers rather than expanding their audience, that approach is counterproductive. The way to stronger earnings growth is to start by assembling the best statistics on the merits, assess their implications, determine the factors influencing them, and then craft policies consistent with those explanations. The typical worker is tens of thousands of dollars better off than in the past. We can and should do better, but it will do us no good to act as if we have made no progress.