\n<\/p>\n
The Puzzle of Rising Wages and Low Payroll Growth<\/strong><\/p>\n Here is a question that has been nagging at economists: why has wage growth held up so well when payroll growth has been so weak?<\/p>\n In 2025, the U.S. economy added just 181,000 workers to payrolls<\/strong> \u2014 a number that would have been considered anemic in any recent year. Yet nominal wages grew at an above-average 3.8 percent across the economy and 4.3 percent in manufacturing. On Friday, the March jobs report came in at 178,000 against expectations of just 59,000 \u2014 a blowout that caught the consensus flat-footed. Unemployment fell to 4.3 percent.<\/p>\n While some of the best economists we know are still worried about softness in the labor market<\/strong>\u2014which they blame on what they see as the Federal Reserve\u2019s overly restrictive monetary policy\u2014we\u2019re increasingly persuaded that they may be not be seeing the full picture. If the labor market were genuinely weak, wage gains would be far weaker and unemployment would be rising.<\/p>\n The apparent contradiction has led many analysts to describe the economy as \u201cfrozen\u201d \u2014 paralyzed by uncertainty<\/strong>. Trump\u2019s critics blame tariff policy or the Iran conflict. That diagnosis also looks wrong.<\/p>\n In our view, almost every major indicator is being misread, and the error is the same in every case: analysts are using a model built for a growing labor force to interpret an economy in which the labor force is holding steady and may even be shrinking<\/strong>.<\/p>\n For the entire postwar period, the American labor force grew every year. A growing labor force went unremarked, and the assumptions around growth were almost unnoticed\u00a0\u2014 like the proverbial fish not noticing that water is wet. When we say \u201cthe economy added 181,000 jobs,\u201d we are speaking a language calibrated to a world of labor abundance. The phrase carries an implicit model: there is a growing pool of workers<\/strong>, employers create positions to absorb them, and the number tells you how strong demand is. That model worked for a century. It no longer works.<\/p>\n Since President Trump began enforcing immigration law, the United States has experienced something unprecedented: a rapid, policy-driven contraction<\/strong> in the labor force. The Dallas Fed estimates a net outflow of 548,000 unauthorized workers in 2025. The Department of Homeland Security says that roughly three million people left the country, divided between some 800,000 deportations and 2.2 million voluntary repatriations.<\/p>\n Last week, the Federal Reserve\u2019s own Board staff confirmed what these numbers imply. In a new FEDS Note, economists Seth Murray and Ivan Vidangos argued that labor force growth in 2026 could be near zero<\/strong>, with the pool of available workers growing by fewer than 10,000 per month \u2014 unprecedented in at least 65 years. According to their analysis, the economy could lose 100,000 jobs in a single month, and it would not signal an economic slump. It would be normal noise around a near-zero center.<\/p>\n The most important correction is conceptual. What BLS reports on \u201cjobs Friday\u201d each month is not \u201cjob growth.\u201d<\/strong> It is net payroll growth \u2014 hires minus separations. In the old world of expanding labor supply, the distinction did not matter. The net number effectively isolated demand because supply only pushed in one direction. If payrolls rose 250,000, demand was strong.<\/p>\n When supply turns negative, that identity breaks. The net number becomes the residual of two large forces pushing in opposite directions. A firm can be hiring at full speed and still show a payroll decline if enough workers departed that month. Hiring takes time \u2014 posting, recruiting, interviewing, onboarding. Departures, especially deportations, are immediate. So in any month when outflows spike, the payroll snapshot captures the loss but not the replacement effort<\/strong> already underway.<\/p>\n This means 181,000 net jobs in 2025 was not a purely demand number. It was what remained after gross hiring offset millions of departures<\/strong>. The actual volume of recruiting, interviewing, and onboarding was far larger than the headline suggests. That is why wages grew above average even as payroll growth looked weak: employers were competing hard for a smaller pool. The wage data was reflecting the labor market that firms actually experience.<\/p>\n Manufacturing payrolls fell by roughly 89,000 between April 2025 and February 2026. Critics pointed to this as proof the tariffs are hurting U.S. manufacturiing\u2014or at least not helping much. But roughly 11 percent of what is euphemistically called the \u201cunauthorized workforce\u201d is in manufacturing. Apply that share to three million departures and you get north of 300,000 manufacturing workers who may have been removed<\/strong> from the labor force, nearly four times the reported decline.<\/p>\n If the math holds\u2014and it is just a reasonable inference rather than a precise estimate\u2014the sector may have added over 200,000 replacement or new authorized workers while the headline showed contraction. That would help explain why manufacturing wages outpaced<\/strong> the broader economy despite payrolls shrinking and why job openings in the sector rose by more than 100,000.<\/p>\n It would also explain why jobs declined while gross output climbed 4.76 percent, and equipment investment surged 9.55 percent. These are not the signatures of decline. They are the signatures of adaptation under constraint.<\/p>\n It\u2019s become a commonplace to describe the current labor market as a \u201clow hire, low fire\u201d environment<\/strong>, which suggests a kind of paralysis. But employers are not paralyzed. They\u2019re hiring millions of workers to replace the millions who exited, either by leaving the country or by retiring. Hiring looks low not because demand is weak \u2014 openings and wage data say otherwise \u2014 but because the available pool shrank<\/strong>.<\/p>\n Layoffs have been extremely low. Last week, the four-week moving average of initial jobless claims hit 207,750, which is in the lowest five percent in records going back to 1967<\/strong>. It\u2019s likely that some companies which would be reducing headcount because business is weak have found that ICE has done the work for them. No need to hand out pink slips when your employees are leaving the country or being deported. And deported workers do not file for unemployment insurance. They show up as separations from the employer\u2019s side but never appear in claims data. Strip out policy-driven exits, and the job security of remaining workers is higher than any headline suggests.<\/p>\n None of this means the population contraction is costless. The departing workers were consumers too<\/strong>, and their spending disappeared almost immediately. Productive capacity held up \u2014 through capital deepening, productivity gains, and replacement hiring \u2014 but the customer base shrank, especially for parts of the economy catering to lower income and immigrant populations. That creates genuine disinflationary pressure<\/strong> that the Fed will have to weigh.<\/p>\n But before anyone can debate whether the tradeoffs are worth it, they have to understand what the economy is actually doing. Right now, most do not. A single variable \u2014 labor supply \u2014 flipped from perpetually positive to negative, and it broke every interpretive framework<\/strong> simultaneously.<\/p>\n The immigration reversal changed the meaning of the data before most economists changed the way they read it. Until they do, nearly every conclusion drawn from the headline numbers will be wrong.<\/p>\n<\/p><\/div>\nReading the Economy with an Outdated Map<\/h2>\n
Payroll Growth Is Not Job Growth<\/h2>\n
Manufacturing: The Clearest Case<\/h2>\n
The \u201cLow Hire, Low Fire\u201d Misdiagnosis<\/h2>\n