The IRS says the average tax refund this spring is $3,571. The White House used Tax Day to declare that “Americans are keeping more of what they earn.” And yet, in the same month, the University of Michigan’s consumer sentiment index fell to 50.8, its lowest preliminary reading since June 1952. One number says households are getting more money back. The other says Americans have not felt this pessimistic about the economy in 74 years. The tension between those two data points reveals something neither figure can explain on its own.
The refund numbers are real
The IRS released its mid-season filing snapshot covering data through March 20, 2026. The agency reported that total refunds exceeded $202 billion, with the average refund landing at $3,571. For context, the average refund during the same window of the 2025 filing season was roughly $3,453, according to prior IRS filing-season reports, making this year’s figure a modest increase rather than a dramatic leap. Electronic filing accounted for about 90% of processed returns, and most refunds were delivered within 21 days via direct deposit.
The White House statement cited a slightly different figure, “over $3,400,” and claimed that 45% of all filers used tax breaks enacted during the current term. The administration attributed a 53-million-filer count to the Treasury Department, and the Associated Press confirmed that figure before the filing deadline. The White House also highlighted expanded credits for families and small businesses but did not release a breakdown by income bracket or filing status.
That missing detail matters. A larger refund does not automatically mean a household paid less in taxes over the course of the year. Refunds can grow simply because an employer withheld more from each paycheck, meaning the filer loaned extra money to the government interest-free and got it back months later. Without IRS data on effective tax rates by income group, the refund figure alone cannot prove that families are financially better off on a net basis.
The confidence collapse is just as real
The University of Michigan’s consumer sentiment index, one of the longest-running measures of public economic mood in the United States, posted a preliminary April 2026 reading of 50.8. The UMCSENT series, maintained by the Federal Reserve Bank of St. Louis, tracks the index over time. Because the underlying data carry a one-month embargo at the source’s request, the most recent publicly available figure on FRED reflects conditions from several weeks prior, but the preliminary reading released by the university itself confirms the severity of the decline.
“We have not seen readings this low since June 1952,” said Joanne Hsu, director of the University of Michigan’s Surveys of Consumers, in the preliminary release accompanying the April data. The survey, whose methodology has remained consistent for decades, asks respondents about their current financial situation and their expectations for the economy over the next one and five years. When the index falls this far, it signals broad unease cutting across income levels and regions.
The survey does not isolate a single cause, but the decline has unfolded against a specific backdrop: escalating tariffs on imports from multiple trading partners, persistent grocery and housing inflation, and uncertainty about whether recent trade actions will push consumer prices higher in the months ahead. Reuters and the Associated Press have both pointed to tariff-driven price anxiety as a significant factor in the April drop. No primary statement from the White House or Treasury has directly addressed the sentiment collapse or attempted to reconcile it with the refund figures officials have been promoting.
A backward-looking payment meets forward-looking dread
A tax refund reflects last year’s income, last year’s withholding, and last year’s tax code. Consumer sentiment captures how people expect prices, wages, and job security to behave in the months ahead. The two measures answer fundamentally different questions, which is why they can move in opposite directions without either one being wrong.
Consider a household that receives a $3,571 refund in March but has watched grocery costs climb steadily. Bureau of Labor Statistics Consumer Price Index data have shown the food-at-home category rising at an annualized pace above the broader CPI in recent reports. A one-time deposit, however welcome, does not change the math on a monthly budget that keeps getting squeezed. That household might cash the refund check and still tell a pollster it feels worse about the economy than it did a year ago.
Mark Zandi, chief economist at Moody’s Analytics, has noted in public commentary that consumer sentiment often diverges from backward-looking income measures during periods of policy uncertainty, because households weigh what they fear losing more heavily than what they just received. That dynamic appears to be playing out this spring: the refund is real money in hand, but anxiety about tariffs, prices, and job stability is shaping how people feel about the months ahead.
What is still missing from the picture
The IRS filing-season data cover only returns processed through March 20. Full-season totals, year-over-year comparisons, and breakdowns by income group will not be available until the agency publishes its complete statistics later in 2026. Until then, any claim about who benefited most from this year’s tax changes is incomplete.
The gap between the White House’s “over $3,400” and the IRS’s $3,571 has not been publicly explained. The administration may be referencing a different cutoff date or a broader pool of filers, but without a technical reconciliation, the discrepancy invites unnecessary confusion.
The 53-million-filer claim also remains difficult to verify independently. The White House attributed the number to Treasury, and the Associated Press reported it, but Treasury has not published a standalone release with the underlying methodology or data tables. Until that documentation appears, the figure rests on a single official assertion passed through political and wire-service channels rather than on a transparent statistical release.
Perhaps most importantly, neither the refund data nor the sentiment index addresses the durability of the tax provisions in question. Several of the credits and deductions the White House highlighted are scheduled to phase down or expire in coming years unless Congress acts. For households trying to plan beyond this spring, the question is not just how large this year’s refund was but whether they can count on similar treatment next year.
Where refund checks and rising bills collide at tax preparers’ desks
Tax preparers across the country have described a recurring scene this filing season: clients see a refund number that looks solid on paper, then immediately start listing the bills it needs to cover. Car insurance premiums, rent increases, and grocery costs have all climbed, and a lump-sum deposit in March does not reset the pressure that has been building for months. The pattern is anecdotal at the individual level, but it aligns with what the sentiment data suggest at scale: a one-time payment does not reshape how people experience an economy that feels more expensive with each passing month.
That disconnect is the story the headline numbers on both sides keep missing. The White House can point to $3,571 refund checks. The University of Michigan can document a 74-year low in confidence. Neither data point captures the full experience of a household that deposited one and still reported the other.
