What large US companies actually pay in tax, by sector

The aggregate effective rate across these 375 companies was 26.8% in FY2015 and 18.6% in FY2025, and the single largest step down was FY2017, from 25.1% to 21.8% — the Tax Cuts and Jobs Act reaching the income statements. The clearest evidence is not the average but the count in the last column: companies with an effective rate above 25% went from 177 to 56.

What happened afterwards is the more interesting half. The rate bottomed at 16.8% in FY2020 and has drifted up since, without a statutory change to explain it. Some of that is mix, some is the minimum-tax rules that arrived after 2022, and some is companies running out of the credits and one-off benefits that made the first post-reform years look lower than the new normal.

The bottom tail is where the argument usually is: 53 of the 349 profitable companies in FY2025 reported an effective rate under 10%. That is a real number and it is not by itself evidence of anything improper — research credits, stock-compensation deductions and foreign earnings taxed elsewhere all land there legitimately — but it is the number the debate is actually about, and it is rarely published with its denominator.

Two things this is not. It is not cash tax paid: expense includes deferred tax the company expects to owe later, and cash taxes in any single year can be far apart from it. And it is not a rate on economic profit — the denominator is accounting pre-tax income, which is itself the product of choices about depreciation and amortisation. Both make this a good measure of what reduced reported earnings and a poor measure of what a company "should" have paid.