in imports of real equipment from the surge in AI investment. By our reckoning, around $50bn of the AI-investment boom in 2025 reflected additional domestic production, adding 0.2 percentage points to annualised GDP growth (see chart 2). The AI frenzy has also fuelled America’s stockmarket, the source of the second boost to growth. Between Mr Trump’s election and the end of 2025, the S&P 500 index of large American firms rocketed by 15% in real terms— unusually fast by historical standards. That added $5trn more to household wealth than would have accrued in a typical year. Americans tend to spend a small share of such windfalls. Still, using a conservative rule of thumb that each dollar of equity wealth raises spending by two cents in the first year, this probably raised consumption by $100bn in 2025, or 0.5% of total consumer spending. Given shoppers’ central role in America’s economy, the wealth effect may have added 0.3 percentage points to growth. The third boost to America’s economy has come from Mr Trump’s growth- promoting policies. His administration has eased corporate mergers, ordered federal agencies to cut red tape and loosened constraints on private credit. The tax-cutting bill passed in 2025 injected trillions of dollars’ worth of stimulus into the economy. It also probably improved the economy’s rate of long-term growth, by making pre-existing tax cuts permanent, restoring

firms’ ability to fully expense spending on research and development, and allowing them to depreciate assets more rapidly. All of this encourages investment. On average, the independent forecasts we reviewed—including those by the Congressional Budget Office, Tax Foundation, the Tax Policy Centre and the Yale Budget Lab—estimated that the legislation would add 0.2 percentage points to GDP growth in its first year and 0.4 percentage points in 2026. Combine all three boosts and America’s economy ought to be breaking the decibel-meter. Before the presidential election—and before economists were able to size up Mr Trump’s ideas— the consensus forecast was for 2% growth in 2025. Adding in AI investment, soaring stockmarkets and tax cuts might have got America to 2.7%. That is more than a half percentage point faster than reported growth. Another way to calculate the MAGA tax is to try to capture its economic drag directly. Economists have done this for some of Mr Trump’s policies. According to the Peterson Institute, a think-tank, his tariffs reduced real GDP growth by about 0.2 percentage points in 2025, by squeezing household purchasing power and compressing firms’ profit margins. The Brookings Institution, another think-tank, estimates that in 2025 the president’s deportations and border shutdown turned net migration negative for the first time in at least half a century. This reduced labour supply and, since migrant workers spend money, consumer demand. All this may have slowed growth by 0.2 percentage points. Such figures are instructive, but they do not capture the full cost of uncertainty stemming from Mr Trump’s erratic policymaking. Tariffs are announced, delayed, revised and revived. Immigration agents are deployed, recalled and redeployed elsewhere. Wars are waged. An index of economic- policy uncertainty developed by Scott Baker of Northwestern University and colleagues rose by over 100 points from before Mr Trump’s election to the end of 2025. Swings of that magnitude are typically followed by growth in business investment slowing by five to ten percentage points, as firms postpone capital spending and supply-chain adjustments.

Indeed, strip out the splurge on information-processing gear and software— the categories most closely tied to AI—and the picture looks grim. Over the past four quarters non-residential fixed investment, excluding AI-related categories, has contracted at an annualised rate of roughly 3%. It grew by over 5% per year during the previous decade (see chart 3). Investment in industrial and transport equipment has fallen by more than 2% over the past year. Manufacturing construction is down by 20%. In total, non-AI investment is running about $130bn below its trend from the last decade. This capex recession is reducing GDP growth by 0.4 percentage points. Could AI itself explain the weakness? The contraction in non-AI investment is too large and broad to be down to firms merely reallocating capital from other sectors towards data centres. It spans oil and gas, carmaking and factory construction. Trade-policy uncertainty probably played a large role. In a survey a year ago, the Federal Reserve Bank of Atlanta found that 45% of executives planned to cut capital spending owing to policy uncertainty. Another potential explanation—that strong demand and heavy government borrowing have pushed up interest rates and crowded out private investment —is also unpersuasive. Credit remains plentiful. Investment-grade corporate borrowing has seldom been so cheap, compared with Treasury yields, since

the 1990s. It is therefore a good bet that presidential policymaking is to blame for sour sentiment. Together, the squeeze from tariffs on real incomes, reduced labour supply and capex-shy companies adds up to 0.8 percentage points. That is in line with the earlier figure we arrived at by comparison with a counterfactual American economy (see chart 4). Looking ahead, there is little sign of relief. Tariffs remain in flux, sustaining high uncertainty for businesses and households. The war in Iran and the closure of the Strait of Hormuz have triggered an energy shock that will further compress real incomes and corporate margins, dampening investment even more. A natural reaction to such figures is to despair at how much damage bad policies can cause. Another, though, is to marvel at the awesome power of America’s economic engine. Despite everything Mr Trump has put in its way, GDP may grow at an annualised rate of 4% in the current quarter, if you believe the latest real-time forecast by the Federal Reserve’s Atlanta branch. Without the deadweight of the MAGA tax, in other words, America might be rocketing ahead with annualised growth of nearly 5%. It has notched up such performance in just nine quarters this century, and only five if you exclude the recovery after the covid-19 pandemic. If only the president would let it, it could be doing so again. ■

For more expert analysis of the biggest stories in economics, finance and markets, sign up to Money Talks, our weekly subscriber-only newsletter. This article was downloaded by zlibrary from https://www.economist.com//finance-and-economics/2026/05/18/how-much-is-donald- trump-costing-americas-economy

Finance & economics | Loss leaders The insurers on the hook for war in Iran Some have been hammered; even those who haven’t might be soon May 21st 2026 The insurance industry exists to contemplate worst-case scenarios, but the events of the past three months have nevertheless been hard to take in. The Iran war has unleashed havoc in the Middle East. It has also presented two classes of specialty insurer—those offering protection against political violence and marine war, which covers ships in conflict zones—with their biggest shock in decades. The impact on those who insure against political violence has been most dramatic. Such policies protect businesses from terrorist attacks, sabotage, civil unrest, war and other nasty events. Since Iran started firing missiles at its neighbours, local firms in the Gulf have scrambled to find coverage for exposed infrastructure and property. Their clamour, and the higher risks they

now face, has transformed the market. Insuring some assets now costs 40 times the pre-war rate. Some of this is because previous prices were too low, says Fergus Critchley, head of terrorism and political violence at WTW, an insurance broker. A run of strong performance in the sector had brought in competition, pushing down underwriters’ premiums. They had room to fall, in turn, because few foresaw a conflict on anything like the present scale. Claims are thought to run into the billions of dollars, with many stemming from attacks on energy infrastructure. For a niche industry that generates about $1.2bn of annual income from premiums, the resulting losses may erase years’ worth of revenue. Many insurance contracts also promise reimbursement for indirect losses, such as interruption of business. The Gulf is packed with offices belonging to multinationals, so such payouts will be steep. The market will tighten as a result. Contracts now being signed are dearer than in peacetime and offer narrower coverage. What if peace is declared? Mr Critchley says that premiums would probably stay at their current, expensive levels for at least three to six months. In part this is because any news of a peace deal would meet significant scepticism. Even if a lasting one is struck, insurance would not be as cheap as before the war. Oleksii Omelianchuk of FortuneGuard, a firm that assesses conflict risks, points out that insurers who did not price in the risk of the present war are unlikely to make the same mistake twice. Whether war or peace prevails, political-violence insurers may soon scale back. They typically negotiate their contracts with reinsurers—to whom they offload some of their risk—each year on January 1st. When this date rolls around in 2027, reinsurers may well reduce the coverage they offer and raise prices, just as insurers are doing already. Oliver Martin, head of political violence at Atrium, an underwriting syndicate, says that could deter insurers from signing bumper policies today, since they may not be able to reinsure them after next year’s renegotiation. Prices have also risen sharply for insurance against marine war. For ships attempting to traverse the Strait of Hormuz, premiums for such protection are now often between 10 and 20 times the pre-war rate, according to WTW.