Why the US economy will muddle through Trump’s tariffs. Probably.

“Soft” survey data is still weakening, and we’re getting good surprises only because expectations have plummeted. Meanwhile, the hard data on spending has been fine.

The good news: There has historically been no relationship between the soft data and future economic growth.

The bad news: There might be this time.

The gloomiest interpretation of the gap between the soft and hard data is that the hard data is out of date. On this view, it is capturing the effects of the boom in inventories as companies stocked up ahead of tariffs, while the soft data captures the collapse in sentiment as CEOs, consumers and investors anticipate tariff-driven price hikes and weaker demand.

Assuming across-the-board tariffs survive the legal challenges—or are reintroduced under other laws—a combination of delayed capital spending and expectations of higher inflation could easily start to show up in weaker spending and job losses, hurting the hard data.

The basic theory is shown by another way of cutting the data, the Conference Board’s leading and coincident economic indicators. The leading indicator has slumped and is close to warning of an imminent recession, while the coincident indicator suggests growth is grinding on regardless. Put simply: It might look sunny, but there’s a storm on the way.

Nonsense, cry the bulls. Don’t look at what people say, look at what they do. Just as in the past, panic rarely becomes self-fulfilling. So long as people keep spending—and retail sales remain decent—there’s no reason to worry. We had a bout of the Trump Derangement Syndrome of which the president’s critics are so often accused, but the soft data will recover once President Trump moves on from tariffs to the tax cuts and deregulation likely to stimulate the economy. Enjoy the sunshine.

Bulls can easily dismiss the leading indicators, which warned of a recession that never appeared from 2022 through to 2024—and then flashed red again before the election.

The uncertainty about the impact of government policy is huge, but I lean toward a middle ground.

On the plus side, the soft data probably isn’t as bad as it looks. Consumer surveys are now so partisan that their results are hard to trust. The purchasing managers’ indexes don’t include retail, which has been reasonably strong as consumer spending drove the economy. And, as Paul Hollingsworth, an economist at BNP Paribas, points out, since the pandemic distortions, seasonal adjustment has behaved oddly.

Even the soft data, bad as it is, has recently begun coming in ahead of expectations, showing just how negative forecasters had become (the surprise isn’t that the soft data is improving—it’s just that economists thought it would be even worse than it is).

True, companies are delaying big capital projects, according to David Garfield, co-CEO of consulting firm AlixPartners, which should hit growth. But, he says, they are going ahead with plenty of smaller ones aimed at improving productivity or the flexibility of supply chains, and implementing machine learning.

One of my favorite, albeit imperfect, economic indicators is the change in manufacturing new orders minus inventories. The idea is that rising new orders point to higher future production, which can be offset by higher inventories, which mean less need for future production.

According to the S&P Global Market Intelligence flash estimates, this comparison fell sharply in May—but only because inventories rose at their fastest in the 18 years it has been running the survey. New orders rose the fastest since February last year. Yes, there might be a bit of a slowdown as inventories are used up, but surely if a recession were on the way, customers would be cutting back their orders in anticipation?

Tariffs ought to slow growth, but if they remain at their lowered levels, shouldn’t kill it. I still worry that Trump will suddenly raise tariffs again, while the “big beautiful bill” sent to the Senate includes a clause allowing heavy taxes on foreigners that would crush inward investment.

But I put less weight on the weak soft data than the bears, and less weight on the strong hard data than the bulls. The U.S. economy can probably muddle through—so long as the government keeps out of it.

Write to James Mackintosh at james.mackintosh@wsj.com