Is the economy really this strong?

Second-quarter GDP grew at an annualised 2.4 per cent, confounding the many who forecast recession and suggesting consumers are still spending freely. On paper the American economy looks unstoppable. Economists who read the national accounts closely are less sure. Gross Domestic Income (GDI), which in theory mirrors GDP, describes an economy that has barely grown at all.

A beat that surprised

The advance estimate beat what most forecasters had expected and was faster than the first quarter's pace. Consumer spending slowed but kept growing, while business investment picked up, helped by a wave of factory construction encouraged by federal subsidies for chips and clean energy. Coming a day after the Fed raised its policy rate to a range of 5.25 to 5.5 per cent, the figure fed the narrative of an economy that shrugs off high borrowing costs. Equity investors took it as fresh evidence for a soft landing.

What the release did not include was the income side. The Bureau of Economic Analysis publishes GDI only with the second estimate of GDP, a month after the first. The latest available readings, for the fourth quarter of last year and the first quarter of this one, both showed income-based output falling while spending-based output rose.

Two measures, two stories

GDP measures spending and GDI measures the income that spending generates, so over time the two should move together. In recent quarters GDI has shrunk or grown at a fraction of GDP's pace. If GDI is the truer read of company cash flow and household pay, then much of the GDP boom is an illusion, puffed up by inventory swings and statistical discrepancies.

In accounting terms every dollar spent is a dollar of income for someone, so the two totals ought to be identical. In practice they are built from different sources. GDP comes mainly from retail sales, construction spending, trade data and business surveys; GDI comes from payroll records, corporate profits, interest and rental income, largely drawn from tax data. The gap between them is booked as the statistical discrepancy, and it has recently grown unusually wide.

Which one to trust

There is research on the question. Jeremy Nalewaik, an economist at the Federal Reserve, argued more than a decade ago that GDI often gives an earlier and more accurate signal of turning points, and that it tracked the start of the 2007-09 recession better than the first estimates of GDP did. The Philadelphia Fed now publishes a combined series, GDPplus, that blends the two, on the view that each contains information the other misses. Neither measure has a monopoly on truth, and the BEA itself now publishes their average alongside both.

That history matters because first estimates of GDP around recessions have been badly wrong before. In 2008 the initial readings understated the collapse in output by a wide margin and were revised down heavily as fuller data arrived. The same could be happening now in reverse: early spending data may be capturing activity that later tax returns will show was weaker than it seemed.

Where the gap comes from

The recent divergence has several possible sources. Corporate profits, a large component of GDI, have come off their pandemic highs as input costs rose and pricing power faded. Inventory accumulation, which adds to GDP when firms stock up whether or not the goods sell, can flatter spending-based output in the short run. And some spending may be financed by drawing down savings or by borrowing, which supports GDP without generating fresh income in the same quarter.

Real incomes also tell a softer story than headline spending. Households have been running down the savings they built up during the pandemic, and consumer credit has kept rising. Spending that is financed from savings and borrowing can continue for a while, but it cannot outlast the income that supports it.

The counterargument

The strongest case for trusting the headline is that the labour market backs it up. Payrolls are still growing by more than 200,000 a month, unemployment is near multi-decade lows, and payroll tax receipts have not collapsed. An economy with that much hiring is unlikely to be flat. It is also worth remembering that last year's first half produced two quarters of falling GDP while employment boomed, and few now think the economy was in recession then. Measurement noise cuts both ways.

That is a fair point, and it is why the column does not claim a recession is under way. The claim is narrower: the reported strength of 2.4 per cent growth rests on the spending side of accounts that the income side does not yet confirm. When two official measures of the same economy diverge this much, the sensible response is to treat the stronger one with suspicion.

Why the Fed should care

The distinction matters for policy. If the Fed believes the GDP figures, it may conclude that its 5.25 to 5.5 per cent policy rate is not restrictive enough and raise again. If GDI is closer to the truth, policy is already biting harder than officials realise, and a further rise would add to the risk of a sharper downturn. Bond investors face the same dilemma: strong growth argues for higher yields, weak income for lower ones.

Equity investors have their own stake in the answer. Corporate profits sit on the income side of the accounts, and they are what share prices depend on in the long run. A spending boom that is not producing matching profits is one in which firms are selling more while earning less on each sale, squeezed by wages and interest costs. If that is the real picture, earnings forecasts built on 2.4 per cent growth are too high, and the sectors most reliant on consumer credit and discretionary spending would be the first to disappoint.

The figures to watch are the second estimate of second-quarter GDP, which will bring the first GDI reading for the quarter, and the annual revisions to the national accounts due in the autumn, which often reshape the recent past. With GDP booming and GDI flat, America's resilience may owe more to the statisticians than to the economy, and this year's recession forecasters may yet be proved merely early. If the revisions pull GDP towards GDI, the Fed will find it has already done enough, and borrowers will feel the slowdown before officials acknowledge it.