THE American economy is in a befuddling state. Firms are on a six-year hiring spree that shows little sign of abating; payrolls swelled by an average of 190,000 a month between May and July.

Competition for workers is pushing up wages. The median pay rise in the year to July was 3.4%, according to the Federal Reserve Bank of Atlanta. Americans are spending that cash; in the second quarter, consumption per person grew at an annual pace of 5.5%, equalling its fastest growth in a decade. Yet real GDP is expanding by only 1.2% a year. The culprit seems to be business investment, which has fallen for three consecutive quarters. It is now 1.3% lower than a year ago—the biggest annual decline since early 2010 when the country was staggering out of the financial crisis. If firms are hiring and consumers are spending, why is investment weak?

Initially, the decoupling was caused by the prolonged fall in oil prices. Cheaper petrol benefited Americans by about $1,300 per household, boosting consumption. Simultaneously, it caused investment in the oil industry to fall by more than half in 2015, as shale oil and gas firms stopped drilling. Investment elsewhere carried on merrily, rising by 4.3%. But the malaise is no longer so contained. Even excluding oil, investment shrank slightly in the first half of 2016 (see chart).

Some of this reflects contagion. When energy firms tighten their belts, their suppliers feel the squeeze. Yet other sectors with little exposure to oil have also pulled back. Financial firms, for instance, invested about 21% less in the first quarter of 2016 than a year earlier. Investment is even down in the consumer-staples industry.

There are three potential explanations for this widespread reluctance to invest. The first is weak demand for the firms’ goods. This explains exporters’ restraint, given lacklustre global demand and a pricey dollar. But it makes less sense at home, with consumer spending strong, and firms happy to hire and to raise wages.

The second is tighter credit. Since the Federal Reserve raised interest rates in December, the average rate banks charge firms to borrow is up by about half a percentage point. After five years of loosening standards, more banks have tightened than eased credit standards for business lending in 2016, according to a Fed survey. In February, financial-market turmoil caused credit spreads in bond markets, the best measure of credit conditions, to surge.

Yet it is unlikely that slightly tighter credit has substantially crimped investment, because American firms are flush with cash. At the end of last year they had $1.7 trillion on hand, enough to pay for Hillary Clinton’s infrastructure plan six times over (though much of this cash is held overseas for tax reasons). Indeed, firms are accumulating cash at the fastest rate since July 2011, according to the Association for Finance Professionals, an industry group.

That leaves the third explanation: that in spite of strong spending, slow trend growth is reducing opportunities for profitable long-term investments. On this view, the recent downturn in business investment was less of a cyclical blip than a sign of things to come.

Economies get bigger when they add people or get more from their existing workforce. America is doing less of both. The Bureau of Labour Statistics projects that the labour force will grow by an average of 0.5% a year from 2014 to 2024, down from 1.2% annually from 1994 to 2004, because of ageing baby-boomers and low fertility. And productivity growth has stalled. From 2005 to 2015, output per hour worked grew by only 1.3% a year, down from growth of 3% a year between 1995 and 2005. In the year to the second quarter of 2016, productivity actually fell, by 0.4%.

Optimists argue that this is part of a lengthy hangover from the recession, which should soon end. One contributor to productivity is the amount of capital—for example, machinery or computers—that each worker has at their disposal. The recession sent this ratio soaring as firms laid off workers and left machines sitting idle. Why would firms invest again before they had replenished their payrolls?

But this explanation is becoming less convincing. The capital-to-worker ratio returned to its long-run trend in 2014 (the last year for which data are available). It is past time for productivity growth to have recovered; instead, it is sinking further.

Pessimists think the productivity problem is chronic. Technological advances, they say, are ever-less revolutionary: Uber is less of an advance than the car itself, the smartphone has not changed office work the way the PC did. Nonsense, reply “techno-optimists”, who foresee huge advances in machine learning and robotics.

For now, the data support the pessimists. The best measure of technological advance is total factor productivity, which measures output after controlling for both the number of workers and the amount of capital. In 2015 it grew by just 0.2%, compared with an average of 1.1% in the two decades prior to the financial crisis.

The economy has gradually become more stale. The number of startups per 100,000 people halved from 160 in 1977 to 80 in 2013, according to data from the Kauffman Foundation, a think-tank.

Disruption thrives in hubs like the Bay Area and New York, but workers have become less geographically mobile (perhaps due to high housing costs in many such areas). Notwithstanding pockets of disruption, the market share of the biggest firms is rising in most industries, suggesting a dearth of competition. A recent analysis by Goldman Sachs, a bank, found that the fraction of workers in innovative occupations, defined as those growing more than three percentage points faster than overall employment, is falling.

Businesses anticipating slower long-term growth cannot be expected to invest much. And politicians cannot easily conjure up technological progress. But they can boost competition, simplify taxes and regulation, and invest in infrastructure and education, all of which would help to raise American productivity.