US Workers' Share of Output Hits Record Low as Retail Money Hunts Risk

US earnings season is in full force, with companies reporting second-quarter results and overall performance appearing strong. Eight sectors are reporting double-digit earnings growth, led by energy, communications, consumer discretionary and technology, according to John Butters, vice president and senior earnings analyst at FactSet, the US financial data provider.

Survey data points the same way. The S&P Global US Services PMI, a monthly survey of purchasing managers in which any reading above 50 signals expansion, registered 54.6 in July, up from 51.2 in June and the strongest reading in nine months. Six of seven sectors saw output rise, according to the S&P Global US Sector PMI.

S&P Global cautioned against reading too much into it. Chris Williamson, chief business economist at S&P Global Market Intelligence, said in the release that the largest improvement in demand came from consumer-facing services, where spending rose at a rate not seen in more than four years on the back of the FIFA World Cup and the US Independence Day events. Export orders fell at close to the steepest rate since late 2022, and input cost inflation reached its highest since May 2025. The survey implied annualized GDP growth of 2.3% for the third quarter, following 1.5% for the second.

Stepping back, US corporate profit margins have risen to levels not seen since the 1950s.

While capital thrives, labour lags

While US businesses perform well, the average worker has not. Rising profit margins have been accompanied by slower growth in workers' compensation.

The result is a rising share of national income going to capital, while the labour share declines. The labour share of nominal GDP, which the US Bureau of Labor Statistics (BLS) defines as the percentage of output accruing to workers as compensation, fell to 52.9% in the second quarter from 53.7% in the first, the BLS said on 6 Aug. That is the lowest since the series began in the first quarter of 1947 when the figure stood at about 70%. It held near 63% through the late 20th century.

Other measures put the level differently. Labour's share of gross domestic income sits closer to 51% on Bureau of Economic Analysis data, and at about 54.1% on a measure used by Federal Reserve economists. The direction is the same across all three.

The squeeze is not confined to the size of the slice. Payroll growth slowed sharply through the first half of 2026. After adding 214,000 jobs in March, the strongest month in more than a year, monthly gains fell to 63,000 in May and 20,000 in June before turning negative in July, when payrolls dropped by 23,000 against expectations for an 80,000 gain. Nine of the past 19 months have shown outright declines.

Higher markups, the gap between what firms charge and what production costs them, have been reflected in stronger equity returns even as household sentiment has stayed weak.

The shift in income share between labour and capital is not recent. After a partial recovery in the mid-1980s, the labour share fell sharply following the collapse of the dot-com bubble in the early 2000s. After a brief reversal before the pandemic, it has fallen again since.

The issue is not confined to the US. Apart from the UK, most developed economies have seen similar trends, which points to structural rather than cyclical causes.

Structural drift or a familiar cycle

Not everyone reads it that way. Economists at the Federal Reserve Bank of New York examined whether the post-pandemic fall is a new phenomenon or a continuation of an old one, and concluded it looks cyclically familiar. Writing on the bank's Liberty Street Economics blog on 24 Jun, Richard Audoly, Miles Guerin, Srinidhi Narayanan and Rachel Schuh found the pattern of cyclicality and the contribution of reallocation similar to earlier episodes, with the decline coming almost entirely from payroll shares dropping within industries rather than from output shifting between them. They found little evidence that this episode will evolve differently from past ones.

That matters for how long the effects described below persist. A cyclical decline eventually reverses, though the New York Fed's own historical comparison suggests a reversal would require a longer expansion than the US has managed since 2000. A structural one does not reverse at all.

Wealth has become more important than income for economic standing, and the period of work with an average salary needed to close the gap between bottom quarter to top quarter has lengthened sharply within a generation, John Burn-Murdoch wrote in the Financial Times.

What it means for policy and markets

The changes under way have consequences for both the policy landscape and the investment environment.

In the short run, a weakening labour share weighs on the economy because it reduces consumption, which accounts for roughly two thirds of US output. Consumption growth has been stable in recent months, but consumer sentiment has stayed low, and the payroll figures above give households a reason beyond prolonged elevated inflation and high energy prices to feel that way. Rising investment in artificial intelligence capabilities and data centres has pulled the other way, supporting GDP growth.

Concentrated wealth has driven the push for policy responses. California's Proposition 40, which qualified for the 3 Nov ballot, would impose a one-time 5% tax on the net worth of state residents worth more than $1bn, with 90% of proceeds directed to public health care. Sponsored by the union SEIU United Healthcare Workers West, it fixes residency as at 1 Jan 2026 and measures net worth at 31 Dec 2026, so a billionaire who leaves California during the year would still owe the tax. It faces likely constitutional litigation.

The investment environment has changed too, helped by the widening of retail access through trading platforms, fractional shares, exchange-traded funds (ETFs), leverage and options trading.

People are seeking higher returns because declining income shares offer fewer prospects of building wealth through work. That pushes them towards assets promising higher returns over shorter horizons.

Where the squeezed retail money goes

The widening gap between capital holders and labour will also shape politics, with the US midterm elections the next test. The divisions are sharpened by rising spending on artificial intelligence by the hyperscalers and by ever-larger data-centre investment, which sets technology billionaires against ordinary workers.

The flow data suggests the money is going somewhere more specific than risk assets in general. Combined monthly volume across Kalshi, the CFTC-regulated event-contract exchange, Polymarket and its separate US-regulated arm reached $44.8bn in June, up 75% from $25.66bn in May, according to The Block's data dashboard. Kalshi took $31.5bn of that, up 87% from $16.81bn, against $10.26bn on Polymarket and $3.04bn on Polymarket US. Sandmark examined how World Cup betting pushed prediction markets into the mainstream in July, and that tournament is the main reason for the June figures, so they are evidence of an appetite rather than proof of a durable shift.

Crypto tells close to the opposite story on the year. US spot Bitcoin (BTC) ETFs stood at about $4.44bn of net outflows for 2026 as at 8 Aug, and ether funds at roughly $873mn, even though that week drew $1.1bn into the two combined, the strongest since April. Bitcoin itself is down about 30% this year and roughly 50% from a year ago, and Ether (ETH) has fallen through $1,900. When July's weak payroll print landed, gold and silver rallied, while bitcoin barely moved.

So the pattern is not squeezed households buying crypto as a store of value. It is squeezed households paying for short-horizon, event-based speculation, while withdrawing on balance from the asset that was supposed to be the long-duration alternative to wage-building.

That distinction is significant for anyone positioning around retail flow. It favours the venues that sell defined-outcome risk over those that sell exposure. More awkwardly for crypto, the venue growing fastest is the regulated one that never touches a blockchain, while the onchain platform, where crypto accounted for 37% to 40% of retail volume in the first quarter on a Bitget Wallet analysis of onchain data, is taking the smaller share of the flow.

One possible outlet for a further squeeze is therefore more retail capital into event contracts and other short-duration alternatives, not necessarily into crypto itself. Whether that persists once the World Cup effect washes out of the volume data is the thing to watch.