US: Democrats and Republicans share a delusion about productivity and wages

US wage

Progressive Democrats and “national conservative” Republicans are increasingly aligned on the failure of American late-stage capitalism. Real wages have been stagnant for decades, they say, even as productivity keeps going up. Economic growth is channeling bigger profits to owners of capital while failing to raise living standards for ordinary families. They don’t always agree on what to blame for this dysfunction – how much is due to free trade, say, as opposed to labour-saving technology, monopoly power or the unbounded greed of the capitalist class – but they no longer quarrel much about the disease.

This is a problem, and not only because the claims about wages, growth and how they relate to each other are wrong. That’s the least of it. The new consensus draws attention away from issues that actually need addressing. Worse, it suggests treatments that will cause the very sickness they’re meant to cure.

The core of the anti-neoliberal consensus is the apparently yawning gap between growth in pay and rising productivity. No matter how often this supposed decades-long divergence has been debunked, the rebuttals keep bouncing off. The story of capitalist exploitation is just too compelling.

Back in 2015 Harvard’s Robert Z. Lawrence explained that the apparent gap between wages and productivity after 1970 mostly disappears once you measure things consistently – meaning, once you compare wages and productivity for the same workers; add non-wage compensation (such as health insurance and employer-paid payroll tax) to earnings; deflate nominal wages using producer prices not consumer prices (so that real wages are measured against the prices of the goods workers actually make); and deduct depreciation from gross production (since output that merely replaces worn-out structures and equipment isn’t available for other uses). The resulting metrics of net output per hour (“productivity”) and real product compensation (“wages”) rise more or less in tandem.

Lawrence noted ten years ago that a smaller gap did seem to emerge after 2000, and he set about explaining it. (Most likely, too little investment, he thought.) But a new study by the American Enterprise Institute’s Scott Winship finds that this later gap has faded away. One of Winship’s various like-with-like comparisons is especially striking, because it uses data that go back to 1929. Between then and 2022, net value added in the non-farm business sector grew by a factor of 23; over the same period, workers’ compensation in the same sector grew by a factor of 23. Drawing on this and other such results, he reasonably concludes, “Overall growth in worker compensation has kept pace with overall productivity growth.”

Granted, this conclusion does present a puzzle: The finding that consistently measured pay and productivity have mostly risen together is at odds with the claim that owners of capital are collecting more and more of the country’s income — another widely adopted belief about dysfunctional capitalism. This alleged fact that the capital share is persistently rising and the labour share persistently falling was most notably celebrated in Thomas Piketty’s best-seller, Capital in the 21st Century. Again, though, it’s an artifact of mismatched data. As MIT’s Matthew Rognlie explained, once you look at output net of depreciation, there’s no longer a steady upward trend in the capital share, and the remaining fluctuations in net capital income are mostly driven by the value of housing (an asset widely held by ordinary Americans, as opposed to “capitalists”).

Despite these inconvenient statistics, the idea that labour must inevitably lose ground to capital is tenacious. That’s partly because it fits the narrative but also, one must admit, because it’s plausible. After all, economies invest and capital accumulates. As the amount of capital per worker goes up, it stands to reason that capital will collect a bigger share of the resulting output. Yet however intuitive, this presumption is wrong. There’s no general law of economics that says capital’s share of output should rise, fall or stay the same as capital accumulates.

Its share depends, in part, on how easily capital can be substituted for labour. Other things being equal, if new equipment can be easily substituted for workers, you’d expect additional capital to lower labour’s share of output. But new investment often requires more labour, not less. (Build a new warehouse and you need people to staff it.) In such cases capital and labour are complements, not substitutes, so additional capital tends to raise labour’s share. There’s a further complication – namely, how far technological progress (for any given amount of capital and labour) is biased toward displacing workers as opposed to expanding the range of tasks for which they are required. People assume that the first always overwhelms the second. Economic history says otherwise.

It so happens that the shares of capital and labour have been pretty stable over time, but the balance of these complex forces can and does fluctuate. Meanwhile, though, the fallacy that capital is relentlessly crushing labour distracts attention from problems that need addressing.

Growing inequality is a fact, even though the depredations of capital are not the cause. Productivity growth is increasingly concentrated in particular industries, generating inequality within the labour force. Median earnings have grown more slowly than average earnings – and in a labour force with more low-paid workers, more high-paid workers, and a small number of extremely high-paid workers, median earnings (at the middle of the income distribution) are no longer so “typical.” Other labour-force shifts are also of concern. Women’s wages have been rising fast, which is good, but men’s wages haven’t, which is not so good. In these and other ways, the problem isn’t capital versus labour, but labour versus labour. What’s needed are policies that grapple intelligently with the resulting inequalities.

Even more important is to restore the centrality of productivity growth for economic policy. If the pay-and-productivity gap were real, measures to raise productivity would be beside the point – and the slowdown in productivity growth over recent decades wouldn’t much matter. But the productivity slowdown does matter. When it comes to raising living standards over the longer term, nothing else counts. More investment – more capital – would raise productivity. Better schools, better infrastructure, stronger competition, liberal trade and a smaller burden of pointless regulation (occupational licensing, paralysis by permitting) would all raise productivity. In other words, they’d raise pay.

So keep firmly in mind that Marx was wrong. Capital isn’t labour’s enemy. It’s the indispensable means to making workers better off.

Credit: Bloomberg