Pay those workers more. Photographer: Baran Azdemir via Getty Images
Pay those workers more. Photographer: Baran Azdemir via Getty Images

For many years, pay in the U.S. has failed to keep pace with productivity. Indictments of American capitalism, nostalgia for powerful trade unions and many a dumb policy proposal follow directly.

"Since 1973," Harold Myerson tells us yet again, "as unions have weakened, productivity has increased by 80 percent and compensation by just 11 percent." As well as pining for stronger unions, he praises Bill Galston for a plan that, in Myerson's words, can "galvanize Democrats, both centrist and liberal." Firms that fail to share their productivity gains with their workers, says Galston, should have to pay more tax.

Since I almost always agree with Galston -- a moderate Democrat who's accustomed to being beaten up by his own side -- it pains me to say this falls in the category of dumb ideas. Exactly how dumb would depend on the details, but the essential point is simple: What he's proposing amounts to a tax on labor-saving innovation. Taxes on innovation don't promote growth in living standards.

The idea is based, I think, on a common misconception about the link between productivity growth and incomes. The connection at the enterprise level is different from the connection for the economy as a whole.

Suppose a manufacturer replaces 90 percent of its workers with robots. Output per worker surges. In a well-functioning economy, you wouldn't expect that gain in productivity to be divided between owners and the remaining much-diminished labor force. You'd expect the reduction in costs to cut the price of the product, which raises real incomes in the aggregate. Sharing gains in productivity at the enterprise level is not the mechanism that raises living standards in a well-functioning market economy.

The scenario that Galston and Myerson envisage would have strange features. Workers who were otherwise identical would be paid at different rates according to the productivity history of the firm that employed them. If you worked in a modernized enterprise, where productivity had gone up, you would get paid more than if you worked in a new enterprise with no productivity history (free, therefore, to hire at the market rate). There'd be efforts to exploit those anomalies, causing other unintended consequences.

I don't deny there's a real issue here -- even though, as a separate matter, the figures cited for the wage-and-productivity gap are often exaggerated. (According to the Congressional Budget Office, middle-income households saw their after-tax incomes rise by 40 percent between 1979 and 2010: slower than in the post-war golden age, but not nothing.)

Nonetheless it's true that middle incomes are lagging behind growth in productivity. It's also true, in an accounting sense, that the startling rise in incomes at the top is one of the reasons. Slow growth in middle and low incomes is a pressing concern, but taxing higher productivity is a bad way to address it.

(Clive Crook is a member of the Bloomberg View editorial board. Follow him on Twitter @clive_crook.)

To contact the writer of this article: Clive Crook at ccrook5@bloomberg.net

To contact the editor responsible for this article: Tobin Harshaw at tharshaw@bloomberg.net.