Delusions of ‘Productivity’

February 11, 1952

One of the grounds of the steel-workers’ union demand for higher wages before the War Stabilization Board is “increased productivity.” The new wage pattern set by the WSB decision will ostensibly apply only in steel, but precedent may impose it on all industry. The principle of “annual improvement” increases in wages therefore deserves far closer scrutiny than it has hitherto received.

1—Most of the clauses in wage contracts providing for “productivity” increases are fraudulent, because the specified annual wage increase must be granted unconditionally, whether or not the anticipated increase in productivity actually occurs.

2—Virtually all labor operates with tools or machines of some sort. There is no such thing as pure “labor productivity”; there is only combined labor-land-capital productivity. The production ascribable to labor cannot be physically separated from the production ascribable to machines.

3—An increase in the productivity of a particular machine, plant, company or even industry does not necessarily imply a general increase in productivity. Or vice versa. The Bureau of Labor Statistics has estimated that from 1939 to 1950 productivity per man-hour (which in longhand means man-machine-method-management hour) went up 23 percent in the bituminous coal industry but down nearly 13 per sent in anthracite. Would John L. Lewis favor, say, raising wages 23 percent in the bituminous mines and cutting them 13 percent in the anthracite mines?

4—The overall increase in man-hour productivity is very difficult to measure. An estimate hitherto widely accepted among statisticians is that in the long-run past, productivity has risen about 2½ percent a year. But there is little evidence that this rate has held in recent years. Estimates recently released by the Manufacturing Census of 1947 show that the physical output of manufacturing was 84 percent higher in 1947 than in 1939—using 1939 “weights”—and 69 percent higher using 1947 weights. But man-hours in manufacturing increased in that same period by 68 percent. Therefore, using 1939 weights, productivity per man-hour increased only 9½ percent in those eight years, or at an annual rate of barely more than 1 percent. And if we use 1947 weights we find no increase in productivity per man-hour at all! Yet vast sums of capital were invested in that period in new tools.

5—Between 1939 and 1947 straight-time hourly earnings increased 89 percent, and gross hourly earnings 95 percent. In the same period wholesale prices of manufactured products rose 82 percent. In the absence of a real increase in physical productivity, higher hourly wages can only be paid, in the long run, out of higher prices.

6—Throughout our history, competition and the free market economy have been solving, infinitely better than any bureaucrat could, the problem of how the gains of increased productivity should be distributed. To the extent that added productivity has been brought about by an increase in the individual worker’s own effort or skill, the gain has gone directly to the worker. To the extent that the gain has been brought about by more capital investment (more or better factories and machines) enough of the gain has gone to the investor to pay an adequate return. But any increase of profits to the pioneers and risk-takers has been transitory. In the long run the gains from higher productivity have been diffused through the whole nation in the form of lower prices than otherwise to consumers. In other words, we have increased the ratio of wages to prices.

7—The proposition that the whole increase in productivity brought about by a new or better machine should go to the particular worker who operates the machine is absurd on its face. It completely contradicts the principle of equal pay for equal work. Its application would give no incentive to any investor or employer to buy or build a machine, no incentive for progress, no gain for consumers. It would not merely be inflationary, but reactionary and disruptive. And no one would be worse hurt by it than labor itself.