Randall Wray raises an interesting question about efficiency using a thought experiment:

First, a purely hypothetical one. Let us say we’ve got a society of 150 people, of whom 100 work and 50 are dependent aged and young who do not work. The total output produced by the 100 workers is allocated through some mechanism (perhaps a market, perhaps through equal shares) to the entire population of 150. For simplicity, assume only one product is produced and consumed; call it corn. A study is undertaken to determine the productivity of workers, which finds (horror of horrors!) some workers are less productive than others—measured by quantity of corn produced per hour of work. It is decided to lay-off the 20 least productive workers in the interest of increasing efficiency. We now have 80 workers to produce corn for the population of 150. Since these 80 were more productive than the now unemployed workers, total output did not fall by a full 20%, but clearly there is less output to share. Only if we get the 80 “efficient” workers to work longer or harder will this society be able to consume as much as before. The 20 “inefficient” workers are now producing nothing.

In what meaningful sense have we increased this society’s “efficiency”?

The argument has been that these surplus workers will go do something else that needs to be done, although if they really weren’t productive, they might not do this new activity very efficiently either. But if scarcity isn’t the primary problem we face–we don’t live anymore in a society where half our work force grows food, for instance–and we refuse to grasp the liberating effects of a fiat currency–we are limiting ourselves over artificial, not real constraints–then it seems efficiency über alles shouldn’t be the north star of economic policy.

Though, in reality, ‘efficiency’ isn’t what we’re talking about here, is it? It’s not like we’re driving down sequencing costs so we can spend the money on more even data and analysis. Shipping the fifty year old’s job overseas isn’t creating new opportunities–the gains in ‘efficiency’ are just lining someone’s pockets. We’re not generating new opportunities. All we’re doing is creating surplus labor and driving down wages….oh, I see. Never mind.

As Wray notes, efficiency arguments are ultimately about money, not efficient production in an engineering sense (boldface mine):

Think of a blue collar factory worker at her machine producing widgets. Put her at the top of a pyramid of humans who serve as a support staff. She’s got a doctor and a hospital infrastructure and staff that keep her healthy enough to stay on the job. Accountants with a team of assistants manage her finances and compute her tax bill for April 15. She’s got a pre-school to keep her youngest children off the streets so she can work; and an entire public school system to prepare her oldest kids from age 6 through to 22 or 28 years old so that they, can enter the workforce of the future. And there are construction companies building shopping malls for her, and the roads that will take her from home to work and to shopping. A fleet of jetliners stands ready to fly her to a sunny clime, where pool boys and cocktail waitresses ready things for her vacation. It takes thousands of pink and white collar service workers as well as blue collar workers to keep that worker at her job. And of course, she will use the inputs of agricultural and manufacturing workers to make the widgets at her machine.

What sense does it make to measure her productivity while ignoring all of these inputs? How can we say in any meaningful sense that she “produced” some specific number of widgets per hour of her labor?

Further, and here is the more important point, the widgets that flow out of her machine are not yet commodities. They become commodities only when they exchange for money-denominated IOUs….But not only does she have a huge staff responsible for getting those widgets sold, but it makes no sense to measure her productivity in terms of widgets for the purposes of economic analysis. In other words, not only do the economists misunderstand the engineering concept of efficiency, they also misapply it. In our “monetary production economy” (A.K.A. “capitalism”) what matters are the monetary inputs and monetary outputs: how much money did the capitalist start with, and how much did she/he end up with?

Those market efficiency fairies could care less about widgets—what they want is money.

The capitalist uses the widget factory not to make widgets but to make monetary profits. He can increase profits through a variety of means: pay his workers less; pay his suppliers less; charge his consumers more; work his workers harder; replace the “less efficient” workers with “more productive” ones. In real world production processes—as discussed above—it is difficult if not impossible to assign productivity measures to individual workers. Further, workers are social beings with a great deal of discretion over how they approach the work process. Even mainstream economists have finally recognized this as they introduced concepts such as “efficiency wages” (surprise, surprise, if you pay workers more, they work harder and smarter!), “shirking” (work below capacity in low paid or otherwise undesirable jobs), and “insider/outsider” cooperation (union workers won’t cooperate with scab labor).

Just worth thinking about the next time you hear someone glorifying efficiency–the context in which efficiency is used (and if it’s appropriate) really matters. In other words, conflating efficiency with productivity (as conventionally measured) happens far too often, and is really harmful.