What's in a name?


In the case of utility, it’s all in the name.

In the eighteenth and nineteenth centuries Jeremy Bentham’s ‘utility’ was a general measure of the welfare of society. But then, during the last quarter of the nineteenth century ‘utility’ was reinterpreted as a measurement of the mental state of an hedonistic economic subject in the so-called neoclassical revolution. With von John Neumann and Oskar Morgenstern’s Theory of Games and Economic Behavior (1944), the meaning of utility changed again, from a psychophysical measure of the individual’s preference for different goods to an axiomatically defined intensity of the individual’s preferences. Morgenstern in particular realized that using utility this way was somewhat unfortunate as it suggested a link to earlier economic research that did not exist. But no better label suggested itself.

As a result, economic contemporaries of von Neumann and Morgenstern such as Paul Samuelson and William Baumol, mistakenly understood the Theory of Games and Economic Behavior to claim that neoclassical economics’ utility could be measured numerically – or ‘cardinally’ as they liked to put it after Hicks and Allen.

In turn, that drew attention to Daniel Bernoulli’s essay Specimen theoriae novae de mensura sortis, published in 1738. All of a sudden, the clouded landscape of the past revealed an essay that seemed very relevant for present discussions. As few economists could read the Latin original or its German translation of the early nineteenth century, Econometrica editor William Simpson took the initiative to produce an English translation.

Initially, that seemed easy enough. Early 1952, Simpson commissioned Latin professor Louise Sommers of American University (Washington DC) to translate the text, with the purpose of publishing it later that year. But as Sommers was not an economist, Simpson first asked Baumol to check a few economic words and concepts in the translation to see if they fitted contemporary economic jargon.

Baumol quickly responded that “the problem of rephrasing the translation is a minor one to say the least. As it stands, I am afraid it is almost completely unintelligible” (Baumol’s letter to Simpson, 6 November, 1952, Economists’ Papers Project, Duke University). Making this historical text play a role in current discussions on the measurability of utility turned out to be much more difficult than merely translating the text. The best way to do it, all three agreed, was for Sommers and Baumol to (physically) sit together and “revise” the translation, so that it would be “in accord” with “terminology, etc [..] used in current controversies” (Baumol’s letter to Sommers, 10 December, 1952). At the same time, making the text as accessible as possible to contemporary economists, while also sticking as close to a literal translation as possible, truly was the only motivation.

But as always, historiography showed no mercy. Particularly difficult was the word “emolumentum,” used by Bernoulli to denote the purchasing power of various monetary outcomes. Emolumentum was not to be confused with the marginally decreasing “moral value” individuals derive from increasing monetary outcomes, as argued by Bernoulli. Thus, Baumol reasoned, emolumentum is an objective measure of the value of different outcomes to the individual – just as von Neumann and Morgenstern had suggested – and inserted “utility” as its translation. Econometrica published the translation in 1954.

As a result, Bernoulli’s paper became reified as something in between a precursor of some hundred and fifty years to neoclassical economics’ reformulation of utility, and a critique of Baumol’s mistaken understanding of von Neumann and Morgenstern

Share your perspective