Friday, March 7, 2014

The “Law” of Diminishing Marginal Utility

The law of diminishing subjective marginal utility states that, as a person consumes an additional unit of the same good (or a homogenous good), then the satisfaction or utility derived from the consumption of that good diminishes and continues to diminish with each additional good.

So one’s “marginal utility” is the additional satisfaction one obtains from consuming one additional unit of a good. At some point, an additional unit would yield zero utility and you would not purchase such an additional unit.

Of course, there are problems with how you define the “same good” or “homogenous good”: does each additional beer of the same brand and type when it is served in a new glass constitute the same good? Presumably it does, although technically your “good” – the beer – is not exactly the same. What if the glasses of beer taken from the same keg diminish slightly in quality as your reach the dregs? Are you consuming the same good?

It would seem that there needs to be a very high degree of similarity between the different units of the same good (such as its brand, type, style, quality, etc.), but not necessarily perfect or exact homogeneity. For think of two articles of clothing: two shirts of the same type, brand, style and colour, and to the human eye indistinguishable.

However, there will be slight differences between the stitching and perhaps loose threads of the “same” two shirts that are an inevitable outcome of the manufacturing process. Are they still the same good? I think most people would say they are, and that this is perfectly reasonable: you do not need absolute and perfect homogeneity between two units for them to be the same good.

Another important assumption is that one must assume constant preferences in at least the short run for the law of diminishing marginal utility to be valid.

In mainstream neoclassical economics, it seems that the “law” is sometimes interpreted as a general principle, not as some universal law of nature like the tautologous law of demand.

But that does not stop Austrian apriorists telling us (literally) that the principle of diminishing marginal utility is a universal law and irrefutably true, a view which is wholly unconvincing and derived from the absurd belief in Kant’s synthetic a priori knowledge, and the use of tautologous analytic a priori statements which are “irrefutable” only because they assert nothing necessarily true of the real world.

Therefore we are justified in rewriting the “law” of diminishing marginal utility as a synthetic a posteriori generalised statement in the following form:
(1) it is generally (but not universally) the case that as a person consumes each additional unit of the same good, then the utility of each additional unit falls.
Understood in this way, there is clearly a lot of empirical evidence to support this idea as true. For example, most people’s demand for washing machines, toasters, or fridges is pretty stable. Perhaps some people (with big families and enough room in the laundry) might derive significant utility from a second washing machine, but by the time we get to the 3rd or 4th unit, such washing machines are pretty much useless.

However, are there exceptions?

Any person, even after a few minutes of reflection, on the basis of their own personal experience or empirical evidence about other people can see significant exceptions:
(1) coin or card operated video or arcade games
In video games in which one has to purchase each new game by coins or card credit, it is easy to find people who become obsessively addicted to a single game, and where it is clear that each time the game is played constitutes the same good. At the same time, a person’s utility from each additional unit may well be constant and even rise as that person gets better at the game and wants to play more to get to the end level. Even if and when one gets to the end level, one may still derive utility from further units of the game because one enjoys playing it. This repeated playing can go on for weeks, months or (in extreme cases) years.

When such a person eventually stops playing the game, it is not because their marginal utility has fallen to zero, but because their preferences have changed: perhaps they are now interested in some new video game or hobby, or for teenagers they now might have their time taken up with friends and social life, and so on.

(2) Addictive drugs and substances
Here we have a straightforward counterexample to the principle of diminishing marginal utility: a person addicted to the same drug will derive constant and possibly even rising marginal utility from each additional unit consumed.

(3) Addictions to fashionable articles of consumption
Some people have shoe “addictions,” or repeated buying of the same type of fashionable article of consumption. The purchase of additional goods can go on for a long time over many years, with varying degrees of utility derived from each additional unit and no apparent diminishing marginal utility stopping the addiction.

Here, of course, the marginalist can always respond by denying that things like new pairs of shoes or jeans or handbags are the same good.

For isn’t a different pair of shoes of a different brand or style a different good, so that even if you kept deriving constant or varying (upwards and downwards) utility from each new purchase of a new pair of shoes and, say, collected at least 3,000 of them (a real life example right here), then the shoes are technically different goods, not the same good?

Of course, this might be a reasonable objection, but now we are back to the question: how homogenous do units have to be in order to be classified as the same good? It is obvious that they do not need to be perfectly homogenous, and so goods with a high degree of similarity seem to come close to violating the principle of diminishing marginal utility.

(4) Collectors of rare objects, used goods, or works of art
One need only think of stamp collecting, or collectors of antique cars or sports cars, or rare art works by the same artist.

Again one might encounter the objection that these things constitute different goods, but that objection seems less convincing here: for example, some people might collect the same type and brand of used goods such as a particular type of motor cycle made by the same company. Here it appears that the same good really is being collected and the utility from each unit need not diminish.

(5) Money
It is possible that some people may actually have a greed for money so strong that they simply wish to accumulate more and more so that the principle of diminishing marginal utility does not apply.

I only say “some people” because there is strong evidence that for most people, once you earn about $75,000, the marginal utility of an additional unit of money does fall, and people report that their own subjective happiness (or “emotional well-being”) does not increase much or levels off at that point.

There would appear to be some people, however, for whom an ever increasing accumulation of money does appear like a pathological addiction, given the way increasing money wealth can confer power, special status, social acceptance, and so on.
But further problems arise from utility theory: say, the belief that when people (supposedly) act to maximise their utility, then this will result in the best or ideal society. That falls apart once it is seen that there are many activities where if people are allowed to maximise their utility in addictive, self-destructive or short-sighted ways, then deleterious consequences follow for such people and for others affected by their behaviour.

Joan Robinson (1964: 49) gave two obvious examples of this: (1) drug addicts and (2) children who would not want to go to school if left to “maximise their utility.”

The social consequences of drug addiction are easy to document, and if we lived in a society where many people choose to not be educated as children, then there would be severe problems in finding skilled labour for production to even continue (and perhaps even in finding people properly literate!).

Further Reading
Philip Pilkington, “Joan Robinson’s Critique of Marginal Utility Theory,” Fixing the Economists, February 17, 2014.

Thorsten Polleit, “What can the Law of Diminishing Marginal Utility teach us?,” Mises Daily, February 11, 2011.

Robinson, Joan. 1964. Economic Philosophy. Penguin, Harmondsworth.


  1. Is there a timeframe imposed on this 'law' ?
    This holds generally true for continuous consumption, but not for intermittent consumption.
    A second beer today may not be as satisfying as the first, but a second beer tomorrow, after digging the garden, would be more satisfying.

    1. Yes, the relevant time frame for the law is yet another serious problem in its definition.

  2. sure you love the Mises references...
    have you read his "epistemological problems..."?

    I feel like his writings on action, choice, preference, etc. make it clear enough that a pedantic concern with terminology (synthetic a priori) doesn't actually make more sense than his fairly clear layout of his thoughts. Guido Hulsmann's foreword does a very good job of making the basics clear as well.

    I check back in here from time to time because I appreciate any effort to seek truth and argue against assumptions. But I don't think your arguments have been more accurate or on point than the Misesians(specifically Mises, Rothbard, Hoppe, and Hulsmann). I certainly don't consider all Austrians in the same camp on value theory, marginalism, or epistemology. So to me, it's very important to keep the distinctions in mind.

    1. If one can extract any substantive point from the rambling comment above, it appears to be that some Austrians are not so extreme as Mises, Rothbard? How relates to the post I wrote above remains a mystery.

  3. Responded here...