Home > Uncategorized > Utility maximization — explaining everything and nothing

Utility maximization — explaining everything and nothing

from Lars Syll

Despite the rise of behavioral economics, many economists still believe that utility maximization is a good explanation of human behavior. Although evidence from experimental economics and elsewhere has rolled back the assumption that human agents are entirely self-interested, and shown that altruism and cooperation are important, a prominent response has been to modify individual preference functions so that they are “other-regarding”. But even with these modified preference functions, individuals are still maximizing their own utility.

5197ebbdd9c758d5c73657c270f97340Defenders of utility maximization rightly reject critical claims that specific behavioral outcomes undermine this assumption. They do not. But this is a sign of weakness rather than strength. The problem is that utility maximization is unfalsifiable as an explanation of behavior. As I show more fully in my 2013 book entitled From Pleasure Machines to Moral Communities, utility maximization can fit any real-world evidence, including behavior that appears to suggest preference inconsistency.

But note that utility maximization is not a tautology. Tautologies are true by assumption or definition. Utility maximization is not a tautology because it is potentially false. But empirically it is unfalsifiable.

Where does that leave us? Utility maximization can be useful as a heuristic modelling device. But strictly it does not explain any behavior. It does not identify specific causes. It cannot explain any particular behavior because it is consistent with any observable behavior. Its apparent universal power signals weakness, not strength.

Geoff Hodgson

Interesting post from one of yours truly’s favourite economists.  

On the question of tautology, I think it is only fair to say that the way axioms and theorems are formulated in mainstream (neoclassical) economics, they are often made tautological and informationally totally empty. 

Modern (expected) utility theory is a good example of this. Leaving the specification of preferences without almost any restrictions whatsoever, every imaginable evidence is safely made compatible with the all-embracing ‘theory’ — and a theory without informational content never risks being empirically tested and found falsified. Used in mainstream economics ‘thought experimental’ activities, it may of course be very ‘handy’, but totally void of any empirical value.

Utility theory has like so many other economic theories morphed into an empty theory of everything. And a theory of everything explains nothing — just as Gary Becker’s ‘economics of everything’ it only makes nonsense out of economic science.

The ever-growing literature on human capital has long recognized that the scope of the theory extends well beyond the traditional analysis of schooling and on-the-job training … Yet economists have ignored the analysis of an important class of activities which can and should be brought within the purview of the theory. A prime example of this class is brushing teeth.

BeckerGaryCartoon2009_07_10The conventional analysis of toothbrushing has centered around two basic models. The “bad taste in mouth” model is based on the notion that each person has a “taste for brushing,” and the fact that brushing frequencies differ is “explained” by differences in tastes. Since any pattern of human behavior can be rationalized by such implicit theorizing, this model is devoid of empirically testable predictions, and hence uninteresting.

The “mother told me so” theory is based on differences in cultural upbringing. Here it is argued, for example, that thrice-a-day brushers brush three times daily because their mothers forced them to do so as children. Of course, this is hardly a complete explanation. Like most psychological theories, it leaves open the question of why mothers should want their children to brush after every meal …

In a survey of professors in a leading Eastern university it was found that assistant professors brushed 2.14 times daily on average, while associate professors brushed only 1.89 times and full professors only 1.47 times daily. The author, a sociologist, mistakenly attributed this finding to the fact that the higher-ranking professors were older and that hygiene standards in America had advanced steadily over time. To a human capital theorist, of course, this pattern is exactly what would be expected from the higher wages received in the higher professorial ranks, and from the fact that younger professors, looking for promotions, cannot afford to have bad breath.

Alan Blinder

  1. April 5, 2016 at 3:41 am

    Interesting, but no Kewpie Doll. Utility maximization has the same issues as any other form of optimization. First, when and how can it be known when utility is maximized. Could it be more maximized, or less, and sill have the same results. Second, what circumstances impede or even negate utility achievement of some utility and to what level? Third, the optimized world is never the world we see when we’re not being told to optimize. So the world of maximizing or partly maximized utility is not researchable in terms of the world that is not and is not planned to be optimized. Like other optimizations, utility maximization is a fiction used to fill in the blanks while we search for what’s really going on. A “placeholder,” nothing more. If that placeholder helps us organize or see more clearly how the experiences called “economic” are organized then it’s useful. Otherwise, it’s a waste of not just just research time but also resources and energy.

  2. merijnknibbe
    April 5, 2016 at 7:05 pm

    One of the features of the utility model is that it ´explains´ the downward sloping demand curve, a cornerstone of economics. Which means that neoclassical demand theory seems a pretty coherent building with sound foundations.


    It does not explain the downward sloping demand curve. It is only consistent with this curve. And in 1962 Gary Becker showed, in an article called ´Irrational behavior and economic theory´, that many models can ´explain´ a downward sloping demand curve when money is limited, including the throw of a dice. Ockhams razorblade requires us to use the simplest model… http://mcadams.posc.mu.edu/econ/Becker,%2520Irrational%2520Behavior.pdf

    Becker himself seems not to have grasped the implications of his article, which shredded neoclassical demand theory. Accepting that demand curves very often slope downwards does not mean that one has to accept utility theory.

    On top of this: in one of the most important markets, houses, higher prices often lead to more demand and lower prices to less demand (for very rational reasons), as shown in this article by Moen, Nenov and Sniekers: http://www.voxeu.org/article/house-prices-buying-or-selling-first-matters.

  3. April 6, 2016 at 1:03 am

    You cannot falsify what is undefined. Ever since Alfred Marshall, ‘utility’ is a benefit, or a satisfaction, or a pleasure, et cetera, infinitely plastic in its meaning. And, when one does not permit preferences ever to show negative values, then the notion is that ‘utility’ is always positive substitutes for real differences in tastes.

    “Utility’ is, however, easily and empirically falsifiable if it is ‘satisfaction/pleasure’. If one allows for all of the mainstream assumptions about preferences, it is easy to demonstrate that existential necessities come before satisfaction or pleasure with what is affordably ‘consumed’.

    If anyone is interested I will post why this is so.

  4. April 6, 2016 at 9:40 am

    I would agree that utility maximization explains everything, and nothing too.

    see ‘backward bending (or sloping) utility curve’.

    one has exact same issue in physics, everything (or almost everything) can be written as a maximum entropy solution using same calculus of variations as utility maximization. the solultion is always p~exp(-beta H) where beta is inverse temperature and H is the Hamiltonian or energy.

    If H is an ideal gas, and seperable or additive, then you get a unique solution (like general equilibrium).

    If its not, even in the most simplest case —-like the ising model–you have multiple optima. If you deal with spin glasses—generalized ising models—one quickly ends up with higher order polynomials. (you can get same result with ising model with a little more trouble. Onsager got a noble prize in physics for finding a solution to the 2 dimensional ising model—and thats last solution you will get—you won’t find anything beyond 2 dimensjons any more than you will with a random walk in dim>2).

    Amartya Sen and Herbert Simon sort of pointed this out long ago —‘menu dependence’ and bounded rationality. Its connected with Marshall-Arthur-Davidson ‘path dependence’ and nonergodicity. I stilll use it until something better comes along, and that is what i am working on—most likely a dead end.

    A very recent article in something i can’t recall maybe AMM (american mathematical monthly) by H T Davis ‘mathematical adventures in social sciences’ 1938 ) discussed bentham’s utilitarian calculus (as well as Slutsky’s theorem—also called yule-slutsky effect—spurious correlations.) —he laid it down in a very simple way. V I Arnold

    pauli.uni-muenster.de/~munsteg/arnold.html seems to share my opinion tho he got it befoe me. arnold is famous for newer version of ergodic theorem—KAM theorem. dobryshunin is also in there .

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.