Talk:Utility/Archive 1

Archive 1

The "Utility in Game Theory" section is dodgey

I agree. The Utility in Game Theory section is ambiguous and non-standard.

I wrote a section on "Utility functions", which is a first attempt at replacing the material. However, I still need to write about Bernoulli utility function and the Expected Utility theorem, before we can "obselete" the existing section. When I get some time...

I like the approach/notation in Mas-Colell et al much better than Fudenberg/Tirole and Myerson. I haven't looked at Osbourne/Rubinstein.

I agree -- the entire Utility in Game Theory section should be scrapped. I'll take a stab sketching the expected utility theorem. Amcfreely 23:45, 8 August 2005 (UTC)

--Clausen 07:18, 9 Oct 2004 (UTC)

Ordinal Utility

Ahh, I'm bit confused here. One can observe someone's act or activities (buying, walking, sleeping, whatever). But what purpose does it serve to define such activities in context of choise? If someone paid $1 for a Coke, isn't that the only observable fact? Why bring Pepsi, Milk or Ice Cream? Can anyone expand on this? <FWBOarticle>

Also, that whole paragraph is full of spelling and grammatical errors, and is rather badly written as well.

Take a look at my rewrite. Amcfreely 04:58, 4 April 2006 (UTC)

Most modern economists are just applied mathematicians, and as such need to express everything in those terms, even if it doesn't apply. 12.47.123.121 13:59, 5 May 2006 (UTC)

Merge Util

Should the the unreferenced page Util be merged here and redirected? The page Util is unreferenced and the topic appears better covered here, but the term is not identified in the article Utility. Jeepday 14:24, 16 February 2007 (UTC)

Not only is the util article unreferenced, it discusses utility theory rather than the util itself (and does a poor job of it). There's certainly potential for an article about the unit, but turning that page into a redirect would be an improvement to it. -- Jonel | Speak 16:31, 16 February 2007 (UTC)]

Thanks for being BOLD Jonel. Signed Jeepday 13:03, 17 February 2007 (UTC)

vNM payoff is a major part of utility payoff, it should deserve to be a main article by its own right. It has much to do with game theory, rather than calculations. Moreover, it involves in people's choices and how they react, kind of like social science.

Merger (Expected utility)

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section.

The result was to merge Neumann-Morgenstern utility into Expected utility hypothesis, and to link this article from Utility#Expected utility. Moves, expansions, etc. can be considered after the merger. Further discussion should go to Talk:Expected utility hypothesis. --B. Wolterding 10:55, 14 July 2007 (UTC)

---

There are currently a number of articles discussing the concept of "Expected utility" and the Neumann-Morgenstern utility function:

• Utility#Expected utility
• Expected utility hypothesis
• Neumann-Morgenstern utility

Several merger proposals have been added over time, and I am trying to consolidate these and get to a consensus. In my point of view, these articles double up much content and should be merged into one central location. I think the best place would be Utility#Expected utility, but other suggestions are welcome. Please add your comments below. Proposed as part of the Notability wikiproject. --B. Wolterding 11:21, 3 July 2007 (UTC)

• Note: See also some discussion on Talk:Expected utility hypothesis about the merger. (If you wish to add new comments, please do so here.) --B. Wolterding 11:29, 3 July 2007 (UTC)

• I'd favour merging Expected utility hypothesis and Neumann-Morgenstern utility. As the standard economic theory of choice under uncertainty, this deserves an article of its own, with a link from the Utility. Note that there are already articles on various alternatives to and generalizations of EU such as Prospect theory.JQ 12:19, 3 July 2007 (UTC)

• • I agree with the comments that both articles are very weak. I'll see if I can put a bit of time into them.JQ 12:21, 3 July 2007 (UTC)

I agree strongly that Expected Utility should remain a separate article from Utility-- expected utility is fundamentally about uncertainty, while utility is fundamentally about preferences over bundles and outcomes. Jeremy Tobacman 19:27, 5 July 2007 (UTC)

• I'm fine with the proposal to merge the content of Neumann-Morgenstern utility into Expected utility hypothesis, and to link this from Utility#Expected utility. (Maybe also to remove some duplicate content from Utility#Expected utility.) Is that the consensus? If not, please protest; otherwise I will close the debate in a few days. --B. Wolterding 17:53, 7 July 2007 (UTC)

• I favor consolidating articles on EU and EU maximization, but having these be separate from the article on utility itself (which I think should discuss EU maximization briefly, with a link to the consolidated article). If the article on utility itself is both properly general (as it now is certainly not) and truly exhaustive, then it will also be exhausting to a great many readers. —SlamDiego_←T 00:23, 12 July 2007 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

disutility & inutility vs. utility

we need different pages about disutility & inutility, or we should explain them in this article. Jackzhp (talk) 19:37, 4 December 2008 (UTC)

utility of money

is the utility of money being called Bernoulli utility? Jackzhp (talk) 19:37, 4 December 2008 (UTC)

What happened?

Why has utility (in the moral sense) been made into an article about economics? I don't get it. Moral utility is clearly the most common use of the word (in the context of the philosophy of utilitarianism). WTF happened? 24.174.82.195 (talk) 04:37, 12 July 2009 (UTC)

Because in economics, Utility is all over the place. Philosophy of utilitarianism is less known, IMO. --Forich (talk) 23:30, 12 July 2009 (UTC)

Because economics is essentially politics (decision-making) without morals. Morality has no place in the so-called "objective science" of economics. As a result, the primacy of politics in decision-making has mostly been replaced by economics. As a result, few resources remain available for the non-human world, not to mention the poorest people in the world, who - lacking money - are considered to have no demand (for resources) by this field. —Preceding unsigned comment added by 72.150.179.59 (talk) 15:31, 14 August 2009 (UTC)

Suggestion to rewrite the "rational preferences" section in mathematical terms

Mathematics has this very nice subject called binary relation. I suggest some editor (I may attempt to do it) rewrites the section on rationality, and adapt it in those terms. For example, Fishburn (1994, p. 1401) says: "The formal development of utility (...) is based on the notion of a binary relation. A binary relation is (...) . Special types of binary relations are identified by conjunctions of properties as follows: (...) weak order: assymetric and negatively transitive (...) an ordered binary relation is isomorphically embeddable in the ordered real numbers (...). Weakly ordered preferences have been widely assumed in economic theory, where one presumes that X is the nonnegative orthant R+n of a finite-dimensional Euclidean space, preference increases along rays out from the origin, and the indifference classes or isoutility contours are smooth hypercurves concave." This is a very nice synthesis by Fishburn, and it avoids the ugly term that neoclassical economics has coined: "rational".--Forich (talk) 00:45, 20 January 2010 (UTC)

Historical Origins

I think the history of "Utility" should be worked out: in particular, I know that David Hume used "Utility" in the modern sense prior to 1776 (and hence prior to Adam Smith, Jeremy Bentham, and JS Mill.) But who was the first? Also, who was the first to formulate it as a function? Bentham? Pareto? Marshall? Samuelson? Rick lightburn (talk) 15:42, 30 December 2008 (UTC)

There is a well known article by Stigler on the history of utility.--Forich (talk) 21:11, 28 March 2010 (UTC)

Cardinal Utility

Utility functions of both sorts assign real numbers (utils) to members of a choice set. For example, suppose a cup of orange juice has utility of 120 utils, a cup of tea has a utility of 80 utils, and a cup of water has a utility of 40 utils. When speaking of cardinal utility, it could be concluded that the cup of orange juice is better than the cup of tea by exactly the same amount by which the cup of tea is better than the cup of water. One is not entitled to conclude, however, that the cup of tea is two thirds as good as the cup of juice, because this conclusion would depend not only on magnitudes of utility differences, but also on the "zero" of utility.

I don't understand why one cannot conclude that the cup of orange juice is 2/3 as good as the cup of juice... it seems to me that it must be exactly 2/3, given that we are talking about CARDINAL utility here. Can somebody explain it to me?--Forich (talk) 16:28, 15 February 2009 (UTC)

Different authors draw different distinctions when they use the terms “cardinal” and “ordinal”, and seem oblivious to the fact that they're not really speaking of quite the same things.

• Utility could be demonstrably ordinal, when no quantification could be fit to a rational preference system. (See Mc Culloch, James Huston; “The Austrian Theory of the Marginal Use and of Ordinal Marginal Utility”, Zeitschrift für Nationalökonomie 37 (1973).) Most economists are sadly oblivious to this possibility.
• Utility could be very weakly quantified such that a quantification could be fit, but any monotonic transformation would fit just as well. This is what most authors mean by “ordinal”, though some call it “cardinal”.
• Utility could be weakly quantified such that a quantification could be fit, but any affine transformation would fit just as well. Many authors still call this “ordinal”; many others call it “cardinal”.
• Utility could be more strongly quantified such that we can start to compare the utility of one person to that of another, much as we could compare the temperature of two physical bodies. Most authors who call that weakly quantified notion “ordinal” would call this notion “cardinal”.

Really, the article ought to distinguish these four notions (and perhaps others), explain the different uses of “cardinal” and “ordinal”, and then itself avoid using those ambiguous terms further. —SlamDiego_←T 09:34, 16 February 2009 (UTC)

I think you nailed it, SlamDiego. I've seen exactly those last three examples of utilities, for example "a quantification could be fit but any monotonic transformation would fit just as well": this is labeled as Ordinal in most textbooks. However, is an ambiguous definition, because the so called weakly-quantified-type-of-utilities share the same property, but somehow they are now Cardinal. I've never understood the explanation found in textbooks.

Now, onto the subject at hand: before attempting to use your proposal of taxonomy (it feels too much like original research, this is not the place to do that), I suggest we look for the actual mathematical definition of cardinality and ordinality. Only then can we pinpoint exactly where it is that the ambiguity arises. Your source (McCulloch 1977, p. 274) says somehing about mathematicians stating a precise definition of cardinality in 1949, it would be useful to start digging there. --Forich (talk) 23:09, 18 February 2009 (UTC)

What perhaps needs to be explicitly stated here is that, however any economist may define “cardinal” and “ordinal”, the cardinal is distinguished from the ordinal by a strengthening of assumptions; the cardinal has all of the positive properties of the ordinal.

I do fear that the issue of “original research” may bite hard here. The confusion that prevails amongst so many economists suggests that it would be difficult to find a clarifying exposition in the literature.

I don't propose that we use the terminology that I used above, which is both original and ad hoc. (Nor would I propose that we here use the terminology that I have actually used in presenting papers.) I simply propose that we use the categories above, after showing that they correspond to important treatments of the subject. I certainly don't think that we would overstep the bounds set by WP:OR if we cited and quoted different economists to make it made plain that they weren't defining these words in the same way as each other, and to show how they differed.

Meanwhile, though I am quite sure that mathematicians' notions should be included in the article (in part at least for the sake of those who come to this article with such notions), if using “cardinal” and “ordinal” only in ways that are familiar to mathematicians and these ways are unfamiliar to economists, then that itself would probably violate WP:OR. And I'm not sure that finding the mathematicians' definitions of these terms will actually pinpoint most of the ambiguity. McCulloch ends-up calling “essentially cardinal” everything that isn't what I've called “demonstrably ordinal” (which he calls simply “ordinal”). But it wouldn't be helpful to refuse to recognize the distinctions amongst these “essentially cardinal” concepts, even if every one of them fits a mathematicians' notion of “cardinality”. This article plainly needs to talk about the distinction between the concept where any monotonic transformation is allowed and that where only affine transformations are permitted; and it plainly needs to talk about the further requirements to justify interpersonal comparisons of utility.

If it cannot do these things withoug violating WP:OR, then we face the grotesque case that the article is intrinsically broken.—SlamDiego_←T 18:08, 19 February 2009 (UTC)

I have done major changes to the entry of cardinal utility. Take a look over there and discuss if necessary. --Forich (talk) 21:14, 28 March 2010 (UTC)

Philosophical Utility

This article is too much focused on 20th century concepts of utility in economics. Utility in philosophy before and during that time perio* Risk - some investments may have an alluring average/classical utility, yet are too risky for some investors to realistically enter. For example, a given business venture has a 1 in 10 chance of multiplying the principle investment by 20, and a 9 in 10 chance of a zero return. The average/classical utility of this investment is a 2-fold return on investment. This sort of investment is prudent only if there are enough such opportunities for the statistical average to be attained with reasonable certainty. And the investor must have deep enough pockets to finance such an extensive venture. An investment company may have the resources to do so, but individual investors may not.

• Classical Utility - for example an investment has a 50% chance of doubling the principle and a 50% chance of halving the principle. The classical utility of this investment is the average of 2 and 1/2; that is 1.25. A buyer may purchase a more expensive good because it will need replacement at a later date. Classical utility isn't inherently statistical; a given tool may have a add 50% to the value of a raw material and its classical utility is a measure of such.
• preference - a buyer may choose to purchase the item with the lower classical utility, simply because the buyer prefers that item. Marketing is heavily focused on this aspect of commerce. A buyer may purchase goods from the outlet with a freindly staff and higher priced, lower quality inventory.
• indifference - the classical utilities of two items are very similar, there is no significant preference for either item, the buyer is indifferent. Yet, one item is purchased the other is not. This may seem banal, but utility is very mathematical in its very nature, and such a function is necesary to create realistic models of economic activity.

To illistrate these four concepts/functions in an economic model:

• goods are purchased based on risk: if there is too high a probability of an unfavorable return, the purchase will not be made by 'reasonable' buyers. Regardless of the statistical, classical utility of the commodity.
• Assuming risk is not a deciding factor in choosing between two items, the choice will often be made by 'reasonable' buyers based on the higher average return on investment.
• Buyers do not always choose based on reason. Preference may outweigh risk and classical utility. For example, gamboling.
• In the event that no significant bias for either item, the choice will be made at random.

Who wrote the comment above? It makes little sense; in case we do a file archiving I suggest we eliminate it. --Forich (talk) 16:22, 18 June 2010 (UTC)

I searched in the history, and it seems like user: Dhh28, made a comment that somehow got messed up with another comment. --Forich (talk) 16:28, 18 June 2010 (UTC)

risk aversion vs. precautionary saving

2nd derivative vs. 3rd derivative of the utility function. Jackzhp (talk) 15:54, 1 December 2008 (UTC)

See Prudence#Prudence in economics and finance. Rinconsoleao (talk) 08:15, 6 October 2010 (UTC)

Barzilai's criticisms

This edit introduced the idea that Barzilai has criticized utility theory (and other stuff), in an article which can be read here. I'm very uncomfortable with this. It is (or soon will be?) published--I don't know if the material for the book is peer-reviewed or not. I find myself scratching my head in looking over the article, probably partially because it's one of those things that will take deep thought to understand, but I fear also because it's obscure. Mostly, I think this falls into "strong claims require strong evidence". There are plenty of criticisms of utility theory and expected utility theory. If Barzilai's are new and unique then they should be given time to garner response before being included here. To include them now is, I think, giving them undue weight. CRETOG8(t/c) 07:23, 6 October 2010 (UTC)

Barzilai's criticisms have been met with little response from peers, other than disregard: see http://www.columbia.edu/cu/psychology/Krantzlabweb/Papers/measurement_comments.pdf and http://www.creativedecisions.net/~rozann/Barzilai%20Rebuttal%20Whistler.pdf for instance. I haven't found any economists commenting on his work and, on Google Scholar, pretty much the only citations of his papers on the errors of utility theory are made by his own (newer) papers on the same topic. There are thousands upon thousands of papers written on utility theory more relevant than Barzilai's, so it makes little sense to include him in the article. I have removed the paragraph about Barzilai's work. 201.252.127.228 (talk) 21:21, 19 December 2010 (UTC)

Discussion and criticism should be cleaned up

It reads fairly poorly; the first paragraph lacks citations and seems kind of weasely to me; the second is mostly incomprehensible and certainly non-grammatical. —Preceding unsigned comment added by 131.107.0.101 (talk) 16:35, 5 January 2010 (UTC)

Some of the criticism to cardinal utilty in this entry is not accurate. It says: "it is impossible to measure the level of satisfaction "quantitatively" when someone consumes or purchases an apple". This is NOT a property of a cardinal utility function, only a property of "measurable utilities". Economists do not take measurable utility seriously nowadays, but cardinal utility remains valid, as in expected utility theory. Please read about this distinction of measurability and cardinal utility here. We need to fix this erroneus critique--Forich (talk) 18:23, 23 June 2012 (UTC)

Category theory

"Repeating in category language:" ... really? Does express it in terms of category theory have positive utility to anyone other than to the author? Angry bee (talk) 02:45, 17 June 2012 (UTC)

From the perspective of mathematics, it is really nonsense (compared to abstract nonsense). There is no morphism between categories. There are only morphisms between objects within a category and functors between category. I'll flag it with cn. 18.51.3.218 (talk) 03:08, 4 September 2013 (UTC)

Definition

The first sentence is a definition. It was inconsistent: it cannot be both relative and total customer satisfaction. I changed it according to investopedia.com, and also included a reference to this definition. Avi Harel (talk) 16:48, 20 November 2011 (UTC)

Why is the word 'customer' used at all? (Is that to suggest utility is non-applicable concept in an endowment economy?) I have removed it. Furthermost (talk) 04:23, 2 December 2011 (UTC)

"utility is a representation of preferences over some set of goods and services." Say what? "a representation of preferences over <something>"? Wuzzat? --Jerome Potts (talk) 18:53, 10 September 2013 (UTC)

substitution

How about creating a section on the substitution postulate: the trade-off between baskets of goods, giving up a commodity for some other one, the relation between substitubility and the slope of indifference curves, the implicit denial of the "priority of needs fallacy", and the relation between slope of indifference curves and "value"?--Forich (talk) 00:24, 16 February 2015 (UTC)

3D graph of a utility surface

Let's add a 3d graph of a utility function!--Forich (talk) 05:17, 17 March 2015 (UTC)

Dr. Bertoletti's comment on this article

Dr. Bertoletti has reviewed this Wikipedia page, and provided us with the following comments to improve its quality:

1) "Gorman polar form" is not a utility function 2) the "continuity/archimedean" property of preferences should not be labelled "convexity"

We hope Wikipedians on this talk page can take advantage of these comments and improve the quality of the article accordingly.

Dr. Bertoletti has published scholarly research which seems to be relevant to this Wikipedia article:

• Reference : Paolo Bertoletti & Federico Etro, 2013. "Monopolistic Competition when Income Matters," DEM Working Papers Series 055, University of Pavia, Department of Economics and Management.

ExpertIdeasBot (talk) 18:55, 27 June 2016 (UTC)

Not devoid of original meaning

The last sentence in the first paragraph reads: "It is devoid of its original interpretation as a measurement of the pleasure or satisfaction obtained by the consumer from that choice." This is not true, whilst utility has become a more abstract concept within economics, it is still clear that the function originally derives from pleasure/satisfaction of the consumer. I will replace this for now. Andre908436 (talk) 14:06, 29 April 2020 (UTC)