Blogs > Liberty and Power > Twenty Questions on Mises, Part I

Nov 23, 2006 7:21 am

Twenty Questions on Mises, Part I

In economics I am almost completely an autodidact. I also know that autodidacts can sometimes be a crankish, blinkered, downright ornery lot.

It is therefore in a spirit of humility that I am reading Ludwig von Mises' The Theory of Money and Credit. Boldly -- or modestly, I can't decide which -- I plan to ask a series of questions as I go; perhaps some economists out there can enlighten me by answering them. I foresee this as a major blogging project, something that will go on for several weeks as I wend my way through Mises' dense, abstract, idea-rich prose.

I expect that most of the posts in this projected series will be short, simple, and hopefully helpful to other nonspecialists. To the best of my ability, I will include a synopsis of what Mises meant so that even those who have not read the text will be able to understand my questions and hopefully discuss them with me. (I understand, of course, that many of the questions will be the products of my own misunderstanding. I ask that you be kind to me when they are.)

Obviously, I intend the comments to be the most interesting part of the process. All posts in the series will be crossposted at Positive Liberty, but I will repost the most interesting and/or helpful comments at Liberty & Power as well. By commenting, you consent to letting me borrow your words in this way; I will keep attributions and links to the originals.

Archetypically, I'm aiming for twenty questions in twenty posts. The first is below the fold.

In the foreword, Murray Rothbard writes,

[N]eglect of the Cuhel-Mises theory of ordinal marginal utility allowed Western economists, led by Hicks and Allen in the mid-1930s, to throw out marginal utility altogether in favor of the fallacious"indifference curve" approach, now familiar in micro textbooks.

Rothbard, of course, did not use links to Wikipedia. But you get the idea. Here's my question: Why is the indifference curve"fallacious" in light of ordinal marginal utility?

Here is what I presume is a good concise definition of marginal utility, by way of Wikipedia. It seems consonant with Mises' own views, expressed later in TM&C.

In economics, marginal utility is the least urgent use of an object in satisfying a want - in other words, the use that is in the"margin." Marginal utility is subjective, because it depends on each person's wants and tastes. The same object may have different marginal utilities for different people.

For example, let us assume that a person has three wants and the satisfaction of each want each requires one gallon of water, so that satisfying all his wants requires three gallons of water. In descending priority, the most urgent want is to satisfy his thirst, the second most want is to give water to his dog, and his least urgent want is to water his roses. The least urgent use (the marginal utility) of one gallon of water when he has two is therefore to give water to his dog. If he has three gallons of water, then the least urgent use would be to water his roses. The"marginal utility" of any given gallon of water depends on how much water he has.

The concept of marginal utility is said to explain the"diamond-water paradox" most usually associated with Adam Smith. Smith asked, if water is more useful than diamonds, then why does water have a lower market price than diamonds? Marginalists answer that it is not the total usefulness that matters, but the usefulness of each unit of water. It is true that the total utility of water to people is tremendous, because they need to survive. However, since water is in such large supply in the world, the marginal utility of water is low. In other words, each additional unit of water that becomes available can be applied to less urgent uses as more urgent uses for water are satisfied. Therefore, any particular unit of water becomes worth less to people as the supply of water increases.

Unless I miss some subtlety (the curse of the autodidact is always to miss some subtlety), this definition seems adequate. Marginal utility is said to be"ordinal" because, in the context of a given good, it concerns the satisfaction of our most pressing want, followed by our second most pressing want, followed by our third, and so forth. Yet it is not absolute, and it cannot, as Mises argued, be reduced to any abstract unit of satisfaction.

The reasons for this are twofold: We cannot speak of units of satisfaction because satisfaction of want, insofar as it inheres in obtaining a good, comes always in the context of marginal utility. Satisfaction is a moving target, and quantified units of satisfaction have no fixed point to base themselves upon. It's all marginal, and therefore every exchange takes place at a different level of relative satisfactions.

We also cannot speak of units of satisfaction because while our wants are ordinal, there is no psychological method (yet?) of quantifying the difference in satisfaction that one takes between giving water to a dog and giving water to roses. While we may prefer one to the other, any quantity attached to this preference must be the product of our imagination (an ordinal preference, though, is not imaginary, since we can generally order our wants through introspection, and these, by definition, are actually what we want).

Now, here's an explanation of the indifference curve, which I also recall from my microeconomics textbook:

An indifference curve is a graph showing combinations of goods for which a consumer is indifferent, that is, it has no preference for one combination versus another, as they render the same level of satisfaction (or the same amount of utility) for the consumer. Curves are a device to represent preferences and are used in choice theory...

There follows a good deal of mathematical discussion which does not transcribe well into WordPress.

To repeat the question: Why is the indifference curve fallacious? I cannot see, simply by introspection, why such a thing should be fallacious in itself. On the contrary, it seems obvious to me that it is real -- and obvious even to the point of triviality. Nor can I see how the indifference curve stands in conflict with the notion of ordinal marginal utility: It might well be that I do not care about the change in quantity along a curve. One Coke? One Pepsi? Personally, I do not care at all, I can't tell them apart, even on a good day -- so an indifference curve certainly exists here for me. What am I not getting?

One question down, probably about nineteen to go. Stay tuned.

comments powered by Disqus

More Comments:

Sudha Shenoy - 11/28/2006

Hands use tools & machinery: not just fingers, sinews, muscles alone. In the DCs, it is the huge quantities & varieties of tools & _machinery_ that, in conjunction with hands, yield, & continue to yield, such a vast cornucopia of goods. Highly productive machinery -- capital -- is so ubiquitous in DCs that people (esp. engineers) treat them as natural growths. Compare the kinds of 'equipment' produced & used in LDCs, by & for ordinary farmers, for example; or the kinds of tools used by the vast majority of craft workers there.

Andrew D. Todd - 11/28/2006

Here is something which might interest you:

I should think it is probably impossible to grow up in the United States without being taught to work with one's hands. I don't know exactly what your background is like. However, you seem to be starting from very different tactile presuppositions than I start from.

Sudha Shenoy - 11/28/2006

Machines aren't rabbits. Machines have to be produced. Reproduction of rabbits & other such creatures is a biological process. Production of machines is not. Production requires real resources which are _not_ reproduced biologically nor do they just drop from the sky. The capital resources needed to produce machines are 'scarce'. That's why machines cannot be simply reproduced ad infinitum: they have 'costs'.

In a DC, capital has been accumulated over several generations. Vast quantities of goods are produced & poured out daily in shops, factories, & the like. Therefore it appears that life is close to being a Land of Cockaigne, where roast fowls fly through the air, crying 'come eat me! Especially to an engineer.

Andrew D. Todd - 11/28/2006

Your understanding of reverse engineering is rather different from mine. I am an engineer, so I rather think my understanding is correct. Reverse engineering means that you take a device apart to find out how it works, and write documents explaining how it works. You can then design similar machines from the ground up. To someone skilled in engineering, a piece of equipment dismantled on the table is like a conversation, or a book. Incidentally, the Japanese have for a long time been the best mass-production machine designers in the world, so there really isn't anyone for them to copy from. It reminds me of the line from _The Magnificent Seven_: "If he is the best, then with whom does he compete? With himself, of course!"

The American automobile industry puts an enormous effort into engineering style changes, the whole point being that if you buy a new car, people will be able to recognize it as a new car, even though real engineering changes in automobiles are glacial in pace. Judged by the key performance criteria of how fast it gets you there, the automobile has not improved at all in the last fifty years. So the only way to sell more automobiles is to make them into fashion items. If you buy a new car, people will therefore be able to recognize you as a big spender, but the danger is that they may merely dismiss you as a parvenu. American automobile engineers, in a sort of angry self-disgust, call this kind of thing "perfuming the pig." They have developed a highly evolved inferiority complex, vis a vis the Japanese, the way the English used to have, vis a vis the Americans. Obviously, if, like the Chinese, you want transportation pure and simple, you don't go in for all that kickshaw. Basic transportation is at substantially the level of the work which American engineering students do, as an incident of their training.

What makes a developed country developed has much less to do with physical capital than with mental capital or social capital. The worst of the physical damage of the Second World War in Europe was substantially repaired within several years, because most of the mental capital was still substantially intact, and still more of the social capital.

For manufacturing machinery, your starting point is general-purpose machine tools-- things like metal-cutting lathes, planers, drill presses, etc. You can use your general-purpose machine tools to make additional general purpose machine tools-- in effect, they can be made to breed like rabbits. When you have enough general-purpose machine tools, you can use some of them to make special-purpose machine tools, which do a more efficient job of manufacturing certain devices-- for example, the transfer machines which are used to make automobile engines. I should think you might be able to get the machines to replicate each other at a rate of 1000% annually, maybe more. The supply of machines is not a limiting factor. The immediate limiting factor is likely to be the supply of people who know how to use the machines. Human capital, in short.

Taken as a whole, China seems to have little or no difficult producing just about everything up to the limits of the formal and informal trade restrictions of the various developed countries, and further, in producing nearly all the various "pre-requisite" goods and services which do not involve being land-rich (eg. the mines of Australia, the grainlands of North America). Where China may be running into difficulties is in producing goods cheaply enough that ordinary Chinese can afford to buy them.

At the same time, there are ultimate limits to the export trade. It is, I believe, an economic dogma that wants are unlimited. This is somewhat problematic. The reality is that consuming is more or less laborious. This has been noted in the computer, electronics, and software industries. There comes a point where consumers resist upgrades and new products for fear of having to learn how the new system works. Just because clothing is very cheap does not mean that western consumers will go in for unlimited dressing-up. Similarly, the obvious upgrade from an automobile is an airplane, but airplanes are very hard to learn to fly. There simply isn't much middle ground between a nonpilot and a fighter pilot. Even if the Chinese could make a private airplane for almost nothing, flying would still be an exclusive club.

China has a problem with water shortages, of course. But the Israelis have done some very interesting things with desalination, drip irrigation, etc.

Jason Kuznicki - 11/27/2006

I must say I found the Hoppe piece unenlightening. Let me summarize as follows:

Robert Nozick argued that the concept of a homogenous commodity implies indifference, since we are indifferent about the exchange of unit A or unit B of the same commodity, provided that the units are of identical quantity. Thus ordinal marginal utility cannot possibly show indifference to be fallacious and remain a consistent account of economic behavior.

Block, then, argued that when we choose unit B as the unit to exchange, we have created heterogeneity in a previously homogenous domain.

Hoppe charges Block of using an ad hoc argument. Yet, almost incredibly, Hoppe turns around and uses one himself -- of the very same type! He writes,

To say that I am indifferent to strawberry and vanilla ice cream is to say, for instance, that a correct description of my action should simply speak of ice cream or something cold and creamy. Getting a strawberry ice cream in exchange for my dollar is then simply not a part of the description of my choice. Instead, my choice demonstrates that I prefer an ice cream or something cold and creamy to a dollar.

I submit that this is every bit as ad hoc as the argument Hoppe set out to refute, and that it is ad hoc in precisely the same direction.

Not, though, that this is necessarily a problem. It just means, after all, that we must resurrect indifference as a useful economic concept, even in a theory based on ordinal marginal utility: Indifference may be defined as that state which arises whenever some people would split a commodity into heterogenous types, while others would lump it into a homogenous single type. The latter group experiences indifference, and, combined with the perhaps discontinuous indifference curves mentioned elsewhere in this discussion, it does not seem to be at odds with Mises or ordinal marginal utility at all.

Sudha Shenoy - 11/27/2006

1. 'Reverse engineering' can provide one or two crude copies. To equip entire industries with machines requires the availability of _real_ resources.

If the Chinese have so much capital around that _they_ can _produce _all_ the industrial machinery they use_, then they're _already developed_. And if they can _produce_ such huge quantities of industrial machinery, why aren't they producing enough luxury grains to feed themselves? Why are they still producing so much animal fodder?

2. The Chinese now produce a whole _range_ of industrial consumer goods -- which they sell _to the entire world_, not just the US. Eg, some 80% of Chinese clothing exports go to the world overall -- principally East & SE Asia; only 20% are sent to the US.

3. Incidentally, the Chinese are hopeless at 'reverse engineering'. The real experts are the South Koreans & the Japanese.

4. The book by the Brazilian author came out in 1933. I understand that Brazilian studies have moved much, much further in the last 73 years. In any case, this is for someone knowledgeable about Brazil to deal with.

Andrew D. Todd - 11/27/2006

Well, it is a commonplace in industrial circles that the Chinese are notorious for buying one of a machine, reverse-engineering it, and cranking out copies. If they use the machines internally rather than exporting them, they don't have to worry about what view a western legal system might take. Things like automobiles, sewing machines, etc. are basically nineteenth and twentieth century technology. There really aren't a whole lot of secrets about them, and if you have a sufficiently large supply of wretched peasants, you can easily produce them in the traditional way. Now, producing them in a developed country, without wretched peasants, is considerably more difficult, just as it was somewhat difficult to find ways of growing cotton without stoop labor. There are a few technologically advanced products, where delivery of the product does not constitute disclosure of the information necessary to replicated it. A case in point would be the most advanced microcircuits, but these only involve tiny numbers of workers in any case.

For what it is worth, I have heard that some of the big discount store chains are setting up "consolidation warehouses" in China, where they can collect goods, sort and repack them to the last detail, and minimize the amount of work which has to be done in the United States. Global air freight is only about a dollar a pound, so it might well be profitable to make up mail orders in China, particularly for comparatively expensive lightweight goods such as clothing, and ship them by air freight to regional airports within the United States.

Of course, the ultimate issue is "telepresence," remote control over global distances. I already get telephone sales call from people in India. As the bandwidth widens, the effective distance decreases. Of course, these Indians fail in their efforts to pass themselves off as Americans-- their accent gives them away.

Here's a question for you, does the concept of "enervation," which Gilberto Freyre speaks of in _The Masters and the Slaves: A Study in the Development of Brazilian Civilization:_ (1964), apply to a situation in which the developed countries are entirely electronically plugged in to that portion of the third world which is on the verge of starvation?;q=%2Bchina+%2B%22sewing+machines%22&btnG=Google+Search

Sudha Shenoy - 11/26/2006

1. Forty years ago, exactly the same things were being said about the Japanese & about Hong Kong: 'they live on the smell of an oil rag, we'll all starve to death if Japanese & Hong kong imports are allowed in', etc., etc. Now it's the turn of the Chinese.

2. Why do the Chinese (& other Asians) eat millet, which is only fed to pigs in the US? In other words, the Chinese have no bread, why don't they eat cake?

DCs _produce_ far, far more than LDCs do, which is why in the DCs everyone can eat a luxury grain like wheat, & use the 'inferior' grains as animal fodder.

3. _Some_ Chinese workers may live in dormitories; most live at home. Factory wages are _far_ higher than they could earn elsewhere. (As they have very little bread, they cannot eat cake, however nutritious it might (or might not) be.)

When goods are produced at lower cost than before, this means resources are available to produce _other_ things. Chinese-produced goods are cheaper. Thus buyers have money to spend on _other_ things. To produce these, capital & labour have to be freed -- from the firms that are _now_ too high-cost to survive.

(This is what happened in Australia for example. The heavily-subsidised steel industry closed down. Later surveys showed that the workers had moved into a variety of other jobs. The heavily-subsidised clothing industry closed down. Oz exports of_ processed_ foods expanded -- to Asia.)

Also, Chinese clothing factories (_eg_) have to import industrial sewing machines, button-making, weaving, & spinning machinery, dyeing & finishing equipment, dyes & chemicals, etc, etc, _first_ -- in order to produce any clothing at all. Such imports have to _continue_ for the Chinese to continue production.

All this came from the DCs, & continues to come from there. Not to speak of the transport investments, etc, needed to get the clothing (& other goods) to final buyers. All this is _not_ obvious, until inquired into.

Sheldon Richman - 11/26/2006

"I dare say I am using the term positivism rather loosely, not just to refer to Wittgenstein as such...."

Wittgenstein was no positivist. The Vienna Circle might have claimed him at first, they eventually learned that was an error.

Andrew D. Todd - 11/26/2006

To: Steven Horwitz

Well, for the record, I don't pretend to know what the ultimate implications of free trade are-- and I don't think anyone else knows either. Certain types of queries are "computationally infeasible," in the sense that you can demonstrate that the permutations of, say, a thousand objects are well in excess of ten to the hundredth power. When you take account of all the possible nth-order interactions between different groups of people, and recognize that all of these are only imperfectly measurable, and the cumulative error makes it impossible to draw a net balance. Free trade is good for some people, bad for others, and how do you strike a balance?

I dare say I am using the term positivism rather loosely, not just to refer to Wittgenstein as such, but to the whole kind of body of thinking, reaching back to the eighteenth century, which says that you can reduce anything as complicated as human society to numbers or class-archetypes. Obviously, adherence to this idea is relative rather than absolute. If you read Henry Mayhew, for example, you will find a circumstantial richness which is of an entirely different order from the specimens of Von Mises which I have read. I notice that Roderick Long has just put up a review of Barbara Ehrenreich’s _Nickel and Dimed_. Mayhew's _London Labor and London Poor_ is the ur-text of that type, the standard against which such works are compared.

To: Sudha Shenoy:

Re: Tariffs:

You are assuming that a firm is in one country or another country. At this stage, the reality is that of a firm which directly or indirectly employs workers in various different countries, and moves bits and pieces of work in progress around between these different countries. It is often possible to design the product itself so that bits and pieces do not have to travel to places where they might get hit with comparatively steep import duties. The firm reasons approximately as follows: "We can either do this little manufacturing operation ourselves in the United States, or we can outsource it to a firm in China. If we do it in the United States, we would have to invent a robot to do it. If our engineers should attempt to develop such a robot and fail, we would still be out of pocket for their expenses. If we outsource it to a firm in China, they will worry about the details, and they will give us a certificate signed by the Chinese government to the effect that there is no child labor or anything like that. The United States Department of Commerce will accept this certificate. It is not our job to investigate the Chinese government for corruption!"

If it turns out that the goods in question are being produced by half-starved six-year-olds, then the American firm is "Shocked, Shocked, I tell you!" So it goes, according to business school logic.

The prevailing wage in China's export industries is only about twenty-five cents per hour. Some of that reflects the extremely low standard of living of the workers, who usually live in factory dormitories, under conditions little better than American farm animals. I gather they tend to live on things like millet porridge. In the United States, millet (sorghum) is usually fed to hogs. Some of the low wage reflects the even lower standard of living of the Chinese peasants who subsidize the factories. Therefore, an American worker would have to be forty or fifty times as productive as a Chinese worker to keep pace. Designing a machine good enough to give that kind of edge is not always easy.

Given the overwhelming character of the difference between the United States and China, if you are going to make free trade a point of doctrine, you do have to address the cheap labor issue.

Sudha Shenoy - 11/24/2006

1. This is precisely what Mises said: the real world is historical, & to 'understand' any historical context, we must use (whether we realise it or not) 'ideal types'. The latter incorporate such historical characteristics as seem necessary to that specific historical problem. SV 'ideal types' in Human Action (available online.)

2. The suggestion that Mises was a positivist deserves at least a book to support it -- & it will be a sensation when it comes out, since everyone knows Mises was a redoubtable _anti_-positivist. Try reading Epistemological problems of Economics, also The Ultimate Foundation of Economic Science. Menger, Mises, Hayek all recognised that the social reality was historical. The analytical problem was therefore to devise the appropriate analytical tools for 'comprehending' what people did, analytically, together with historical 'understanding' of the specific historical circumstances being investigated. This means (inter alia) replacing the inchoate, contradictory, unsystematic concepts that people in fact use (whether they realise this or not) with systematic, well-thought-out, clear, etc, analytical categories. See, again, Epistemological Problems (available online, I believe), & esp. Hayek's essay on 'Facts of the Social Sciences'.

3. Tariffs are subsidies, from everyone else to the high-cost, unwanted outputs that particular firms are producing. firms that produce viable outputs do not need subsidies, by definition.If the subsidised firms 'automate' or whatever, it is because they have in effect robbed all other firms of the resources _they_ need to improve _their_ machinery. That the subsidised firms do well is so obvious that even journalists, politicians & other lay observers can see it. That all other firms lose out is _not_ obvious -- a tariff doesn't come with a label round its neck saying this. Analytics has to be learned.

Steven Horwitz - 11/24/2006

Yeah, well the Austrians abandoned positivism before it existed, so can we move to talk about the economics actually associated with Mises and Hayek?

Note further that your belief that "better" economies are ones that "reduce costs the most in the long run" would suggest producing nothing. That would reduce costs to zero, no? The whole vision of what economies do that is implicit in your argument is exactly the mechanistic, positivistic, engineering model that you later criticize.

Andrew D. Todd - 11/24/2006

Well, the counterargument for free trade runs as follows: high tariffs discourage employers from looking for cheap labor, and create incentives for mechanization and automation, which reduce costs the most in the long run. The trade issue thus becomes one of the feasibility and desirability of automation. Not only does this depend on the particular circumstances, but it is not a question which economists have any special expertise to answer. The same kinds of arguments apply for the related issue of economy of scale (Adam Smith's hypothetical pin factory).It all depends.

Most of the social sciences abandoned the pretension of positivism by 1970, or thereabouts. Economics has become an outlier, making claims which no one else makes, and without any discoverable factual basis.

Jason Kuznicki - 11/24/2006

Andrew --

I can understand how the various competences, rulesets, and priorities of all the different individual economic actors can make macroeconomics a deeply problematic exercise as it is usually conceived. Yet there still seem some macroeconomic conclusions that can't easily be escaped: How can you claim, for instance, that economists have no particular competence with regard to trade barriers? The argument, from neoclassicals as well as from Austrians, has always been that comparative advantage makes free trade the most productive situation, and that even dropping trade barriers unilaterally is in a country's best interests. This prediction (and at one time it was chiefly a prediction) has been amply borne out in the modern world.

I guess my question to you would be as follows: It seems very, very likely that there are some things we can say intelligently about macroeconomics. If individual choice is a bad foundation to build macroeconomics upon, what foundations are there for a proper (and possibly even systematic) theory?

Sheldon Richman - 11/24/2006

I just wanted to interject a thank you to Jason for raising the question, since it elicited those illuminating responses from Steve and Sudha, and the links from Stephan. I look forward to more. I find foundational Austrian economics endlessly rich and exciting. Sometimes a restatement of something I've thought about before can shine a whole new light on aspects I've never considered. I'm looking forward to the rest.

Andrew D. Todd - 11/24/2006

Well, if you carry Von Mise's idea to the level of someone like Barbara Ward, and her study of Malaysian shopkeepers, there is no such thing as a generalized businessman, a generalized worker, etc. There are vast numbers of groups of businessmen, vast numbers of groups of workers, etc., each group with its own imperatives and rule set. For an example, see Abner Cohen's _Custom and Politics in Urban Africa: A Study of Hausa Migrants in Yoruba Towns_ (1969). This is a study of an outpost, in Ibadan, Nigeria, of a trading diaspora involved in supplying the coastal jungle Guinea Coast region of Nigeria with beef from the greasy upland Western Sudan region, and covers such items as the economic role of religious identity. These vast numbers of rule sets, summed together, do not necessarily add up to anything more than noise, like the noise of a thousand unrelated conversations taking place at once. This implies that there is no such thing as macroeconomic law. That means that economists do not have any special competence in respect of things like budget deficits, taxes, and tariff barriers.

Sudha Shenoy - 11/23/2006

What do indifference curves 'model'? What sort of _action_ by people is best grasped through the device of indifference curves?

Neoclassical economics is interested in mathematising things at all possible points. That is because its aim is to be scientific. Mises (& others in the same analytical line) want to comprehend/understand what people actually _do_. That's another kettle of fish altogether.

Jason Kuznicki - 11/23/2006

I had a purely selfish reason for linking the Amazon version: I get paid for every link from my blog that results in a sale. (Also, I find it acceptable, but much less pleasurable, to read books online. Paper remains the easier choice for me, as for most I think.)

I will have more to say once I've read the papers you linked. Thank you for the pointers.

Stephan Kinsella - 11/23/2006

Not sure why you link to the Amazon version, as it's online: Rothbard's introduction is here :, and the book is online here: . Also, a PDF version of the whole book is at here: And if you want to buy it, buy it from the heroic Mises Institute here instead of Amazon:

Further, for recent discussion of the "indifference" issue see Hans-Hermann Hoppe, A Note on Preference and Indifference in Economic Analysis, available here:, from v. 8, no. 4, of the QJAE; also the draft rejoinder by Walter Block, A Rejoinder to Hoppe on Indifference, available here:

Jason Kuznicki - 11/23/2006

I had the same impression of the Wikipedia description of marginal utility, that it was quite Austrian in perspective. I've read through page 60 of TM&C, and the sections where Mises discusses marginal utility seemed more or less recapitulated there. Still, my question is unanswered: If I can imagine indifference, why is an indifference curve fallacious?

I can see one potential answer, and I think I'll throw it out for discussion purposes: "You," Rothbard might say "are not REALLY indifferent. Deep within you, there is some slight preference somewhere. If you had no preference, then the two goods would actually be ONE good, and there would be no point in discussing them differently."

I don't know, of course, if this is what he meant. Which is why I asked.

Steven Horwitz - 11/23/2006

Oy. :) I'll take my best shot at some of this Jason.

The wikipedia explanation of marginal utility is VERY Austrian (and very right :) ). That is indeed ordinal marginal utility. But in the history of economics that Rothbard is referring to, this notion of MU got lost as the formalization of the discipline increased. Specficially, MU came to mean the first derivative of a total utility function. Such a function is, of mathematical necessity, smooth and not discrete, thus eliminating the notion of distinct units of goods.

At the same time, utility was increasingly understood in a hedonic fashion, as something one feels, rather than as the "usefulness" of a means to an end. Note how those two fit together - if utility is a feeling, the idea that it could be adjusted in marginal amounts by differentiating a total utility function makes sense.

What happened in the 30s is that the formalists threw out the marginal utility baby with the hedonic bathwater. The profession wanted to drain the "psychological" content out of utility as it couldn't be measured. Seeing no other way to talk about MU, they tossed it overboard. This is MNR's first point: there was this whole history of ordinal MU (as opposed to "cardinal" MU) that Hicks et al ignored in their search for mathematical formalism. Economists stopped talking about MU at that point.

Instead, they talked about marginal rates of substitution, which referred to objective quantities of goods and not utility. The indifference curve illustrates this. The problem is that it is premised on the fact that "they render the same level of satisfaction (or the same amount of utility) for the consumer." Note the hedonic and mathematical use of utility there. That is precisely what Mises and the Austrians said utility was NOT. The IC is an attempt to skirt problems created when the profession ignored the ordinal marginal utility of Mises and others. It creates many problems of its own.

One, of course, is whether indifference is a meaningful economic concept. From a Mises/Austrian perspective, economic action is about removing felt uneasiness. If we were truly "indifferent" we would not act. In the real world, we certainly try to allocate our resources in ways that ensure that we are satsifying the most important of all of our ends - using the means with the highest marginal utility. But that's not exactly what the whole IC analysis is trying to do.

Perhaps some of my more micro-economically inclined economics colleagues will help me out here. :)

BTW, TTOMAC is a GREAT book and one that has been formative to my own work in Austrian macro, so I look forward to the rest of your questions.