How Exactly is Marginal Utility Theory Based on Circular Reasoning?

Try giving this a chew over:

https://fixingtheeconomists.wordpress.com/2014/02/17/joan-robinsons-critique-of-marginal-utility-theory/

"Marginalist doctrine claims that we cannot measure utility directly. We know of a person’s utility only due to the fact that they buy something — this is called ‘revealed preferences‘ in the literature. So, we only know the cause — i.e. the utility of a purchase — by the effect it produces — i.e. the actual purchase that is made by the consumer. If we consider preferences as being fixed then this makes some sense. But if we allow that preferences fluctuate the whole edifice falls apart because now we cannot be sure to what extent consumer decisions have changed due to price changes and to what extent they have changed due to a change in preferences."

explainthingstome

"We know of a person's utility only due to the fact that they buy something"

I have two questions about this quote.

a) Surely it is the product, and not the buyer, that has utility? Did the writer of the blog post mean to say "subjective value judgement" instead of "utility"?

I think "we know of a person's utility" can be parsed as "we know what a person finds utility in.."

b) Isn't the marginalist view that the purchase reveals only that the buyers subjective value judgement is either higher or at the same level as the price of the purchased product? I.e. that the purchase only shows what the subjective value judgement of the buyer isn't? (For example if I buy a book for 100 dollars, we only know that my subjective value judgement isn't below 100 dollars.)

In neo-classical economics "marginal-utility" is related to the extra utility that is gained by purchasing an extra unit of a said good. If marginal utility is greater than marginal cost (the cost of buying one more unit) then the consumer will not buy that additional unit. So the theory goes:

https://www.investopedia.com/terms/m/marginalutility.asp

https://www.youtube.com/watch?v=DD128Y0P9YU

https://youtu.be/VVp8UGjECt4?t=159

In a hurry but in Australian economist Steve Keen's book Debunking Economics he makes what seems like a pretty solid claim that the mathematical core of marginalism is incoherent. He also criticises Marxism along similar lines although I don't think explaining relative prices is a key concern for communists.

explainthingstome

Maybe I don't get the term "marginal utility". What would be a good, "easy" synonym?

Marginal utility is concerned with the change in utility that happens when an extra unit of a good or service is consumed, hence "marginal" rather than "total". There's lots of stuff written about it, the Wikipedia page says the same as what you'll get taught at undergraduate economics.

Sidenote: is it wrong to say that marginalists make the strange assumption that the marginal utility of a commodity is the same for every individual?

Yes that would be wrong, they don't say that. But the will refer to "total utility" that is all the individual utilities in an economy added together. If there really is something called "utility" that can be measured and added up in such a manner is a separate question..

You may find this -- and its various references -- of use:

C.1 What is wrong with economics?

It is from An Anarchist FAQ.

Explainthingstome, the Marxian response could point to the sources of demand. In a simplified capitalist economy, demand comes from workers or from capitalists. Workers pay for the commodities they buy with money that they receive as wages. If we assume simple reproduction (i.e. profit is not reinvested), capitalists use profit to buy articles of personal consumption. They also replace used-up means of production using the part of revenue that represents value transferred (by the labor of workers) from the means of production onto the products that capitalists produce and sell. Thus, for both workers and capitalists, demand is limited by the extent of their income.

Take workers. What a worker can effectively demand on the market changes with the wage she receives. If her wage increases (while other prices remain constant), she can buy more and vice versa. But the wage is a kind of price – it's the price of labor power. (We could go further and ask how this price is determined and talk about the value of labor power and how it's determined by the costs of reproducing labor power.) So, on the one hand, demand supposedly determines prices (neoclassical economics says), but on the other hand, it seems that demand (i.e. how much and of what is demanded and bought on the market) is itself determined by prices (first and foremost by the price of labor power, in the case of workers). Hence the circularity.

We could do the same for capitalists. The extent of profit is ultimately determined by the size of capital and the rate of profit (which itself is determined by the rate of surplus value and the composition of capital). So the effective demand of capitalists is in fact a dependent variable as well. Ultimately, we would arrive at capital accumulation and the investment decisions of capitalists as the prime mover behind demand in a capitalist economy.

BTW, as mentioned by bastarx above, Keen's Debunking Economics is a useful critique of the inconsistencies of neoclassical economics. See Chapter 3 specifically on demand and on the impossibility of arriving at a market demand curve as an aggregate of individual demand curves. Keen shows that neoclassical principles only work under extremely unrealistic conditions that have nothing to do with a capitalist society (i.e., a one-consumer, one-commodity market).

explainthingstome

But can't a neo-classical economist just argue that wages (i.e. prices of labour power) are an exception to the rule (i.e. the rule that prices are determined by subjective value judgements)?

The issue isn't that wages are not determined by demand. It's that wages seem to determine demand itself (i.e. they affect all prices).

In general, the problem that plagues marginal utility theory is this: it has a theory of price that seeks to explain what prices are and how they are formed, but to do that, it already presupposes prices. Inidividuals can't form their utility functions without knowing the prices, but the utility functions were supposed to ground prices in the first place. This is what Mattick is talking about, and he's right to point out that the theory, in its original version, was "abandoned". The problem was well known and led to the development of a different approach based on indifference curves and revealed preferences.

At this point, you'd be better off looking at Keen who looks in detail at the problems inherent there. These problems, such as the the income effect (i.e., changes in demand due to changes in income), or the issues stemming from aggregation (i.e., if we want to talk about market demand, we need to combine, somehow, the preferences of individuals), can be avoided by making some highly unrealistic assumptions. The resulting theory is not circular and is not really dependent on the sort of "utility" that Mattick was talking about. However, it assumes what is in effect a market with a single consumer and a single commodity. These underlying assumptions are then overlooked when neoclassical analysis is applied to talk about real-world markets that don't satisfy the assumptions. The answers you've got on Reddit simply gloss over these problems and talk about market supply and demand as if the concepts were intuitive. But, e.g., the "intuitive" downward-sloping demand curve is not in fact intuitive at all and cannot be reliably derived for markets with many consumers and many commodities.

explainthingstome

So basically, modern neo-classical economics is quite different from Bohm-Bawerks marginal utility theory?

The overall thrust (utility maximization) is the same, but the underlying definitions are different. The differences are not very well reflected in textbooks, since they tend to mix the old psychological approach with "revealed preference theory" (which is an attempt at a behaviorist redefinition of the old notions of utility).

explainthingstome

I have to become more knowledgable about neo-classical economics before I do that, but thanks for the help

Hmm, I really think you should give it a try. It's very accessible. In each chapter, he first lays out the neo-classical position, and then talks about the problems. You'll learn more about neoclassical economics in the process.

I'm a few days late to this, but I just wanted to say – posts like Jura's in this thread are the reason why I still find it worthwhile to visit the Libcom forums, and why I hope they won't be entirely gotten rid of in the planned overhaul of the site.

I'd like to point to OP to this text that I translated from German, which deals with the errors of marginalism in detail:

https://libcom.org/library/critique-bourgeois-science-microeconomics-explanation-value-invention-marginal-utility-m

(This could also be interesting: https://libcom.org/library/world-mathematical-curve-misuse-mathematics-humanities-social-sciences)

The translation is obviously not perfect, at times it is too close to the original and hence exhibits traces of German grammar (MG has a dense style that is difficult to translate), but I think you'll gain insights from it nevertheless.

explainthingstome

I am under the impression that me making my "subjective value judgements" means that I'm sitting in a chair thinking things like "1 car = 154 dollars", "1 bicycle = 10 dollars" etc. I.e. that I'm ranking commodities in my mind. I wouldn't have to know the market prices to be able to do this.

In that situation, you might not be confronted with a price tag, but you're still presupposing the very thing that you set out to explain. What is the quality of the concept "value" that is being expressed in "154 dollars" (or "dollars" as such)? The possibility of there being more or less of it - be it in reality or imagination - doesn't tell us anything. Let's suppose you wanted to explain to someone what "length" as such is: What would it help the person if you told them that you can conceive of an object that is five metres long? Moreover, in your example, you're indeed even presupposing market prices (and your income)! Where else would you have received an idea of what 154 dollars mean? Why that exact sum? Why not one million dollars? Why not 1 dollar?

jura

This is what Mattick is talking about, and he's right to point out that the theory, in its original version, was "abandoned". The problem was well known and led to the development of a different approach based on indifference curves and revealed preferences. [...] The resulting theory is not circular and is not really dependent on the sort of "utility" that Mattick was talking about.

This isn't really true. Both "revealed preferences" and indifference curves work off of the same basic idea that the original theory of marginal utility introduced. The text I linked above demonstrates this satisfyingly.

jura

But, e.g., the "intuitive" downward-sloping demand curve is not in fact intuitive at all and cannot be reliably derived for markets with many consumers and many commodities.

The very notion of demand curves is a problem, not to speak of equilibrium, perfect/imperfect competition etc.

@OP: There's a lot to unpack here, and if you want to understand all of this to compare it to what Marx did for example, you're in even more trouble, since Marx was not an economist. What you've been linked here are criticisms of a particular form of economics from the standpoint of the discipline itself. It amounts to reasserting the position of classical political economy vis-à-vis marginalism. Marx however subtitled "Capital" with "critique of political economy" - he has a standpoint outside of the complete discipline as such.