June 2018

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
Blog powered by Typepad

« Austrian Final Exam Question | Main | 2011 Association of Private Enterprise Education Conference »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Disagree: rationalism vs. anti-rationalism.

But I don't their differing views on the nature of reason explains their differences on Mill's Fourth.

Is there any background on this for non-economists?

Disagree - their theory of the interest rate. Not capital structure or statist overlordship, just the interest rate. If it weren't for Leslie Stephens gift-wrapping that as a handy swipe to whip out, I doubt we would hear so much about it.

Mill writes this further on: "I conceive that a person who buys commodities and consumes them himself, does no good to the laboring classes, and that it is only by what he abstains from consuming, and expends in direct payments to laborers in exchange for labor, that he benefits the laboring classes, or adds anything to the amount of their employment."

That's classic Keynesian thinking. Where Keynes perhaps departed from Mill (I don't know Mill well enough to say one way or another), is that Mill might not have considered the prospect that abstention from consumption might not be expended in the employment of labor. If he had opened that door a little wider, nobody would even talk about this section of Mill as contradicting Keynes. As it stands, it's certainly not a contradiction, although Keynes goes place Mill doesn't.

Go down to the bricklayer/velvet example and the similarity between Mill and Keynes becomes even more obvious, and the reductio ad absurdum about the Poor Law is a great illustration of the multiplier (and quite reminiscent of Keynes's own critique of unemployment benefits).

Disagree: forward looking marginal valuational logic vs. backward-looking, cost-adding valuational logic.

.. and now let me explain the curious fact that, although both Mill and Keynes assumed a backwards looking valuation logic, Keynes' deployment of a new version of a backward looking value logic made it impossible for him to appreciate Mill's fourth postulate.

Roger -- arguable Keynes was the anti-rationalist.

Don't you think?

I mean, what is G. E. Moore ethics / ethical intuitionism if it isn't a rejection of reason, a privileging of (elite) taste over reason?

Greg - I was very surprised on that point about rationalism too.

But I was also surprised about your claim that he was backward-looking and cost-adding in his valuation. That's very hard for me to square with his discussion of the marginal efficiency of capital, but perhaps I'm not understanding what you mean. Could you explain? I would have imagined he would definitely be forward-looking, but also marginal.

"Demand for commodities is not demand for labor." (Mill).

I agree. Hayek thinks this is essentially correct, while Keynes' GT is entirely based on the opposite theory. Demand for commodities can create demand for labor, but it does not necessarily do so. It depends on what commodities are being created, and by whom.

The basis of Keynes' policy prescriptions for government spending during a recession is that increased aggregate demand will result in an increase in employment. At first glance, this would seem to make sense. If you increase demand for products, then more products will be produced, and you need more people to make more products. All of this seems perfectly obvious.

Hayek, however, is on the side of Say, who argues that an economy cannot produce just anything, but must produce what people actually want. When that is not happening, one gets a recession. Stimulating already-existing companies does not help, because those are the very companies which were producing the things nobody wanted. More than that, though, there is the problem of the efficiency of established firms.

The idea behind stimulus is that government spending in the economy will put people back to work in the companies that already exist. The government money is necessarily going to those companies which already exist. Those companies are supposed to increase production and, thus, find a need to hire more employees. Like I said, this seems logical on the surface. However, if we understand the typical history of production in a firm or industrial sector, we will see how and why this is mistaken.

When a company first starts up, based on a new product idea or a new way of doing things, such production is necessarily labor-intensive. There is a period during which one has to discover the best way of doing things. Rarely is it the first way you find. Over time, new ways of doing things are discovered, ways of cutting costs without harming quality are discovered, etc., and fewer and fewer people are needed to produce more and more. More automation is introduced, etc. Thus, as an industry grows, the growth rate of employment is not going to go up at the same rate as it did in the early years of production. In fully mature industries, we in fact see a gradual decrease in the number of workers necessary to do produce those goods. An industry, like agriculture, may even go from almost everyone working in the industry to almost no one working in them as mechanization comes to dominate.

Thus, it is when companies are in their infancy that they need large numbers of workers. Start-ups hire the most people. But stimulus package money does not go to startups. It goes to well-established industries. Those industries can increase production without increasing employment because so much of what they do is automated. Run the machines faster, or buy new ones, and you are making more. Thus, the money does not in fact stimulate the economy. If anything, it results in an increase in production in already-established industries, creating a new bubble in those industries -- and/or driving up prices in them. Capital is thus misallocated in the direction of already-existing firms producing products nobody actually wants. In the end, whatever it is that government is spending the money on will begin to have higher prices. Thus will we see inflation, but not improved employment.

During a recession, misallocated captial can either be reallocated into new forms of production, typically by new firms producing new products that people want, or it can be reallocated into old forms of production, typically resulting in new misallocations and the set-up for a new bubble, which will inevitably burst. Hayek recognizes this process, and wants to put an end to it. Even the strawman version of Say's Law that Keynes sets up, that "supply creates demand" does capture a truth that is actually advocated by a Hayekian approach to the economy -- that you have to produce new things that people didn't realize they wanted in order to create economic growth. People can only own so many houses, cars, televisions, and refrigerators -- but with each new iPod, iPhone, iPad, etc. new demand is created by new supply. The new things require more labor, as new processes are labor-intensive. Thus Keynesian stimulus cannot ever work, because it cannot stimulate the only part of the economy that matters for economic growth: (very often new) companies making new kinds of products.

In other words, this forms the basis of the differences on interest rates, capital, etc.


Well, H. said rationalism was an abuse of reason. I mean, you know, how are we defining our terms? It should also be remembered that Keynes was making fun of his earlier beliefs. By the time he wrote "My Early Beliefs" he was more reasonable, or at least thought himself so. As you know, Bill Butos and I go into all this sort of thing in our 1997 HOPE paper on "TheVarieties of Subjectivism: Keynes and Hayek on Expectations."


ungated version:

This point from Keynes is relevant to understanding what his actual objection to Mill is: "As a corollary of that same doctrine, it has been supposed that any individual act of abstaining from consumption necessarily leads to, and amounts to the same thing as, causing the labour and commodities thus released from supplying consumption to be invested in the production of capital wealth... Those who think this way are deceived, nevertheless, by an optical illusion, which makes two essentially different activities appear to be the same. They are fallaciously supposing that there is a nexus which unites decisions to abstain from present consumptionwith decisions to provide for future consumption; whereas the motives of the latter are not linked in any simple way with the motives which determine the former."

The concern is not that demand for commodities don't create demand for labor at all. You find throughout the General Theory the point that Mill made in that section - that new investment drives employment, not consumption. The concern is that Mill was too sanguine about the extent to which refraining from consumption would be channeled to investment. I'm loathe to jump at Troy again today, but I think part of the reason why his analysis is off base is that he doesn't at all distinguish between what Keynes said about investment and what Keynes said about consumption. Indeed, the two words never show up in his comment. For Keynes (and Mill) the distinction was crucial.

Just a quick comment:

I agree that K and H's disagreements about the interest rate are where the action is. However, THOSE disagreements are further premised on their respective views of capital. You simply cannot talk about their differences over interest without discussing their conceptions of capital.

The quote DK deploys there is one that gets at it: Keynes denies there's a market nexus of intertemporal coordination. But that's not just about aggregate S and I; it's also about the microeconomic processes by which those aggregates are effectuated (hence Keynes's famous quote about abstaining from dinner tonight not leading to the demand for anything specific). And here is where I think Keynes really does become very skeptical not just of aggregate level intertemporal coordination but of the ability of the price system to coordinate at the micro level.

For Austrians, the macro level IC requires both monetary equilibrium AND the effective use of monetary calculation by micro-level entrepreneurs.

Steve -
This - "Keynes denies there's a market nexus of intertemporal coordination" - seems like it is overstated. It's not that there isn't a market nexus for intertemporal coordination - it's just that that market nexus is bundled up with a market nexus for liquidity. Money, for Keynes, is essentially a bundled good with one price. Where Mill assumes that money released from consumption is necessarily invested, Keynes points out that this is not true.

Brad DeLong has previously put it (and Nick Rowe has put it in these terms too I believe) that what we're essentially dealing with in a monetary economy is two goods with only one price, which leads to the potential for slack in one of the two markets. There is absolutely a market nexus of intertemporal coordination in Keynes, but it has the potential not to clear or two clear below full employment.

A lot of people misinterpret that to be that there is no intertemporal coordination for Keynes.

"They are fallaciously supposing that there is a nexus which unites decisions to abstain from present consumption with decisions to provide for future consumption;"

How much plainer do you want it Daniel? Seriously. That such a nexus exists is a "fallacy" - Keynes says so in that many words.

S = f(Y)
I = I (exogenously)

Nothing unites S and I. The only way they are equal ex ante is by sheer accident. And ex post equality comes from adjustments in Y and employment. There is NO ex ante nexus to equate investment and savings. Yes, he's just that clear.

No, of course there is no nexus to equate investment and savings. I said there is no guarantee that the market will clear. But that's quite different from saying there is no microeconomic intertemporal coordination, which was your claim. Investment demand is determined by the marginal efficiency of capital, and that is coordinated with savings in the market according to the level of the interest rate. Now, the interest rate is determined for Keynes in the money market, so there is no guarantee that the interest rate will CLEAR the market (as I've said), but that's very different from saying there's no intertemporal coordination. There is intertemporal coordination. There is no guarantee of optimality. And as you say, the difference is made up in shifting output and employment.

It's not "sheer accident" that high MEC investments do occur, but low ones don't. That's not an "accident" - that's intertemporal coordination with a bad price signal. Your understanding of Keynes on this point would be like saying when there's a minimum wage there's no nexus of coordination in the labor market. There IS STILL market coordination in a labor market with a minimum wage, it just doesn't clear the market because of bad price signals. You don't say the coordination of the price mechanism disappears after the imposition of minimum wage - you say that it isn't optimal because the price signal is distorted.

Steve, I think that Daniel has it right. I understand that Keynes wrote what you say, but I think Daniel's interpretation is the better one.
In any case, Daniel's interpretation is the one that should be in play, not an argument that financial markets have nothing to do with channeling savings into investment.

"The interest rate is determined for Keynes in the money market". There lies the greatest Keynes error.

Whatever Keynes' original opinion the modern day Keynesians (at least those who blog) are very keen on this idea.

When I read Krugman and Dan Kuehn I think of them as "Liquidity-Trap-Keynesians".

I guess we're in an interpretive struggle here.

Yes, it's possible that we are at an interest rate that coordinates savings and investment, but that's because the interest rate just happened to be determined correctly from another market. But since neither S nor I is dependent upon that rate, and because neither is dependent on any other common independent variable, there is no "nexus" that coordinates them.

Financial markets can still channel savings into investment, but there remains no reason whatsoever in Keynes's model to think that savings and investment will stand in ex ante coordination other than by chance based on events in other markets.

I guess it's this issue of there being a common independent variable on which both S and I depend that could conceivably create ex ante coordination that is the fundamental issue for me. It doesn't happen perfectly in Hayek's model of course, but perfect isn't the issue.

Isn't anyone going to give Troy a good grade?

I admit to a great deal of confusion about what Keynes says about S and I, at least in Ch. 6 of GT:


I invite everyone to come over and tell me what is going on, because I'm not sure I know.

Hiding behind the Fourth Postulate is Say's Law, which Mill got from his father. Mill's own theory of macro fluctuations involved collapsed investment after bank failures after the collapse of speculative bubbles. Unlike most Austrians he did not blame those entirely on overly low interest rates, but saw them arising fundamentally from psychological processes.

Not in Hayek.

Barkeley writes,

"Hiding behind the Fourth Postulate is Say's Law"

Hayek unwittingly suggests a direct link between Keynes' ethical irrationalism and his macro economics, and its this.

Keynes was more influenced by Moore's ethics than anything else he ever learned.

Moore's ethics justified Keynes' taste for privileging his elite ethical and social- political "intuitions" -- those with the right status and education had a special right and ability to intuitively grasp or directly intuit ethical value.

Hayek points out that Keynes' macro is grounded in the same sort of arrogance and self-conceit -- Keynes' belief that he had the ability to directly intuit the size and significance of macro aggregates .. and their causal inter-relations.

NOTE WELL -- in measuring these aggregates, Keynes uses an aggregated, backward-looking measuring rod, labor input.

For Hayek on these topics see _The Fatal Conceit_ and the UCLA Oral History interviews with Hayek.

"Where Mill assumes that money released from consumption is necessarily invested, Keynes points out that this is not true."

Actually, in Ch. 6 of GT, Keynes argues that

Income = value of output = consumption + investment.
Saving = income - consumption.
Therefore saving = investment. (63)

In GT, at least, it seems that he argues that it is, in fact, true.


That's all ex post in Keynes. The problem is that there was no assurance it would be true ex ante.

And Keynes's persistent inability or unwillingness to make the ex ante / ex post distinction with respect to the loanable funds market is the evidence that he simply did not understand the Swedish tradition as well as he thought he did. And this is a point Hayek makes in later criticisms of Keynes.

Isn't this directly connected to the fact that -- implicitly and often explicitly -- there are no sunk costs in Keynes, and that value is derived from adding up inputs, i.e. costs from the past.

Keynes is always adding up things based on costs -- which are tallied as aggregates based on prior costs, in other words we are returning to the aggregate backwards valued econ of Ricardo.

Call it a counter-revolution against science -- a retrogression to pre-marginalist econ, in a form _worse_ than that achieved by the classicals.

ex post / ex anti tells us something aboit value theory - if it is missing in Keynes, it tells you something about Keynes' value theory.

Steve writes,

"That's all ex post in Keynes. The problem is that there was no assurance it would be true ex ante."

I disagree that it is the causal difference. It may be the difference in one of their arguments. But it is not the causal difference. The difference is that neither of them succeeded despite solving for two different priorities.

Greg Ransom's Right: "forward looking marginal valuational logic vs. backward-looking, cost-adding valuational logic." There are a dozen ways to express the same conceptual difference. But the principle is the same.

Steve Horowitz is discounting Hayek's priority for solving for productive increases from disruptive innovation, rather than efficiency. Which is what entrepreneurs actually do.

Daniel Kuhn is right on DeLong's interpretation of keynes as clearing below full unemployment. Although the nexus argument is a property of the logic being used, not of the economy itself and therefore suspect.

Ivanfoofoo is right in saying that "The interest rate is determined for Keynes in the money market'. There lies the greatest Keynes error."

Fundamentally these commenters are all making the SAME criticism. The net difference is the bias that each of these men brings by stating different priorities (or preferences), and applying biased logic to their problem accordingly. And until we figure out a way measure measure invention, (Hume's problem of induction) we are as stuck with this argument as are the physicists with dark matter. There is hope. But we need enough data, and are slowly accumulating it albiet in odd places, like Google.

A politician or an investor relies upon keynes's work in an effort to create predictability and justify the DSE Model, from which he can profit by exposing asymmetry of information, and therefore solve the material problem of eliminating the scarcity of hard money without causing even worse problems in the economy than the shortage. An entrepreneur solves with Hayek in order to create opportunity for disruption of existing patterns of human exchange from which he can profit by taking personal risks. To a not insignificant degree these goals conflict only because their priorities conflict. And without a mathematics of innovation, our crude and primitive tools of probabilistic analysis are little better than arguing over Divine intent. It's not that we should quibble which of these men is right. It's that the science we try to practice is currently insufficiently provided for by both of them for us to judge their difference in priorities.

Valuation by its nature is trade-offs at the margin, looking toward the future.

There is no marginal trading off at the margin looking into the future, no valuation, when you are summing quantities measured in the past into aggregates -- and charting their gross relative weights into the future.

Oderus -
I think this is the best rendition: http://delong.typepad.com/sdj/2009/06/a-sokratic-dialogue-liquidity-preference-loanable-funds-and-european-hedge-funds-that-fear-the-collapse-of-us-treasury-b.html

It's front and center in the classic IS-LM model too.

My only concern with its rendition in the IS-LM model is that in that model they force the interest rate to clear BOTH markets, and then put all the slack in the output market. There's no good reason why that has to be the case. Keynes, in a 1937/38 (I forget which exactly) to Hicks and others emphatically said he did not think the interest rate was determined in both market - he just thought it was determined in the loanable funds market. I'm not sure I'm as emphatic as Keynes on this point, but I do agree - there's no reason it should necessarily be expected to clear both.

Yes, as Roger said, the way Hayek (and Oakeshott before him) used 'rationalism' did not imply its opposite was unreason; it was rationalism that was counter to reason. As Oakeshott put it in a letter to Popper, "First, of course, when I argue against rationalism, I do not argue against reason. Rationalism in my sense is, among other things, thoroughly unreasonable."

I presented Hayek versus Keynes last semester as part of the ongoing battle over the meaning and universality of Say's Law. (By the way, Sowell makes very clear in his eponymously named book that "Say's Law" really describes a constellation of ideas developed by Smith, Mill Sr. and Say, and that the ideas of the latter two and their epigones changed over time, so I don't think there is much sense in arguing over what Say's Law "really says" or whether Keynes's formulation of it is a strawman, but the way Keynes actually formulates it is: "the aggregate demand price of output as a whole is equal to its aggregate supply price for all volumes of output," i.e., all levels of output are equilibrium levels, and there are no general gluts, which certainly was the contention of many proponents of the law.


That it is not in Hayek does not mean that it is therefore false that Say's Law lies behind Mill's Fourth Postulate. It does. Just read Mill's account of it. Anything that might appear to lead to a change in aggregate spending must involve some reallocation of the spending instead.

Barkley, don't be tendentious.

Mill does not own the cognitive content his "Fourth Postulate".

As I said, Say's Law does not lie behind Hayek's account of why "Mill's Fourth Postulate" is true.

Barkley writes,

"That it is not in Hayek does not mean that it is therefore false that Say's Law lies behind Mill's Fourth Postulate."

Gene -- at some point all the labels just get in the way, they become substitutes for thinking .. and esp. barriers to re-thinking things.

One reason -- the referent of these labels is usually no one thing, but each of these many different things becomes cast in cement.

Think of the label "methodological individualism" -- it's a substitute for thinking, it refers to many different specific things, and the _exemplars_ which explicate the term both tell us more, and allow for open-ended understanding and further substantive yet grounded explication.

Can't someone block "Sleazy P Martini"'s IP address?

There's a nasty troll out there who's infested my cite with the same MO and style.

If you Google the guys IP address you will likely find he's a very nasty piece of work -- wishing people to die from cancer, using endless profanity, and lovely things like that.

Thanks, Steve. That helps.

Curt, can there be a mathematics of innovation? The closest we may come to is concept of adjacent possible of Stuart Kauffman and/or Hector Sabelli's Bios theory, both of which are very abstract. But that may be enough, depending on your purposes.

Done Greg.

Thanks, Steve!

Two quick questions for Daniel Kuehn:

Do you believe that the accumulation of capital may potentially yield a systemic underemployment of resources?

Do you believe that the demand for capital goods varies in the same direction as the demand for consumer goods?

To both - potentially, but not necessarily.

Although it may be more illuminating and perhaps I may answer better if I knew what you were getting at.


If "Barkley" weren't my middle name, "Tendentious" would be. Everybody knows that, :-).

"Tendentious B" could be your rap name.

For me the most fundamental difference was that Keynes had a mechanical view of the aggregate economy while Hayek had a microfundamented view of the same concept. There wasn't such thing as aggregate demand for Hayek: macroeconomic problems would general discoordinations between millions of individual plans and not a relative lack of a certain aggregate variable.

You can tell this is a George Mason question. No math.

That's right, it deals with a complex and not a simple situation. No math can deal with it. It deals with relevant things instead.

While I agree that math cannot deal with everything, math can deal with many relevant things.

Of course it can. It just seems strange that while physics is about 90% math-driven, chemistry is about 70% math-driven, biology is about 20% math-driven, and psychology is about 10% math-driven, that economics, which deals with even more complex conditions, reverts back to at least 50% math-driven.

Disagree. Keynes looked at an economy in terms of monetary aggregates. Fundamentally, an economy *is* those aggregates, and that's the thinking that persists to day (e.g., "economic health" equals "GDP growth"). Another way of putting it is that Keynes in some sense defined the economy in terms of money. For Hayek, the economy is the buzzing process of millions and billions of individual actions. Money is just one more economic good, and aggregates tend to blur or hide the underlying reality.

The comments to this entry are closed.

Our Books