Archive for the 'Earl Thompson' Category

Samuelson Rules the Seas

I think Nick Rowe is a great economist; I really do. And on top of that, he recently has shown himself to be a very brave economist, fearlessly claiming to have shown that Paul Samuelson’s classic 1980 takedown (“A Corrected Version of Hume’s Equilibrating Mechanisms for International Trade“) of David Hume’s classic 1752 articulation of the price-specie-flow mechanism (PSFM) (“Of the Balance of Trade“) was all wrong. Although I am a great admirer of Paul Samuelson, I am far from believing that he was error-free. But I would be very cautious about attributing an error in pure economic theory to Samuelson. So if you were placing bets, Nick would certainly be the longshot in this match-up.

Of course, I should admit that I am not an entirely disinterested observer of this engagement, because in the early 1970s, long before I discovered the Samuelson article that Nick is challenging, Earl Thompson had convinced me that Hume’s account of PSFM was all wrong, the international arbitrage of tradable-goods prices implying that gold movements between countries couldn’t cause the relative price levels of those countries in terms of gold to deviate from a common level, beyond the limits imposed by the operation of international commodity arbitrage. And Thompson’s reasoning was largely restated in the ensuing decade by Jacob Frenkel and Harry Johnson (“The Monetary Approach to the Balance of Payments: Essential Concepts and Historical Origins”) and by Donald McCloskey and Richard Zecher (“How the Gold Standard Really Worked”) both in the 1976 volume on The Monetary Approach to the Balance of Payments edited by Johnson and Frenkel, and by David Laidler in his essay “Adam Smith as a Monetary Economist,” explaining why in The Wealth of Nations Smith ignored his best friend Hume’s classic essay on PSFM. So the main point of Samuelson’s takedown of Hume and the PSFM was not even original. What was original about Samuelson’s classic article was his dismissal of the rationalization that PSFM applies when there are both non-tradable and tradable goods, so that national price levels can deviate from the common international price level in terms of tradables, showing that the inclusion of tradables into the analysis serves only to slow down the adjustment process after a gold-supply shock.

So let’s follow Nick in his daring quest to disprove Samuelson, and see where that leads us.

Assume that durable sailing ships are costly to build, but have low (or zero for simplicity) operating costs. Assume apples are the only tradeable good, and one ship can transport one apple per year across the English Channel between Britain and France (the only countries in the world). Let P be the price of apples in Britain, P* be the price of apples in France, and R be the annual rental of a ship, (all prices measured in gold), then R=ABS(P*-P).

I am sorry to report that Nick has not gotten off to a good start here. There cannot be only tradable good. It takes two tango and two to trade. If apples are being traded, they must be traded for something, and that something is something other than apples. And, just to avoid misunderstanding, let me say that that something is also something other than gold. Otherwise, there couldn’t possibly be a difference between the Thompson-Frenkel-Johnson-McCloskey-Zecher-Laidler-Samuelson critique of PSFM and the PSFM. We need at least three goods – two real goods plus gold – providing a relative price between the two real goods and two absolute prices quoted in terms of gold (the numeraire). So if there are at least two absolute prices, then Nick’s equation for the annual rental of a ship R must be rewritten as follows R=ABS[P(A)*-P(A)+P(SE)*-P(SE)], where P(A) is the price of apples in Britain, P(A)* is the price of apples in France, P(SE) is the price of something else in Britain, and P(SE)* is the price of that same something else in France.

OK, now back to Nick:

In this model, the Law of One Price (P=P*) will only hold if the volume of exports of apples (in either direction) is unconstrained by the existing stock of ships, so rentals on ships are driven to zero. But then no ships would be built to export apples if ship rentals were expected to be always zero, which is a contradiction of the Law of One Price because arbitrage is impossible without ships. But an existing stock of ships represents a sunk cost (sorry) and they keep on sailing even as rentals approach zero. They sail around Samuelson’s Iceberg model (sorry) of transport costs.

This is a peculiar result in two respects. First, it suggests, perhaps inadvertently, that the law of price requires equality between the prices of goods in every location when in fact it only requires that prices in different locations not differ by more than the cost of transportation. The second, more serious, peculiarity is that with only one good being traded the price difference in that single good between the two locations has to be sufficient to cover the cost of building the ship. That suggests that there has to be a very large price difference in that single good to justify building the ship, but in fact there are at least two goods being shipped, so it is the sum of the price differences of the two goods that must be sufficient to cover the cost of building the ship. The more tradable goods there are, the smaller the price differences in any single good necessary to cover the cost of building the ship.

Again, back to Nick:

Start with zero exports, zero ships, and P=P*. Then suppose, like Hume, that some of the gold in Britain magically disappears. (And unlike Hume, just to keep it simple, suppose that gold magically reappears in France.)

Uh-oh. Just to keep it simple? I don’t think so. To me, keeping it simple would mean looking at one change in initial conditions at a time. The one relevant change – the one discussed by Hume – is a reduction in the stock of gold in Britain. But Nick is looking at two changes — a reduced stock of gold in Britain and an increased stock of gold in France — simultaneously. Why does it matter? Because the key point at issue is whether a national price level – i.e, Britain’s — can deviate from the international price level. In Nick’s two-country example, there should be one national price level and one international price level, which means that the only price level subject to change as a result of the change in initial conditions should be, as in Hume’s example, the British price level, while the French price level – representing the international price level – remained constant. In a two-country model, this can only be made plausible by assuming that France is large compared to Britain, so that a loss of gold could potentially affect the British price level without changing the French price level. Once again back to Nick.

The price of apples in Britain drops, the price of apples in France rises, and so the rent on a ship is now positive because you can use it to export apples from Britain to France. If that rent is big enough, and expected to stay big long enough, some ships will be built, and Britain will export apples to France in exchange for gold. Gold will flow from France to Britain, so the stock of gold will slowly rise in Britain and slowly fall in France, and the price of apples will likewise slowly rise in Britain and fall in France, so ship rentals will slowly fall, and the price of ships (the Present Value of those rents) will eventually fall below the cost of production, so no new ships will be built. But the ships already built will keep on sailing until rentals fall to zero or they rot (whichever comes first).

So notice what Nick has done. Instead of confronting the Thompson-Frenkel-Johnson-McCloseky-Zecher-Laidler-Samuelson critique of Hume, which asserts that a world price level determines the national price level, Nick has simply begged the question by not assuming that the world price of gold, which determines the world price level, is constant. Instead, he posits a decreased value of gold in France, owing to an increased French stock of gold, and an increased value of gold in Britain, owing to a decreased British stock of gold, and then conflating the resulting adjustment in the value gold with the operation of commodity arbitrage. Why Nick thinks his discussion is relevant to the Thompson-Frenkel-Johnson-McCloseky-Zecher-Laidler-Samuelson critique escapes me.

The flow of exports and hence the flow of specie is limited by the stock of ships. And only a finite number of ships will be built. So we observe David Hume’s price-specie flow mechanism playing out in real time.

This bugs me. Because it’s all sorta obvious really.

Yes, it bugs me, too. And, yes, it is obvious. But why is it relevant to the question under discussion, which is whether there is an international price level in terms of gold that constrains movements in national price levels in countries in which gold is the numeraire. In other words, if there is a shock to the gold stock of a small open economy, how much will the price level in that small open economy change? By the percentage change in the stock of gold in that country – as Hume maintained – or by the minisicule percentage change in the international stock of gold, gold prices in the country that has lost gold being constrained from changing by more than allowed by the cost of arbitrage operations? Nick’s little example is simply orthogonal to the question under discussion.

I skip Nick’s little exegetical discussion of Hume’s essay and proceed to what I think is the final substantive point that Nick makes.

Prices don’t just arbitrage themselves. Even if we take the limit of my model, as the cost of building ships approaches zero, we need to explain what process ensures the Law of One Price holds in equilibrium. Suppose it didn’t…then people would buy low and sell high… know the rest.

There are different equilibrium conditions being confused here. The equilibrium arbitrage conditions are not same as the equilibrium conditions for international monetary equilibrium. Arbitrage conditions for individual commodities can hold even if the international distribution of gold is not in equilibrium. So I really don’t know what conclusion Nick is alluding to here.

But let me end on what I hope is a conciliatory and constructive note. As always, Nick is making an insightful argument, even if it is misplaced in the context of Hume and PSFM. And the upshot of Nick’s argument is that transportation costs are a function of the dispersion of prices, because, as the incentive to ship products to capture arbitrage profits increases, the cost of shipping will increase as arbitragers bid up the value of resources specialized to the processes of transporting stuff. So the assumption that the cost of transportation can be treated as a parameter is not really valid, which means that the constraints imposed on national price level movements are not really parametric, they are endongenously determined within an appropriately specified general equilibrium model. If Nick is willing to settle for that proposition, I don’t think that our positions are that far apart.

Helicopter Money and the Reflux Problem

Although I try not to seem overly self-confident or self-satisfied, I do give myself a bit of credit for being willing to admit my mistakes, of which I’ve made my share. So I am going to come straight out and admit it up front: I have not been reading Nick Rowe’s blog lately. Realizing my mistake, I recently looked up his posts for the past few months. Reading one of Nick’s posts is always an educational experience, teaching us how to think about an economic problem in the way that a good – I mean a really good — economist ought to think about the problem. I don’t always agree with Nick, but in trying to figure out whether I agree — and if not, why not — I always find that I have gained some fresh understanding of, or a deeper insight into, the problem than I had before. So in this post, I want to discuss a post that Nick wrote for his blog a couple of months ago on “helicopter money” and the law of reflux. Nick and I have argued about the law of reflux several times (see, e.g., here, here and here, and for those who just can’t get enough here is J. P. Koning’s take on Rowe v. Glasner) and I suspect that we still don’t see eye to eye on whether or under what circumstances the law of reflux has any validity. The key point that I have emphasized is that there is a difference in the way that commercial banks create money and the way that a central bank or a monetary authority creates money. In other words, I think that I hold a position somewhere in between Nick’s skepticism about the law of reflux and Mike Sproul’s unqualified affirmation of the law of reflux. So the truth is that I don’t totally disagree with what Nick writes about helicopter money. But I think it will help me and possibly people who read this post if I can explain where and why I take issue with what Nick has to say on the subject of helicopter money.

Nick begins his discussion with an extreme example in which people have a fixed and unchanging demand for money – one always needs to bear in mind that when economists speak about a demand for money they mean a demand to hold money in their wallets or their bank accounts. People will accept money in excess of their demand to hold money, but if the amount of money that they have in their wallets or in their bank accounts is more than desired, they don’t necessarily take immediate steps to get rid of their excess cash, though they will be more tolerant of excess cash in their bank accounts than in their wallets. So if central bank helicopters start bombarding the population with piles of new cash, those targeted will pick up the cash and put the cash in their wallets or deposit it into their bank accounts, but they won’t just keep the new cash in their wallets or their banks accounts permanently, because they will generally have better options for the superfluous cash than just leaving it in their wallets or their bank accounts. But what else can they do with their excess cash?

Well the usual story is that they spend the cash. But what do they spend it on? And the usual answer is that they buy stuff with the excess cash, causing a little consumption boom that either drives up prices of goods and services, or possibly, if wages and prices are “sticky,” causes total output to increase (at least temporarily unless the story starts from an initial condition of unemployed resources). And that’s what Nick seems to be suggesting in this passage.

If the central bank prints more currency, and drops it out of a helicopter, will the people refuse to pick it up, and leave the newly-printed notes lying on the sidewalk?

No. That’s silly. They will pick it up, and spend it. Each individual knows he can get rid of any excess money, even though it is impossible for individuals in the aggregate to get rid of excess money. What is true for each individual is false for the whole. It’s a fallacy of composition to assume otherwise.

But this version of the story is problematic for the same reason that early estimates of the multiplier in Keynesian models were vastly overstated. A one-time helicopter drop of money will be treated by most people as a windfall, not as a permanent increase in their income, so that it will not cause people to increase their spending on stuff except insofar as they expect their permanent income to have increased. So the main response of most people to the helicopter drop will be to make some adjustments in the composition of their balance sheets. People may use the cash to buy other income generating assets (including consumer durables), but they will hardly change their direct expenditures on present consumption.

So what else could people do with excess cash besides buying consumer durables? Well, they could buy real or financial assets (e.g., houses and paintings or bonds) driving up the value of those assets, but it is not clear why the value of those assets, which fundamentally reflect the expected future flows of real services or cash associated with those assets and the rates at which people discount future consumption relative to present consumption, is should be affected by an increase in the amount of cash that people happen to be holding at any particular moment in time. People could also use their cash to pay off debts, but that would just mean that the cash held by debtors would be transferred into the hands of their creditors. So the question what happens to the excess cash, and, if nothing happens to it, how the excess cash comes to be willingly held is not an easy question to answer.

Being the smart economist that he is, Nick understands the problem and he addresses it a few paragraphs below in a broader context in which people can put cash into savings accounts as well as spend it on stuff.

Now let me assume that the central bank also offers savings accounts, as well as issuing currency. Savings accounts may pay interest (at a rate set by the central bank), but cannot be used as a medium of exchange.

Start in equilibrium where the stock of currency is exactly $100 per person. What happens if the central bank prints more currency and drops it out of a helicopter, holding constant the nominal rate of interest it pays on savings accounts?

I know what you are thinking. I know how most economists would be thinking. (At least, I think I do.) “Aha! This time it’s different! Because now people can get rid of the excess currency, by depositing it in their savings accounts at the central bank, so Helicopter Money won’t work.” You are implicitly invoking the Law of Reflux to say that an excess supply of money must return to the bank that issued that money.

And you are thinking wrong. You are making exactly the same fallacy of composition as you would have been making if you said that people would leave the excess currency lying on the sidewalk.People in aggregate can only get rid of the excess currency by depositing it in their savings accounts (or throwing it away) therefore each individual will get rid of his excess currency by depositing it in his savings account (since it’s better than throwing it away).

There are 1,001 different ways an individual can get rid of excess currency, and depositing it in his savings account is only one of those 1,001 ways. Why should an individual care if depositing it in his savings account is the only way that works for the aggregate? (If people always thought like that, littering would never be a problem.) And if individuals do spend any portion of their excess currency, so that NGDP rises, and is expected to keep in rising, then the (assumed fixed) nominal interest rate offered on savings accounts at the central bank will start to look less attractive, and people will actually withdraw money from their savings accounts. Not because they want to hold extra currency, but because they plan to spend it.

There are indeed 1,001 ways that people could dispose of their excess cash balances, but how many of those 1,001 ways would be optimal under the assumptions of Nick’s little thought experiment? Not that many, because optimal spending decisions would be dictated by preferences for consumption over time, and there is no reason to assume that optimal spending plans would be significantly changed by the apparent, and not terribly large, wealth windfall associated with the helicopter drops. There could be some increase in purchases of assets like consumer durables, but one would expect that most of the windfall would be used to retire debt or to acquire interest-earning assets like central-bank deposits or their equivalent.

So, to be clear, I am not saying that Nick has it all wrong; I don’t deny that there could be some increase in expenditures on stuff; all I am saying is that in the standard optimizing models that we use, the implied effect on spending from an increase in cash holding seems to be pretty small.

Nick then goes on to bring commercial banks into his story.

The central bank issues currency, and also offers accounts at which central banks can keep “reserves”. People use both central bank currency and commercial bank chequing accounts as their media of exchange; commercial banks use their reserve accounts at the central bank as the medium of exchange they use for transactions between themselves. And the central bank allows commercial banks to swap currency for reserves in either direction, and reserves pay a nominal rate of interest set by the central bank.

My story now (as best as I can tell) matches the (implicit) model in “Helicopter Money: the Illusion of a Free Lunch” by Claudio Borio, Piti Disyatat, and Anna Zabai. (HT Giles Wilkes.) They argue that Helicopter Money will be unwanted and must Reflux to the central bank to be held as central bank reserves, where those reserves pay interest and so are just like (very short-term) government bonds, or savings accounts at the central bank. Their argument rests on a fallacy of composition. Individuals in aggregate can only get rid of unwanted currency that way, but this does not mean that individuals will choose to get rid of unwanted currency that way.

It seems to me that the effect that Nick is relying on is rather weak. If non-interest-bearing helicopter money can be costlessly converted into interest-bearing reserves at the central bank, then commercial banks will compete with each other to induce people with unwanted helicopter money in their pockets to convert the cash into interest-bearing deposits, so that the banks can pocket the interest on reserves. Competition will force the banks to share their interest income with depositors. Again, there may be some increase in spending on stuff associated with the helicopter drops, but it seems unlikely that it would be very large relative to the size of the drop.

It seems to me that the only way to answer the question how an excess supply of cash following a helicopter drop gets eliminated is to use the idea proposed by Earl Thompson over 40 years ago in his seminal, but unpublished, paper “A Reformulation of Macroeconomic Theory” which I have discussed in five posts (here, here, here, here and here) over the past four years. Even as I write this sentence, I feel a certain thrill of discovery in understanding more clearly than I ever have before the profound significance of Earl’s insight. The idea is simply this: in any intertemporal macroeconomic model, the expected rate of inflation, or the expected future price level, has to function, not as a parameter, but as an equilibrating variable. In any intertemporal macromodel, there will be a unique expected rate of inflation, or expected future price level, that is consistent with equilibrium. If actual expected inflation equals the equilibrium expected rate the economy may achieve its equilibrium, if the actual expected rate does not equal the equilibrium expected rate, the economy cannot reach equilibrium.

So if the monetary authority bombards its population with helicopter money, the economy will not reach equilibrium unless the expected rate of inflation of the public equals the rate of inflation (or the future price level) that is consistent with the amount of helicopter money being dropped by the monetary authority. But the fact that the expected rate of inflation is an equilibrating variable tells us nothing – absolutely nothing – about whether there is any economic mechanism whereby the equilibrium expectation of inflation is actually realized. The reason that the equilibrium value of expected inflation tells us nothing about the mechanism by which the equilibrium expected rate of inflation is achieved is that the mechanism does not exist. If it pleases you to say that rational expectations is such a mechanism, you are free to do so, but it should be obvious that the assertion that rational expectations ensures that the the actual expected rate of inflation is the equilibrium expected rate of inflation is nothing more than an exercise in question begging.

And it seem to me that, in explaining why helicopter drops are not nullified by reflux, Nick is implicitly relying on a change in inflation expectations as a reason why putting money into savings accounts will not eliminate the excess supply of cash. But it also seems to me that Nick is just saying that for equilibrium to be restored after a helicopter drop, inflation expectations have to change. Nothing I have said above should be understood to deny the possibility that inflation expectations could change as a result of a helicopter drop. In fact I think there is a strong likelihood that helicopter drops change inflation expectations. The point I am making is that we should be clear about whether we are making a contingent – potentially false — assertion about a causal relationship or making a logically necessary inference from given premises.

Thus, moving away from strictly logical reasoning, Nick makes an appeal to experience to argue that helicopter drops are effective.

We know, empirically, that helicopter money (in moderation of course) does not lead to bizarre consequences. Helicopter money is perfectly normal; central banks do it (almost) all the time. They print currency, the stock of currency grows over time, and since that currency pays no interest this is a profitable business for central banks and the governments that own them.

Ah yes, in the good old days before central banks started paying interest on reserves. After it became costless to hold money, helicopter drops aren’t what they used to be.

The demand for central bank currency seems to rise roughly in proportion to NGDP (the US is maybe an exception, since much is held abroad), so countries with rising NGDP are normally doing helicopter money. And doing helicopter money, just once, does not empirically lead to central banks being forced to set nominal interest rates at zero forever. And it would be utterly bizarre if it did; what else are governments supposed to do with the profits central banks earn from printing paper currency?

Why, of course! Give them to the banks by paying interest on reserves. Nick concludes with this thought.

The lesson we learn from all this is that the Law of Reflux will prevent Helicopter Money from working only if the central bank refuses to let NGDP rise at the same time. Which is like saying that pressing down on the gas pedal won’t work if you press the brake pedal down hard enough so the car can’t accelerate.

I would put it slightly differently. If the central bank engages in helicopter drops while simultaneously proclaiming that its inflation target is below the rate of inflation consistent with its helicopter drops, reflux may prevent helicopter drops from having any effect.

The Free Market Economy Is Awesome and Fragile

Scott Sumner’s three most recent posts (here, here, and here)have been really great, and I’ld like to comment on all of them. I will start with a comment on his post discussing whether the free market economy is stable; perhaps I will get around to the other two next week. Scott uses a 2009 paper by Robert Hetzel as the starting point for his discussion. Hetzel distinguishes between those who view the stabilizing properties of price adjustment as being overwhelmed by real instabilities reflecting fluctuations in consumer and entrepreneurial sentiment – waves of optimism and pessimism – and those who regard the economy as either perpetually in equilibrium (RBC theorists) or just usually in equilibrium (Monetarists) unless destabilized by monetary shocks. Scott classifies himself, along with Hetzel and Milton Friedman, in the latter category.

Scott then brings Paul Krugman into the mix:

Friedman, Hetzel, and I all share the view that the private economy is basically stable, unless disturbed by monetary shocks. Paul Krugman has criticized this view, and indeed accused Friedman of intellectual dishonesty, for claiming that the Fed caused the Great Depression. In Krugman’s view, the account in Friedman and Schwartz’s Monetary History suggests that the Depression was caused by an unstable private economy, which the Fed failed to rescue because of insufficiently interventionist monetary policies. He thinks Friedman was subtly distorting the message to make his broader libertarian ideology seem more appealing.

This is a tricky topic for me to handle, because my own view of what happened in the Great Depression is in one sense similar to Friedman’s – monetary policy, not some spontaneous collapse of the private economy, was what precipitated and prolonged the Great Depression – but Friedman had a partial, simplistic and distorted view of how and why monetary policy failed. And although I believe Friedman was correct to argue that the Great Depression did not prove that the free market economy is inherently unstable and requires comprehensive government intervention to keep it from collapsing, I think that his account of the Great Depression was to some extent informed by his belief that his own simple k-percent rule for monetary growth was a golden bullet that would ensure economic stability and high employment.

I’d like to first ask a basic question: Is this a distinction without a meaningful difference? There are actually two issues here. First, does the Fed always have the ability to stabilize the economy, or does the zero bound sometimes render their policies impotent?  In that case the two views clearly do differ. But the more interesting philosophical question occurs when not at the zero bound, which has been the case for all but one postwar recession. In that case, does it make more sense to say the Fed caused a recession, or failed to prevent it?

Here’s an analogy. Someone might claim that LeBron James is a very weak and frail life form, whose legs will cramp up during basketball games without frequent consumption of fluids. Another might suggest that James is a healthy and powerful athlete, who needs to drink plenty of fluids to perform at his best during basketball games. In a sense, both are describing the same underlying reality, albeit with very different framing techniques. Nonetheless, I think the second description is better. It is a more informative description of LeBron James’s physical condition, relative to average people.

By analogy, I believe the private economy in the US is far more likely to be stable with decent monetary policy than is the economy of Venezuela (which can fall into depression even with sufficiently expansionary monetary policy, or indeed overly expansionary policies.)

I like Scott’s LeBron James analogy, but I have two problems with it. First, although LeBron James is a great player, he’s not perfect. Sometimes, even he messes up. When he messes up, it may not be his fault, in the sense that, with better information or better foresight – say, a little more rest in the second quarter – he might have sunk the game-winning three-pointer at the buzzer. Second, it’s one thing to say that a monetary shock caused the Great Depression, but maybe we just don’t know how to avoid monetary shocks. LeBron can miss shots, so can the Fed. Milton Friedman certainly didn’t know how to avoid monetary shocks, because his pet k-percent rule, as F. A. Hayek shrewdly observed, was a simply a monetary shock waiting to happen. And John Taylor certainly doesn’t know how to avoid monetary shocks, because his pet rule would have caused the Fed to raise interest rates in 2011 with possibly devastating consequences. I agree that a nominal GDP level target would have resulted in a monetary policy superior to the policy the Fed has been conducting since 2008, but do I really know that? I am not sure that I do. The false promise held out by Friedman was that it is easy to get monetary policy right all the time. It certainly wasn’t the case for Friedman’s pet rule, and I don’t think that there is any monetary rule out there that we can be sure will keep us safe and secure and fully employed.

But going beyond the LeBron analogy, I would make a further point. We just have no theoretical basis for saying that the free-market economy is stable. We can prove that, under some assumptions – and it is, to say the least, debatable whether the assumptions could properly be described as reasonable – a model economy corresponding to the basic neoclassical paradigm can be solved for an equilibrium solution. The existence of an equilibrium solution means basically that the neoclassical model is logically coherent, not that it tells us much about how any actual economy works. The pieces of the puzzle could all be put together in a way so that everything fits, but that doesn’t mean that in practice there is any mechanism whereby that equilibrium is ever reached or even approximated.

The argument for the stability of the free market that we learn in our first course in economics, which shows us how price adjusts to balance supply and demand, is an argument that, when every market but one – well, actually two, but we don’t have to quibble about it – is already in equilibrium, price adjustment in the remaining market – if it is small relative to the rest of the economy – will bring that market into equilibrium as well. That’s what I mean when I refer to the macrofoundations of microeconomics. But when many markets are out of equilibrium, even the markets that seem to be equilibrium (with amounts supplied and demanded equal) are not necessarily in equilibrium, because the price adjustments in other markets will disturb the seeming equilibrium of the markets in which supply and demand are momentarily equal. So there is not necessarily any algorithm, either in theory or in practice, by which price adjustments in individual markets would ever lead the economy into a state of general equilibrium. If we believe that the free market economy is stable, our belief is therefore not derived from any theoretical proof of the stability of the free market economy, but simply on an intuition, and some sort of historical assessment that free markets tend to work well most of the time. I would just add that, in his seminal 1937 paper, “Economics and Knowledge,” F. A. Hayek actually made just that observation, though it is not an observation that he, or most of his followers – with the notable and telling exceptions of G. L. S. Shackle and Ludwig Lachmann – made a big fuss about.

Axel Leijonhufvud, who is certainly an admirer of Hayek, addresses the question of the stability of the free-market economy in terms of what he calls a corridor. If you think of an economy moving along a time path, and if you think of the time path that would be followed by the economy if it were operating at a full-employment equilibrium, Leijonjhufvud’s corridor hypothesis is that the actual time path of the economy tends to revert to the equilibrium time path as long as deviations from the equilibrium are kept within certain limits, those limits defining the corridor. However, if the economy, for whatever reasons (exogenous shocks or some other mishaps) leaves the corridor, the spontaneous equilibrating tendencies causing the actual time path to revert back to the equilibrium time path may break down, and there may be no further tendency for the economy to revert back to its equilibrium time path. And as I pointed out recently in my post on Earl Thompson’s “Reformulation of Macroeconomic Theory,” he was able to construct a purely neoclassical model with two potential equilibria, one of which was unstable so that a shock form the lower equilibrium would lead either to a reversion to the higher-level equilibrium or to downward spiral with no endogenous stopping point.

Having said all that, I still agree with Scott’s bottom line: if the economy is operating below full employment, and inflation and interest rates are low, there is very likely a problem with monetary policy.

Thompson’s Reformulation of Macroeconomic Theory, Part V: A Neoclassical Black Hole

It’s been over three years since I posted the fourth of my four previous installments in this series about Earl Thompson’s unpublished paper “A Reformulation of Macroeconomic Theory,” Thompson’s strictly neoclassical alternative to the standard Keynesian IS-LM model. Given the long hiatus, a short recapitulation seems in order.

The first installment was an introduction summarizing Thompson’s two main criticisms of the Keynesian model: 1) the disconnect between the standard neoclassical marginal productivity theory of production and factor pricing and the Keynesian assertion that labor receives a wage equal to its marginal product, thereby implying the existence of a second scarce factor of production (capital), but with the market for capital services replaced in the IS-LM model by the Keynesian expenditure functions, creating a potential inconsistency between the IS-LM model and a deep property of neoclassical theory; 2) the market for capital services having been excluded from the IS-LM model, the model lacks a variable that equilibrates the choice between holding money or real assets, so that the Keynesian investment function is incompletely specified, the Keynesian equilibrium condition for spending – equality between savings and investment – taking no account of the incentive for capital accumulation or the relationship, explicitly discussed by Keynes, between current investment and the (expected) future price level. Excluding the dependence of the equilibrium rate of spending on (expected) inflation from the IS-LM model renders the model logically incomplete.

The second installment was a discussion of the Hicksian temporary-equilibrium method used by Thompson to rationalize the existence of involuntary unemployment. For Thompson involuntary unemployment means unemployment caused by overly optimistic expectations by workers of wage offers, leading them to mistakenly set reservation wages too high. The key idea of advantage of the temporary-equilibrium method is that it reconciles the convention of allowing a market-clearing price to equilibrate supply and demand with the phenomenon of substantial involuntary unemployment in business-cycle downturns. Because workers have an incentive to withhold their services in order to engage in further job search or job training or leisure, their actual short-run supply of labor services in a given time period is highly elastic at the expected wage. If wage offers are below expectations, workers (mistakenly = involuntarily) choose unemployment, but given those mistaken expectations, the labor market is cleared with the observed wage equilibrating the demand for labor services and supply of labor services. There are clearly problems with this way of modeling the labor market, but it does provide an analytical technique that can account for cyclical fluctuations in unemployment within a standard microeconomic framework.

In the third installment, I showed how Thompson derived his FF curve, representing combinations of price levels and interest rates consistent with (temporary) equilibrium in both factor markets (labor services and capital services) and two versions of the LM curve, representing price levels and interest rates consistent with equilibrium in the money market. The two versions of the LM curve (analogous, but not identical, to the Keynesian LM curve) correspond to different monetary regimes. In what Thompson called the classical case, the price level is fixed by convertibility of output into cash at a fixed exchange rate, with money being supplied by a competitive banking system paying competitive interest on cash balances. The LM curve in this case is vertical at the fixed price level, with any nominal rate of interest being consistent with equilibrium in the money market, inasmuch as the amount of money demanded depends not on the nominal interest rate, but on the difference between the nominal interest rate and the competitively determined interest rate paid on cash. In the modern case, cash is non-interest bearing and supplied monopolistically by the monetary authority, so the LM curve is upward-sloping, with the cost of holding cash rising with the rate of interest, thereby reducing the amount of money demanded and increasing the price level for a given quantity of money supplied by the monetary authority. The solution of the model corresponds to the intersection of the FF and LM curves. For the classical case, the intersection is unique, but in the modern case since both curves are upward sloping, multiple intersections are possible.

The focus of the fourth installment was on setting up a model analogous to the Keynesian model by replacing the market for capital services excluded by Walras’s Law with something similar to the Keynesian expenditure functions (consumption, investment, government spending, etc.). The key point is that the FF and LM curves implicitly define a corresponding CC curve (shown in Figure 4 of the third installment) with the property that, at all points on the CC curve, the excess demand for (supply of) money exactly equals the excess supply of (demand for) labor. Thus, the CC curve represents a stock equilibrium in the market for commodities (i.e., a single consumption/capital good) rather than a flow rate of expenditure and income as represented by the conventional IS curve. But the inconsistency between the upward-sloping CC curve and the downward sloping IS curve reflects the underlying inconsistency between the neoclassical and the Keynesian paradigms.

In this installment, I am going to work through Thompson’s argument about the potential for an unstable equilibrium in the version of his model with an upward-sloping LM curve corresponding to the case in which non-interest bearing money is monopolistically supplied by a central bank. Thompson makes the argument using Figure 5, a phase diagram showing the potential equilibria for such an economy in terms of the FF curve (representing price levels and nominal interest rates consistent with equilibrium in the markets for labor and capital services) and the CC curve (representing price levels and nominal interest rates consistent with equilibrium in the output market).

Thompson_Figure5A phase diagram shows the direction of price adjustment when the economy is not in equilibrium (one of the two points of intersection between the FF and the CC curves). A disequilibrium implies a price change in response to an excess supply or excess demand in some market. All points above and to the left of the FF curve correspond to an excess supply of capital services, implying a falling nominal interest rate; points below and to the right of the FF curve correspond to excess demand for capital services, implying a rising interest rate. Points above and to the left of the CC curve correspond to an excess demand for output, implying a rising price level; points below and to the right of the CC curve correspond to an excess supply of output, implying a falling price level. Points in between the FF and CC curves correspond either to an excess demand for commodities and for capital services, implying a rising price level and a rising nominal interest rate (in the region between the two points of intersection – Eu and Es — between the CC and FF curves) or to an excess supply of both capital services and commodities, implying a falling interest rate and a falling price level (in the regions below the lower intersection Eu and above the upper intersection Es). The arrows in the diagram indicate the direction in which the price level and the nominal interest rate are changing at any point in the diagram.

Given the direction of price change corresponding to points off the CC and FF curves, the upper intersection is shown to be a stable equilibrium, while the lower intersection is unstable. Moreover, the instability corresponding to the lower intersection is very dangerous, because entering the region between the CC and FF curves below Eu means getting sucked into a vicious downward spiral of prices and interest rates that can only be prevented by a policy intervention to shift the CC curve to the right, either directly by way of increased government spending or tax cuts, or indirectly, through monetary policy aimed at raising the price level and expected inflation, shifting the LM curve, and thereby the CC curve, to the right. It’s like stepping off a cliff into a black hole.

Although I have a lot of reservations about the practical relevance of this model as an analytical tool for understanding cyclical fluctuations and counter-cyclical policy, which I plan to discuss in a future post, the model does resonate with me, and it does so especially after my recent posts about the representative-agent modeling strategy in New Classical economics (here, here, and here). Representative-agent models, I argued, are inherently unable to serve as analytical tools in macroeconomics, because their reductionist approach implies that all relevant decision making can be reduced to the optimization of a single agent, insulating the analysis from any interactions between decision-makers. But it is precisely the interaction effects between decision makers that create analytical problems that constitute the subject matter of the discipline or sub-discipline known as macroeconomics. That Robert Lucas has made it his life’s work to annihilate this field of study is a sad commentary on his contribution, Nobel Prize or no Nobel Prize, as an economic theorist.

That is one reason why I regard Thompson’s model, despite its oversimplifications, as important: it is constructed on a highly aggregated, yet strictly neoclassical, foundation, including continuous market-clearing, arriving at the remarkable conclusion that not only is there an unstable equilibrium, but it is at least possible for an economy in the neighborhood of the unstable equilibrium to be caught in a vicious downward deflationary spiral in which falling prices do not restore equilibrium but, instead, suck the economy into a zero-output black hole. That result seems to me to be a major conceptual breakthrough, showing that the strict rationality assumptions of neoclassical theory can lead to aoutcome that is totally at odds with the usual presumption that the standard neoclassical assumptions inevitably generate a unique stable equilibrium and render macroeconomics superfluous.

Susan Woodward Remembers Armen Alchian

Susan Woodward, a former colleague and co-author of the late great Armen Alchian, was kind enough to share with me an article of hers forthcoming in a special issue of the Journal of Corporate Finance dedicated to Alchian’s memory. I thank Susan and Harold Mulherin, co-editor of the Journal of Corporate Finance for allowing me to post this wonderful tribute to Alchian.

Memories of Armen

Susan Woodward

Sand Hill Econometrics

Armen Alchian approached economics with constructive eccentricity. An aspect became apparent long ago when I taught intermediate price theory, a two-quarter course. Jack Hirshleifer’s new text (Hirshleifer (1976)) was just out and his approach was the foundation of my own training, so that was an obvious choice. But also, Alchian and Allen’s University Economics (Alchian and Allen (1964)) had been usefully separated into parts, of which Exchange and Production: Competition, Coordination, and Control (Alchian and Allen (1977)), the “price theory” part, available in paperback. I used both books.

Somewhere in the second quarter we got to the topic of rent. Rent is such a difficult topic because it’s a word in everyone’s vocabulary but to which economists give a special, second meaning. To prepare a discussion, I looked up “rent” in the index of both texts. In Hirshleifer (1976), it appeared for the first time on some page like 417. In Alchian & Allen (1977), it appeared, say, on page 99, and page 102, and page 188, and pages 87-88, 336-338, and 364-365. It was peppered all through the book.

Hirshleifer approached price theory as geometry. Lay out the axioms, prove the theorems. And never introduce a new idea, especially one like “rent” that collides with standard usage, without a solid foundation. The Alchian approach is more exploratory. “Oh, here’s an idea. Let’s walk around the idea and see what it looks like from all sides. Let’s tip it over and see what’s under it and what kind of noise it makes. Let’s light a fire under it and just see what happens. Drop it ten stories.” The books were complements, not substitutes.

While this textbook story illustrates one aspect of Armen’s thinking, the big epiphanies came working on our joint papers. Unusual for students at UCLA in that era, I didn’t have Armen as a teacher. My first year, Armen was away, and Jack Hirshleifer taught the entire first year price theory. Entranced by the finance segment of that year, the lure of finance in business school was irresistible. But fortune did not abandon me.

I came back to UCLA to teach at the dawn of personal computers. Oh they were feeble! There was a little room on the eighth floor of Bunche Hall where there were three little Compaq computers—the ones with really tiny green-on-black screens. Portable, sort of, but not like a purse. Armen and I were regulars in this word processing cave. Armen would get bored and start a conversation by asking some profound question. I’d flounder a bit and tell him I didn’t know and go back to work. But one day he asked why corporations limit liability. Whew, something to say. It is not a risk story, but about facilitating transferable shares. Limit liability, then shareholders and contracting creditors can price possible recovery, and the wealth and resources of individual shareholder are then irrelevant. When liability tries to reach beyond the firm’s assets to those of individual shareholders, shareholder wealth matters to value, and this creates reasons for inhibiting share transfers. You can limit liability and still address concern about tort creditors by having the firm carrying insurance for torts.

Armen asked “How did you figure this out?” I said, “I don’t know.” “Have you written it down?” “No, it doesn’t seem important enough, it would only be two pages.” “Oh, no, of course it is!” He was right. What I wrote at Armen’s insistence, Woodward (1985), is now in two books of readings on the modern corporation, still in print, still on reading lists, and yes it was more than two pages. The paper by Bargeron and Lehn (2015) in this volume provides empirical confirmation about the impact of limited liability on share transferability. After our conversations about limited liability, Armen never again called me “Joanne,” as in the actress, Joanne Woodward, wife of Paul Newman.

This led to many more discussions about the organization of firms. I was dismayed by the seeming mysticism of “teamwork” as discussed in the old Alchian & Demsetz paper. Does it not all boil down to moral hazard and hold-up, both aspects of information costs, and the potential for the residual claimant to manage these? Armen came to agree and that this, too, was worth writing up. So we started writing. I scribbled down my thoughts. Armen read them and said, “Well, this is right, but it will make Harold (Demsetz) mad. We can’t say it that way. We’ll say it another way.” Armen saw it as his job to bring Harold around.

As we started working on this paper (Alchian and Woodward (1987)), I asked Armen, “What journal should we be thinking of?” Armen said “Oh, don’t worry about that, something will come along”. It went to Rolf Richter’s journal because Armen admired Rolf’s efforts to promote economic analysis of institutions. There are accounts of Armen pulling accepted papers from journals in order to put them into books of readings in honor of his friends, and these stories are true. No journal impressed Armen very much. He thought that if something was good, people would find it and read it.

Soon after the first paper was circulating, Orley Ashenfelter asked Armen to write a book review of Oliver Williamson’s The Economic Institutions of Capitalism (such a brilliant title!). I got enlisted for that project too (Alchian and Woodward (1988)). Armen began writing, but I went back to reread Institutions of Capitalism. Armen gave me what he had written, and I was baffled. “Armen, this stuff isn’t in Williamson.” He asked, “Well, did he get it wrong?” I said, “No, it’s not that he got it wrong. These issues just aren’t there at all. You attribute these ideas to him, but they really come from our other paper.” And he said “Oh, well, don’t worry about that. Some historian will sort it out later. It’s a good place to promote these ideas, and they’ll get the right story eventually.” So, dear reader, now you know.

This from someone who spent his life discussing the efficiencies of private property and property rights—to basically give ideas away in order to promote them? It was a good lesson. I was just starting my ten years in the federal government. In academia, thinkers try to establish property rights in ideas. “This is mine. I thought of this. You must cite me.” In government this is not a winning strategy. Instead, you need plant an idea, then convince others that it’s their idea so they will help you.

And it was sometimes Armen’s strategy in the academic world too. Only someone who was very confident would do this. Or someone who just cared more about promoting ideas he thought were right than he cared about getting credit for them. Or someone who did not have so much respect for the refereeing process. He was so cavalier!

Armen had no use for formal models that did not teach us to look somewhere new in the known world, nor had he any patience for findings that relied on fancy econometrics. What was Armen’s idea of econometrics? Merton Miller told me. We were chatting about limited liability. Merton asked about evidence. Well, all public firms with transferable shares now have limited liability. But in private, closely-held firms, loans nearly always explicitly specify which of the owner’s personal assets are pledged against bank loans. “How do you know?” “From conversations with bankers.” Merton said said, “Ah, this sounds like UCLA econometrics! You go to Armen Alchian and you ask, ‘Armen, is this number about right?’ And Armen says, ‘Yeah, that sounds right.’ So you use that number.”

Why is Armen loved so much? It’s not just his contributions to our understanding because many great thinkers are hardly loved at all. Several things stand out. As noted above, Armen’s sense of what is important is very appealing. Ideas are important. Ideas are more important than being important. Don’t fuss over the small stuff or the small-minded stuff, just work on the ideas and get them right. Armen worked at inhibiting inefficient behavior, but never in an obvious way. He would be the first to agree that not all competition is efficient, and in particular that status competition is inefficient. Lunches and dinners with Armen never included conversations about who was getting tenure where or why various papers got in or did not get in to certain journals. He thought it just did not matter very much or deserve much attention.

Armen was intensely curious about the world and interested in things outside of himself. He was one of the least self-indulgent people that I have ever met. It cheered everybody up. Everyone was in a better mood for the often silly questions that Armen would ask about everything, such as, “Why do they use decorations in the sushi bar and not anywhere else? Is there some optimality story here?” Armen recognized his own limitations and was not afraid of them.

Armen’s views on inefficient behavior came out in an interesting way when we were working on the Williamson book review. What does the word “fair” mean? In the early 1970’s at UCLA, no one was very comfortable with “fair”. Many would even have said ‘fair’ has no meaning in economics. But then we got to pondering the car repair person in the desert (remember Los Angeles is next to a big desert), who is in a position to hold up unlucky motorists whose vehicles break down in a remote place. Why would the mechanic not hold up the motorist and charge a high price? The mechanic has the power. Information about occasional holdups would provoke inefficient avoidance of travel or taking ridiculous precautions. But from the individual perspective, why wouldn’t the mechanic engage in opportunistic behavior, on the spot? “Well,” Armen said, “probably he doesn’t do it because he was raised right.” Armen knew what “fair” meant, and was willing to take a stand on it being efficient.

For all his reputation as a conservative, Armen was very interested in Earl Thompson’s ideas about socially efficient institutions, and the useful constraints that collective action could and does impose on us. (see, for example, Thompson (1968, 1974).) He had more patience for Earl that any of Earl’s other senior colleagues except possibly Jim Buchanan. Earl could go on all evening and longer about the welfare cost of the status rat race, of militarism and how to discourage it, the brilliance of agricultural subsidies, how no one should listen to corrupt elites, and Armen would smile and nod and ponder.

Armen was a happy teacher. As others attest in this issue, he brought great energy, engagement, and generosity to the classroom. He might have been dressed for golf, but he gave students his complete attention. He especially enjoyed teaching the judges in Henry Manne’s Economics & Law program. One former pupil sought him out and at dinner, brought up the Apple v. Microsoft copyright dispute. He wanted to discuss the merits of the issues. Armen said oh no, simply get the thing over with ASAP. Armen said that he was a shareholder in both companies, and consequently did not care who won, but cared very much about what resources were squandered on the battle. Though the economics of this perspective was not novel (it was aired in Texaco v Pennzoil few years earlier), Armen provided in that conversation a view that neither side had an interest in promoting in court. The reaction was: Oh! Those who followed this case might have been puzzled at the subsequent proceeding in this dispute, but those who heard the conversation at dinner were not.

And Armen was a warm and sentimental person. When I moved to Washington, I left my roller skates in the extra bedroom where I slept when I visited Armen and Pauline. These were old-fashioned skates with two wheels in the front and two in the back, Riedell boots and kryptonite wheels, bought at Rip City Skates on Santa Monica Boulevard, (which is still there in 2015! I just looked it up, the proprietor knows all the empty swimming pools within 75 miles). I would take my skates down to the beach and skate from Santa Monica to Venice and back, then go buy some cinnamon rolls at the Pioneer bakery, and bring them back to Mar Vista and Armen and Pauline and I would eat them. Armen loved this ritual. Is she back yet? When I married Bob Hall and moved back to California, Armen did not want me to take the skates away. So I didn’t.

And here is a story Armen loved: Ron Batchelder was a student at UCLA who is also a great tennis player, a professional tennis player who had to be lured out of tennis and into economics, and who has written some fine economic history and more. He played tennis with Armen regularly for many years. On one occasion before dinner Armen said to Ron, “I played really well today.” Ron said, “Yes, you did; you played quite well today.” And Armen said, “But you know what? When I play better, you play better.” And Ron smiled and shrugged his shoulders. I said, “Ron, is it true?” He shrugged again and said, “Well, a long time ago, I learned to play the customer’s game.” And of course Armen just loved that line. He re-told that story many times.

Armen’s enthusiasm for that story is a reflection of his enthusiasm for life. It was a rare enthusiasm, an extraordinary enthusiasm. We all give him credit for it and we should, because it was an act of choice; it was an act of will, a gift to us all. Armen would have never said so, though, because he was raised right.


Alchian, Armen A., William R. Allen, 1964. University Economics. Wadsworth Publishing Company, Belmont, CA.

Alchian, Armen A., William R. Allen, 1977 Exchange and Production: Competition, Coordination, and Control. Wadsworth Publishing Company, Belmont, CA., 2nd edition.

Alchian, Armen A., Woodward, Susan, 1987. “Reflections on the theory of the firm.” Journal of Institutional and Theoretical Economics. 143, 110-136.

Alchian, Armen A., Woodward, Susan, 1988. “The firm is dead: Long live the firm: A review of Oliver E. Williamson’s The Economic Institutions of Capitalism.” Journal of Economic Literature. 26, 65-79.

Bargeron, Leonce, Lehn, Kenneth, 2015. “Limited liability and share transferability: An analysis of California firms, 1920-1940.” Journal of Corporate Finance, this volume.

Hirshleifer, Jack, 1976. Price Theory and Applications. Prentice hall, Englewood Cliffs, NJ.

Thompson, Earl A., 1968. “The perfectly competitive production of public goods.” Review of Economics and Statistics. 50, 1-12.

Thompson, Earl A., 1974. “Taxation and national defense.” Journal of Political Economy. 82, 755-782.

Woodward, Susan E., 1985. “Limited liability in the theory of the firm.” Journal of Institutional and Theorectical Economics. 141, 601-611.

Trying to Make Sense of the Insane Policy of the Bank of France and Other Catastrophes

In the almost four years since I started blogging I have occasionally referred to the insane Bank of France or to the insane policy of the Bank of France, a mental disorder that helped cause the deflation that produced the Great Depression. The insane policy began in 1928 when the Bank of France began converting its rapidly growing stockpile of foreign-exchange reserves (i.e., dollar- or sterling-denominated financial instruments) into gold. The conversion of foreign exchange was precipitated by the enactment of a law restoring the legal convertibility of the franc into gold and requiring the Bank of France to hold gold reserves equal to at least 35% of its outstanding banknotes. The law induced a massive inflow of gold into the Bank of France, and, after the Federal Reserve recklessly tightened its policy in an attempt to stamp out stock speculation on Wall Street, thereby inducing an inflow of gold into the US, the one-two punch knocked the world economy into just the deflationary tailspin that Hawtrey and Cassel, had warned would result if the postwar restoration of the gold standard were not managed so as to minimize the increase in the monetary demand for gold.

In making a new round of revisions to our paper on Hawtrey and Cassel, my co-author Ron Batchelder has just added an interesting footnote pointing out that there may have been a sensible rationale for the French gold policy: to accumulate a sufficient hoard of gold for use in case of another war with Germany. In World War I, belligerents withdrew gold coins from circulation, melted them down, and, over the next few years, exported the gold to neutral countries to pay for food and war supplies. That’s how the US, remaining neutral till 1917, wound up with a staggering 40% of the world’s stock of monetary gold reserves after the war. Obsessed with the military threat a re-armed Germany would pose, France insisted that the Versailles Treaty impose crippling reparations payments. The 1926 stabilization of the franc and enactment of the law restoring the gold standard and imposing a 35% reserve requirement on banknotes issued by the Bank of France occurred during the premiership of the staunchly anti-German Raymond Poincaré, a native of Lorraine (lost to Prussia in the war of 1870-71) and President of France during World War I.

A long time ago I wrote a paper “An Evolutionary Theory of the State Monopoly over Money” (which was reworked as chapter two of my book Free Banking and Monetary Reform and was later published in Money and the Nation State) in which, relying on an argument made by Earl Thompson, I suggested that historically the main reason for the nearly ubiquitous state involvement in supplying money was military not monetary: monopoly control over the supply of money enables the sovereign to quickly gain control over resources in war time, thereby giving states in which the sovereign controls the supply of money a military advantage over states in which the sovereign has no such control. Subsequently, Thompson further developed the idea to explain the rise of the gold standard after the Bank of England was founded in 1694, early in the reign of William and Mary, to finance rebuilding of the English navy, largely destroyed by the French in 1690. As explained by Macaulay in his History of England, the Bank of England, by substantially reducing the borrowing costs of the British government, was critical to the survival of the new monarchs in their battle with the Stuarts and Louis XIV. See Thompson’s article “The Gold Standard: Causes and Consequences” in Business Cycles and Depressions: An Encyclopedia (edited by me).

Thompson’s article is not focused on the holding of gold reserves, but on the confidence that the gold standard gave to those lending to the state, especially during a wartime suspension of convertibility, owing to an implicit commitment to restore the gold standard at the prewar parity. The importance of that implicit commitment is one reason why Churchill’s 1925 decision to restore the gold standard at the prewar parity was not necessarily as foolish as Keynes (The Economic Consequences of Mr. Churchill), along with almost all subsequent commentators, judged it to have been. But the postwar depreciation of the franc was so extreme that restoring the convertibility of the franc at the prewar parity became a practical impossibility, and the new parity at which convertibility was restored was just a fifth of its prewar level. Having thus reneged on its implicit commitment to restore the gold standard at the prewar parity, impairing its ability to borrow, France may have felt it had no alternative but to accumulate a ready gold reserve from which to draw when another war against Germany came. This is just theoretical speculation, but it might provide some clues for historical research into the thinking of French politicians and bankers in the late 1920s as they formulated their strategy for rejoining the gold standard.

However, even if the motivation for France’s gold accumulation was not simply a miserly desire to hold ever larger piles of shiny gold ingots in the vaults of the Banque de France, but was a precautionary measure against the possibility of a future war with Germany – and we know only too well that the fear was not imaginary – it is important to understand that, in the end, it was almost certainly the French policy of gold accumulation that paved the way for Hitler’s rise to power and all that entailed. Without the Great Depression and the collapse of the German economy, Hitler might well have remained an outcast on the margins of German politics.

The existence of a legitimate motivation for the insane policy of the Bank of France cannot excuse the failure to foresee the all too predictable consequences of that policy – consequences laid out plainly by Hawtrey and Cassel already in 1919-20, and reiterated consistently over the ensuing decade. Nor does the approval of that policy by reputable, even eminent, economists, who simply failed to understand how the gold standard worked, absolve those who made the wrong decisions of responsibility for their mistaken decisions. They were warned about the consequences of their actions, and chose to disregard the warnings.

All of this is sadly reminiscent of the 2003 invasion of Iraq. I don’t agree with those who ascribe evil motives to the Bush administration for invading Iraq, though there seems little doubt that the WMD issue was largely pretextual. But that doesn’t mean that Bush et al. didn’t actually believe that Saddam had WMD. More importantly, I think that Bush et al. sincerely thought that invading Iraq and deposing Saddam Hussein would, after the supposed defeat of Al Qaeda and the Taliban in Afghanistan, establish a benign American dominance in the region, as World War II had done in Japan and Western Europe.

The problem is not, as critics like to say, that Bush et al. lied us into war; the problem is that they stupidly fooled themselves into thinking that they could just invade Iraq, unseat Saddam Hussein, and that their job would be over. They fooled themselves even though they had been warned in advance that Iraq was riven by internal ethnic, sectarian, religious and political divisions. Brutally suppressed by Hussein and his Ba’athist regime, those differences were bound to reemerge once the regime was dismantled. When General Eric Shinseki’s testified before Congress that hundreds of thousands of American troops would be needed to maintain peace and order after Hussein was ousted, Paul Wolfowitz and Donald Rumsfeld could only respond with triumphalist ridicule at the idea that more troops would be required to maintain law and order in Iraq after Hussein was deposed than were needed to depose him. The sophomoric shallowness of the response to Shinseki by those that planned the invasion still shocks and appalls.

It’s true that, after the Republican loss in the 2006 Congressional elections, Bush, freeing himself from the influence of Dick Cheney and replacing Donald Rumsfeld with Robert Gates as Secretary of Defense, and Gen. George Casey with Gen. David Petraeus as commander of US forces in Iraq, finally adopted the counter-insurgency strategy (aka the “surge”) so long resisted by Cheney and Rumsfeld, thereby succeeding in putting down the Sunni/Al-Qaeda/Baathist insurgency and in bringing the anti-American Shi’ite militias to heel. I wrote about the success of the surge in December 2007 when that provisional military success was still controversial. But, as General Petraeus conceded, the ultimate success of the counter-insurgency strategy depended on implementing a political strategy to reconcile the different elements of Iraqi society to their government. We now know that even in 2008 Premier Nouri al-Maliki, who had been installed as premier with the backing of the Bush administration, was already reversing the limited steps taken during the surge to achieve accommodation between Iraqi Sunnis and Shi’ites, while consolidating his Shi’ite base by reconciling politically with the pro-Iranian militants he had put down militarily.

The failure of the Iraqi government to consolidate and maintain the gains made in 2007-08 has been blamed on Obama’s decision to withdraw all American forces from Iraq after the status of forces agreement signed by President Bush and Premier al-Maliki in December 2008 expired at the end of 2011. But preserving the gains made in 2007-08 depended on a political strategy to reconcile the opposing ethnic and sectarian factions of Iraqi society. The Bush administration could not implement such a strategy with 130,000 troops still in Iraq at the end of 2008, and the sovereign Iraqi government in place, left to its own devices, had no interest in pursuing such a strategy. Perhaps keeping a larger US presence in Iraq for a longer time would have kept Iraq from falling apart as fast as it has, but the necessary conditions for a successful political outcome were never in place.

So even if the motivation for the catastrophic accumulation of gold by France in the 1928-29 was merely to prepare itself to fight, if need be, another war against Germany, the fact remains that the main accomplishment of the gold-accumulation policy was to bring to power a German regime far more dangerous and threatening than the one that would have otherwise confronted France. And even if the motivation for the catastrophic invasion of Iraq in 2003 was to defeat and discredit Islamic terrorism, the fact remains that the invasion, just as Osama bin Laden had hoped, was to create the conditions in which Islamic terrorism could grow into a worldwide movement, attracting would-be jihadists to a growing number of local conflicts across the world. Although bin Laden was eventually killed in his Pakistani hideout, the invasion of Iraq led to rise of an even more sophisticated, more dangerous, and more threatening opponent than the one the invasion was intended to eradicate. Just as a misunderstanding of the gold standard led to catastrophe in 1928-29, the misconception that the threat of terrorism can be eliminated by military means has been leading us toward catastrophe since 2003. When will we learn?

PS Despite some overlap between what I say above and what David Henderson said in this post, I am not a libertarian or a non-interventionist.

What Is Free Banking All About?

I notice that there has been a bit of a dustup lately about free banking, triggered by two posts by Izabella Kaminska, first on FTAlphaville followed by another on her own blog. I don’t want to get too deeply into the specifics of Kaminska’s posts, save to correct a couple of factual misstatements and conceptual misunderstandings (see below). At any rate, George Selgin has a detailed reply to Kaminska’s errors with which I mostly agree, and Scott Sumner has scolded her for not distinguishing between sensible free bankers, e.g., Larry White, George Selgin, Kevin Dowd, and Bill Woolsey, and the anti-Fed, gold-bug nutcases who, following in the footsteps of Ron Paul, have adopted free banking as a slogan with which to pursue their anti-Fed crusade.

Now it just so happens that, as some readers may know, I wrote a book about free banking, which I began writing almost 30 years ago. The point of the book was not to call for a revolutionary change in our monetary system, but to show that financial innovations and market forces were causing our modern monetary system to evolve into something like the theoretical model of a free banking system that had been worked out in a general sort of way by some classical monetary theorists, starting with Adam Smith, who believed that a system of private banks operating under a gold standard would supply as much money as, but no more money than, the public wanted to hold. In other words, the quantity of money produced by a system of competing banks, operating under convertibility, could be left to take care of itself, with no centralized quantitative control over either the quantity of bank liabilities or the amount of reserves held by the banking system.

So I especially liked the following comment by J. V. Dubois to Scott’s post

[M]y thing against free banking is that we actually already have it. We already have private banks issuing their own monies directly used for transactions – they are called bank accounts and debit/credit cards. There are countries like Sweden where there are now shops that do not accept physical cash (only bank monies) – a policy actively promoted government, if you can believe it.

There are now even financial products like Xapo Debit Card that automatically converts all payments received on your account into non-monetary assets (with Xapo it is bitcoins) and back into monies when you use the card for payment. There is a very healthy international bank money market so no matter what money you personally use, you can travel all around the world and pay comfortably without ever seeing or touching official local government currency.

In opposition to the Smithian school of thought, there was the view of Smith’s close friend David Hume, who famously articulated what became known as the Price-Specie-Flow Mechanism, a mechanism that Smith wisely omitted from his discussion of international monetary adjustment in the Wealth of Nations, despite having relied on PSFM with due acknowledgment of Hume, in his Lectures on Jurisprudence. In contrast to Smith’s belief that there is a market mechanism limiting the competitive issue of convertible bank liabilities (notes and deposits) to the amount demanded by the public, Hume argued that banks were inherently predisposed to overissue their liabilities, the liabilities being issuable at almost no cost, so that private banks, seeking to profit from the divergence between the face value of their liabilities and the cost of issuing them, were veritable engines of inflation.

These two opposing views of banks later morphed into what became known almost 70 years later as the Banking and Currency Schools. Taking the Humean position, the Currency School argued that without quantitative control over the quantity of banknotes issued, the banking system would inevitably issue an excess of banknotes, causing overtrading, speculation, inflation, a drain on the gold reserves of the banking system, culminating in financial crises. To prevent recurring financial crises, the Currency School proposed a legal limit on the total quantity of banknotes beyond which limit, additional banknotes could be only be issued (by the Bank of England) in exchange for an equivalent amount of gold at the legal gold parity. Taking the Smithian position, the Banking School argued that there were market mechanisms by which any excess liabilities created by the banking system would automatically be returned to the banking system — the law of reflux. Thus, as long as convertibility obtained (i.e., the bank notes were exchangeable for gold at the legal gold parity), any overissue would be self-correcting, so that a legal limit on the quantity of banknotes was, at best, superfluous, and, at worst, would itself trigger a financial crisis.

As it turned out, the legal limit on the quantity of banknotes proposed by the Currency School was enacted in the Bank Charter Act of 1844, and, just as the Banking School predicted, led to a financial crisis in 1847, when, as soon as the total quantity of banknotes approached the legal limit, a sudden precautionary demand for banknotes led to a financial panic that was subdued only after the government announced that the Bank of England would incur no legal liability for issuing banknotes beyond the legal limit. Similar financial panics ensued in 1857 and 1866, and they were also subdued by suspending the relevant statutory limits on the quantity of banknotes. There were no further financial crises in Great Britain in the nineteenth century (except possibly for a minicrisis in 1890), because bank deposits increasingly displaced banknotes as the preferred medium of exchange, the quantity of bank deposits being subject to no statutory limit, and because the market anticipated that, in a crisis, the statutory limit on the quantity of banknotes would be suspended, so that a sudden precautionary demand for banknotes never materialized in the first place.

Let me pause here to comment on the factual and conceptual misunderstandings in Kaminska’s first post. Discussing the role of the Bank of England in the British monetary system in the first half of the nineteenth century, she writes:

But with great money-issuance power comes great responsibility, and more specifically the great temptation to abuse that power via the means of imprudent money-printing. This fate befell the BoE — as it does most banks — not helped by the fact that the BoE still had to compete with a whole bunch of private banks who were just as keen as it to issue money to an equally imprudent degree.

And so it was that by the 1840s — and a number of Napoleonic Wars later — a terrible inflation had begun to grip the land.

So Kaminska seems to have fallen for the Humean notion that banks are inherently predisposed to overissue and, without some quantitative restraint on their issue of liabilities, are engines of inflation. But, as the law of reflux teaches us, this is not true, especially when banks, as they inevitably must, make their liabilities convertible on demand into some outside asset whose supply is not under their control. After 1821, the gold standard having been officially restored in England, the outside asset was gold. So what was happening to the British price level after 1821 was determined not by the actions of the banking system (at least to a first approximation), but by the value of gold which was determined internationally. That’s the conceptual misunderstanding that I want to correct.

Now for the factual misunderstanding. The chart below shows the British Retail Price Index between 1825 and 1850. The British price level was clearly falling for most of the period. After falling steadily from 1825 to about 1835, the price level rebounded till 1839, but it prices again started to fall reaching a low point in 1844, before starting another brief rebound and rising sharply in 1847 until the panic when prices again started falling rapidly.


From a historical perspective, the outcome of the implicit Smith-Hume disagreement, which developed into the explicit dispute over the Bank Charter Act of 1844 between the Banking and Currency Schools, was highly unsatisfactory. Not only was the dysfunctional Bank Charter Act enacted, but the orthodox view of how the gold standard operates was defined by the Humean price-specie-flow mechanism and the Humean fallacy that banks are engines of inflation, which made it appear that, for the gold standard to function, the quantity of money had to be tied rigidly to the gold reserve, thereby placing the burden of adjustment primarily on countries losing gold, so that inflationary excesses would be avoided. (Fortunately, for the world economy, gold supplies increased fairly rapidly during the nineteenth century, the spread of the gold standard meant that the monetary demand for gold was increasing faster than the supply of gold, causing gold to appreciate for most of the nineteenth century.)

When I set out to write my book on free banking, my intention was to clear up the historical misunderstandings, largely attributable to David Hume, surrounding the operation of the gold standard and the behavior of competitive banks. In contrast to the Humean view that banks are inherently inflationary — a view endorsed by quantity theorists of all stripes and enshrined in the money-multiplier analysis found in every economics textbook — that the price level would go to infinity if banks were not constrained by a legal reserve requirement on their creation of liabilities, there was an alternative view that the creation of liabilities by the banking system is characterized by the same sort of revenue and cost considerations governing other profit-making enterprises, and that the equilibrium of a private banking system is not that value of money is driven down to zero, as Milton Friedman, for example, claimed in his Program for Monetary Stability.

The modern discovery (or rediscovery) that banks are not inherently disposed to debase their liabilities was made by James Tobin in his classic paper “Commercial Banks and Creators of Money.” Tobin’s analysis was extended by others (notably Ben Klein, Earl Thompson, and Fischer Black) to show that the standard arguments for imposing quantitative limits on the creation of bank liabilities were unfounded, because, even with no legal constraints, there are economic forces limiting their creation of liabilities. A few years after these contributions, F. A. Hayek also figured out that there are competitive forces constraining the creation of liabilities by the banking system. He further developed the idea in a short book Denationalization of Money which did much to raise the profile of the idea of free banking, at least in some circles.

If there is an economic constraint on the creation of bank liabilities, and if, accordingly, the creation of bank liabilities was responsive to the demands of individuals to hold those liabilities, the Friedman/Monetarist idea that the goal of monetary policy should be to manage the total quantity of bank liabilities so that it would grow continuously at a fixed rate was really dumb. It was tried unsuccessfully by Paul Volcker in the early 1980s, in his struggle to bring inflation under control. It failed for precisely the reason that the Bank Charter Act had to be suspended periodically in the nineteenth century: the quantitative limit on the growth of the money supply itself triggered a precautionary demand to hold money that led to a financial crisis. In order to avoid a financial crisis, the Volcker Fed constantly allowed the monetary aggregates to exceed their growth targets, but until Volcker announced in the summer of 1982 that the Fed would stop paying attention to the aggregates, the economy was teetering on the verge of a financial crisis, undergoing the deepest recession since the Great Depression. After the threat of a Friedman/Monetarist financial crisis was lifted, the US economy almost immediately began one of the fastest expansions of the post-war period.

Nevertheless, for years afterwards, Friedman and his fellow Monetarists kept warning that rapid growth of the monetary aggregates meant that the double-digit inflation of the late 1970s and early 1980s would soon return. So one of my aims in my book was to use free-banking theory – the idea that there are economic forces constraining the issue of bank liabilities and that banks are not inherently engines of inflation – to refute the Monetarist notion that the key to economic stability is to make the money stock grow at a constant 3% annual rate of growth.

Another goal was to explain that competitive banks necessarily have to select some outside asset into which to make their liabilities convertible. Otherwise those liabilities would have no value, or at least so I argued, and still believe. The existence of what we now call network effects forces banks to converge on whatever assets are already serving as money in whatever geographic location they are trying to draw customers from. Thus, free banking is entirely consistent with an already existing fiat currency, so that there is no necessary link between free banking and a gold (or other commodity) standard. Moreover, if free banking were adopted without abolishing existing fiat currencies and legal tender laws, there is almost no chance that, as Hayek argued, new privately established monetary units would arise to displace the existing fiat currencies.

My final goal was to suggest a new way of conducting monetary policy that would enhance the stability of a free banking system, proposing a monetary regime that would ensure the optimum behavior of prices over time. When I wrote the book, I had been convinced by Earl Thompson that the optimum behavior of the price level over time would be achieved if an index of nominal wages was stabilized. He proposed accomplishing this objective by way of indirect convertibility of the dollar into an index of nominal wages by way of a modified form of Irving Fisher’s compensated dollar plan. I won’t discuss how or why that goal could be achieved, but I am no longer convinced of the optimality of stabilizing an index of nominal wages. So I am now more inclined toward nominal GDP level targeting as a monetary policy regime than the system I proposed in my book.

But let me come back to the point that I think J. V. Dubois was getting at in his comment. Historically, idea of free banking meant that private banks should be allowed to issue bank notes of their own (with the issuing bank clearly identified) without unreasonable regulations, restrictions or burdens not generally applied to other institutions. During the period when private banknotes were widely circulating, banknotes were a more prevalent form of money than bank deposits. So in the 21st century, the right of banks to issue hand to hand circulating banknotes is hardly a crucial issue for monetary policy. What really matters is the overall legal and regulatory framework under which banks operate.

The term “free banking” does very little to shed light on most of these issues. For example, what kind of functions should banks perform? Should commercial banks also engage in investment banking? Should commercial bank liabilities be ensured by the government, and if so under what terms, and up to what limits? There are just a couple of issues; there are many others. And they aren’t necessarily easily resolved by invoking the free-banking slogan. When I was writing, I meant by “free banking” a system in which the market determined the total quantity of bank liabilities. I am still willing to use “free banking” in that sense, but there are all kinds of issues concerning the asset side of bank balance sheets that also need to be addressed, and I don’t find it helpful to use the term free banking to address those issues.

Aggregate Demand and Coordination Failures

Regular readers of this blog may have noticed that I have been writing less about monetary policy and more about theory and methodology than when I started blogging a little over three years ago. Now one reason for that is that I’ve already said what I want to say about policy, and, since I get bored easily, I look for new things to write about. Another reason is that, at least in the US, the economy seems to have reached a sustainable growth path that seems likely to continue for the near to intermediate term. I think that monetary policy could be doing more to promote recovery, and I wish that it would, but unfortunately, the policy is what it is, and it will continue more or less in the way that Janet Yellen has been saying it will. Falling oil prices, because of increasing US oil output, suggest that growth may speed up slightly even as inflation stays low, possibly even falling to one percent or less. At least in the short-term, the fall in inflation does not seem like a cause for concern. A third reason for writing less about monetary policy is that I have been giving a lot of thought to what it is that I dislike about the current state of macroeconomics, and as I have been thinking about it, I have been writing about it.

In thinking about what I think is wrong with modern macroeconomics, I have been coming back again and again, though usually without explicit attribution, to an idea that was impressed upon me as an undergrad and grad student by Axel Leijonhufvud: that the main concern of macroeconomics ought to be with failures of coordination. A Swede, trained in the tradition of the Wicksellian Stockholm School, Leijonhufvud immersed himself in the study of the economics of Keynes and Keynesian economics, while also mastering the Austrian literature, and becoming an admirer of Hayek, especially Hayek’s seminal 1937 paper, “Economics and Knowledge.”

In discussing Keynes, Leijonhufvud focused on two kinds of coordination failures.

First, there is a problem in the labor market. If there is unemployment because the real wage is too high, an individual worker can’t solve the problem by offering to accept a reduced nominal wage. Suppose the price of output is $1 a unit and the wage is $10 a day, but the real wage consistent with full employment is $9 a day, meaning that producers choose to produce less output than they would produce if the real wage were lower, thus hiring fewer workers than they would if the real wage were lower than it is. If an individual worker offers to accept a wage of $9 a day, but other workers continue to hold out for $10 a day, it’s not clear that an employer would want to hire the worker who offers to work for $9 a day. If employers are not hiring additional workers because they can’t cover the cost of the additional output produced with the incremental revenue generated by the added output, the willingness of one worker to work for $9 a day is not likely to make a difference to the employer’s output and hiring decisions. It is not obvious what sequence of transactions would result in an increase in output and employment when the real wage is above the equilibrium level. There are complex feedback effects from a change, so that the net effect of making those changes in a piecemeal fashion is unpredictable, even though there is a possible full-employment equilibrium with a real wage of $9 a day. If the problem is that real wages in general are too high for full employment, the willingness of an individual worker to accept a reduced wage from a single employer does not fix the problem.

In the standard competitive model, there is a perfect market for every commodity in which every transactor is assumed to be able to buy and sell as much as he wants. But the standard competitive model has very little to say about the process by which those equilibrium prices are arrived at. And a typical worker is never faced with that kind of choice posited in the competitive model: an impersonal uniform wage at which he can decide how many hours a day or week or year he wants to work at that uniform wage. Under those circumstances, Keynes argued that the willingness of some workers to accept wage cuts in order to gain employment would not significantly increase employment, and might actually have destabilizing side-effects. Keynes tried to make this argument in the framework of an equilibrium model, though the nature of the argument, as Don Patinkin among others observed, was really better suited to a disequilibrium framework. Unfortunately, Keynes’s argument was subsequently dumbed down to a simple assertion that wages and prices are sticky (especially downward).

Second, there is an intertemporal problem, because the interest rate may be stuck at a rate too high to allow enough current investment to generate the full-employment level of spending given the current level of the money wage. In this scenario, unemployment isn’t caused by a real wage that is too high, so trying to fix it by wage adjustment would be a mistake. Since the source of the problem is the rate of interest, the way to fix the problem would be to reduce the rate of interest. But depending on the circumstances, there may be a coordination failure: bear speculators, expecting the rate of interest to rise when it falls to abnormally low levels, prevent the rate of interest from falling enough to induce enough investment to support full employment. Keynes put too much weight on bear speculators as the source of the intertemporal problem; Hawtrey’s notion of a credit deadlock would actually have been a better way to go, and nowadays, when people speak about a Keynesian liquidity trap, what they really have in mind is something closer to Hawtreyan credit deadlock than to the Keynesian liquidity trap.

Keynes surely deserves credit for identifying and explaining two possible sources of coordination failures, failures affecting the macroeconomy, because interest rates and wages, though they actually come in many different shapes and sizes, affect all markets and are true macroeconomic variables. But Keynes’s analysis of those coordination failures was far from being fully satisfactory, which is not surprising; a theoretical pioneer rarely provides a fully satisfactory analysis, leaving lots of work for successors.

But I think that Keynes’s theoretical paradigm actually did lead macroeconomics in the wrong direction, in the direction of a highly aggregated model with a single output, a bond, a medium of exchange, and a labor market, with no explicit characterization of the production technology. (I.e., is there one factor or two, and if two how is the price of the second factor determined? See, here, here, here, and here my discussion of Earl Thompson’s “A Reformulation of Macroeconomic Theory,” which I hope at some point to revisit and continue.)

Why was it the wrong direction? Because, the Keynesian model (both Keynes’s own version and the Hicksian IS-LM version of his model) ruled out the sort of coordination problems that might arise in a multi-product, multi-factor, intertemporal model in which total output depends in a meaningful way on the meshing of the interdependent plans, independently formulated by decentralized decision-makers, contingent on possibly inconsistent expectations of the future. In the over-simplified and over-aggregated Keynesian model, the essence of the coordination problem has been assumed away, leaving only a residue of the actual problem to be addressed by the model. The focus of the model is on aggregate expenditure, income, and output flows, with no attention paid to the truly daunting task of achieving sufficient coordination among the independent decision makers to allow total output and income to closely approximate the maximum sustainable output and income that the system could generate in a perfectly coordinated state, aka full intertemporal equilibrium.

This way of thinking about macroeconomics led to the merging of macroeconomics with neoclassical growth theory and to the routine and unthinking incorporation of aggregate production functions in macroeconomic models, a practice that is strictly justified only in a single-output, two-factor model in which the value of capital is independent of the rate of interest, so that the havoc-producing effects of reswitching and capital-reversal can be avoided. Eventually, these models were taken over by modern real-business-cycle theorists, who dogmatically rule out any consideration of coordination problems, while attributing all observed output and employment fluctuations to random productivity shocks. If one thinks of macroeconomics as an attempt to understand coordination failures, the RBC explanation of output and employment fluctuations is totally backwards; productivity fluctuations, like fluctuations in output and employment, are the not the results of unexplained random disturbances, they are the symptoms of coordination failures. That’s it, eureka! Solve the problem by assuming that it does not exist.

If you are thinking that this seems like an Austrian critique of the Keynesian model or the Keynesian approach, you are right; it is an Austrian critique. But it has nothing to do with stereotypical Austrian policy negativism; it is a critique of the oversimplified structure of the Keynesian model, which foreshadowed the reduction ad absurdum or modern real-business-cycle theory, which has nearly banished the idea of coordination failures from modern macroeconomics. The critique is not about the lack of a roundabout capital structure; it is about the narrow scope for inconsistencies in production and consumption plans.

I think that Leijonhufvud almost 40 years ago was getting at this point when he wrote the following paragraph near toward end of his book on Keynes.

The unclear mix of statics and dynamics [in the General Theory] would seem to be main reason for later muddles. One cannot assume that what went wrong was simply that Keynes slipped up here and there in his adaptation of standard tools, and that consequently, if we go back and tinker a little more with the Marshallian toolbox his purposes will be realized. What is required, I believe, is a systematic investigation from the standpoint of the information problems stressed in this study, of what elements of the static theory of resource allocation can without further ado be utilized in the analysis of dynamic and historical systems. This, of course, would be merely a first step: the gap yawns very wide between the systematic and rigorous modern analysis of the stability of simple, “featureless,” pure exchange systems and Keynes’ inspired sketch of the income-constrained process in a monetary exchange-cum production system. But even for such a first step, the prescription cannot be to “go back to Keynes.” If one must retrace some step of past developments in order to get on the right track – and that is probably advisable – my own preference is to go back to Hayek. Hayek’s Gestalt-conception of what happens during business cycles, it has been generally agreed, was much less sound that Keynes’. As an unhappy consequence, his far superior work on the fundamentals of the problem has not received the attention it deserves. (pp. 401-02)

I don’t think that we actually need to go back to Hayek, though “Economics and Knowledge” should certainly be read by every macroeconomist, but we do need to get a clearer understanding of the potential for breakdowns in economic activity to be caused by inconsistent expectations, especially when expectations are themselves mutually dependent and reinforcing. Because expectations are mutually interdependent, they are highly susceptible to network effects. Network effects produce tipping points, tipping points can lead to catastrophic outcomes. Just wanted to share that with you. Have a nice day.

Monetary Theory on the Neo-Fisherite Edge

The week before last, Noah Smith wrote a post “The Neo-Fisherite Rebellion” discussing, rather sympathetically I thought, the contrarian school of monetary thought emerging from the Great American Heartland, according to which, notwithstanding everything monetary economists since Henry Thornton have taught, high interest rates are inflationary and low interest rates deflationary. This view of the relationship between interest rates and inflation was advanced (but later retracted) by Narayana Kocherlakota, President of the Minneapolis Fed in a 2010 lecture, and was embraced and expounded with increased steadfastness by Stephen Williamson of Washington University in St. Louis and the St. Louis Fed in at least one working paper and in a series of posts over the past five or six months (e.g. here, here and here). And John Cochrane of the University of Chicago has picked up on the idea as well in two recent blog posts (here and here). Others seem to be joining the upstart school as well.

The new argument seems simple: given the Fisher equation, in which the nominal interest rate equals the real interest rate plus the (expected) rate of inflation, a central bank can meet its inflation target by setting a fixed nominal interest rate target consistent with its inflation target and keeping it there. Once the central bank sets its target, the long-run neutrality of money, implying that the real interest rate is independent of the nominal targets set by the central bank, ensures that inflation expectations must converge on rates consistent with the nominal interest rate target and the independently determined real interest rate (i.e., the real yield curve), so that the actual and expected rates of inflation adjust to ensure that the Fisher equation is satisfied. If the promise of the central bank to maintain a particular nominal rate over time is believed, the promise will induce a rate of inflation consistent with the nominal interest-rate target and the exogenous real rate.

The novelty of this way of thinking about monetary policy is that monetary theorists have generally assumed that the actual adjustment of the price level or inflation rate depends on whether the target interest rate is greater or less than the real rate plus the expected rate. When the target rate is greater than the real rate plus expected inflation, inflation goes down, and when it is less than the real rate plus expected inflation, inflation goes up. In the conventional treatment, the expected rate of inflation is momentarily fixed, and the (expected) real rate variable. In the Neo-Fisherite school, the (expected) real rate is fixed, and the expected inflation rate is variable. (Just as an aside, I would observe that the idea that expectations about the real rate of interest and the inflation rate cannot occur simultaneously in the short run is not derived from the limited cognitive capacity of economic agents; it can only be derived from the limited intellectual capacity of economic theorists.)

The heretical views expressed by Williamson and Cochrane and earlier by Kocherlakota have understandably elicited scorn and derision from conventional monetary theorists, whether Keynesian, New Keynesian, Monetarist or Market Monetarist. (Williamson having appropriated for himself the New Monetarist label, I regrettably could not preserve an appropriate symmetry in my list of labels for monetary theorists.) As a matter of fact, I wrote a post last December challenging Williamson’s reasoning in arguing that QE had caused a decline in inflation, though in his initial foray into uncharted territory, Williamson was actually making a narrower argument than the more general thesis that he has more recently expounded.

Although deep down, I have no great sympathy for Williamson’s argument, the counterarguments I have seen leave me feeling a bit, shall we say, underwhelmed. That’s not to say that I am becoming a convert to New Monetarism, but I am feeling that we have reached a point at which certain underlying gaps in monetary theory can’t be concealed any longer. To explain what I mean by that remark, let me start by reviewing the historical context in which the ruling doctrine governing central-bank operations via adjustments in the central-bank lending rate evolved. The primary (though historically not the first) source of the doctrine is Henry Thornton in his classic volume The Nature and Effects of the Paper Credit of Great Britain.

Even though Thornton focused on the policy of the Bank of England during the Napoleonic Wars, when Bank of England notes, not gold, were legal tender, his discussion was still in the context of a monetary system in which paper money was generally convertible into either gold or silver. Inconvertible banknotes – aka fiat money — were the exception not the rule. Gold and silver were what Nick Rowe would call alpha money. All other moneys were evaluated in terms of gold and silver, not in terms of a general price level (not yet a widely accepted concept). Even though Bank of England notes became an alternative alpha money during the restriction period of inconvertibility, that situation was generally viewed as temporary, the restoration of convertibility being expected after the war. The value of the paper pound was tracked by the sterling price of gold on the Hamburg exchange. Thus, Ricardo’s first published work was entitled The High Price of Bullion, in which he blamed the high sterling price of bullion at Hamburg on an overissue of banknotes by the Bank of England.

But to get back to Thornton, who was far more concerned with the mechanics of monetary policy than Ricardo, his great contribution was to show that the Bank of England could control the amount of lending (and money creation) by adjusting the interest rate charged to borrowers. If banknotes were depreciating relative to gold, the Bank of England could increase the value of their notes by raising the rate of interest charged on loans.

The point is that if you are a central banker and are trying to target the exchange rate of your currency with respect to an alpha currency, you can do so by adjusting the interest rate that you charge borrowers. Raising the interest rate will cause the exchange value of your currency to rise and reducing the interest rate will cause the exchange value to fall. And if you are operating under strict convertibility, so that you are committed to keep the exchange rate between your currency and an alpha currency at a specified par value, raising that interest rate will cause you to accumulate reserves payable in terms of the alpha currency, and reducing that interest rate will cause you to emit reserves payable in terms of the alpha currency.

So the idea that an increase in the central-bank interest rate tends to increase the exchange value of its currency, or, under a fixed-exchange rate regime, an increase in the foreign exchange reserves of the bank, has a history at least two centuries old, though the doctrine has not exactly been free of misunderstanding or confusion in the course of those two centuries. One of those misunderstandings was about the effect of a change in the central-bank interest rate, under a fixed-exchange rate regime. In fact, as long as the central bank is maintaining a fixed exchange rate between its currency and an alpha currency, changes in the central-bank interest rate don’t affect (at least as a first approximation) either the domestic money supply or the domestic price level; all that changes in the central-bank interest rate can accomplish is to change the bank’s holdings of alpha-currency reserves.

It seems to me that this long well-documented historical association between changes in the central-bank interest rates and the exchange value of currencies and the level of private spending is the basis for the widespread theoretical presumption that raising the central-bank interest rate target is deflationary and reducing it is inflationary. However, the old central-bank doctrine of the Bank Rate was conceived in a world in which gold and silver were the alpha moneys, and central banks – even central banks operating with inconvertible currencies – were beta banks, because the value of a central-bank currency was still reckoned, like the value of inconvertible Bank of England notes in the Napoleonic Wars, in terms of gold and silver.

In the Neo-Fisherite world, central banks rarely peg exchange rates against each other, and there is no longer any outside standard of value to which central banks even nominally commit themselves. In a world without the metallic standard of value in which the conventional theory of central banking developed, do the propositions about the effects of central-bank interest-rate setting still obtain? I am not so sure that they do, not with the analytical tools that we normally deploy when thinking about the effects of central-bank policies. Why not? Because, in a Neo-Fisherite world in which all central banks are alpha banks, I am not so sure that we really know what determines the value of this thing called fiat money. And if we don’t really know what determines the value of a fiat money, how can we really be sure that interest-rate policy works the same way in a Neo-Fisherite world that it used to work when the value of money was determined in relation to a metallic standard? (Just to avoid misunderstanding, I am not – repeat NOT — arguing for restoring the gold standard.)

Why do I say that we don’t know what determines the value of fiat money in a Neo-Fisherite world? Well, consider this. Almost three weeks ago I wrote a post in which I suggested that Bitcoins could be a massive bubble. My explanation for why Bitcoins could be a bubble is that they provide no real (i.e., non-monetary) service, so that their value is totally contingent on, and derived from (or so it seems to me, though I admit that my understanding of Bitcoins is partial and imperfect), the expectation of a positive future resale value. However, it seems certain that the resale value of Bitcoins must eventually fall to zero, so that backward induction implies that Bitcoins, inasmuch as they provide no real service, cannot retain a positive value in the present. On this reasoning, any observed value of a Bitcoin seems inexplicable except as an irrational bubble phenomenon.

Most of the comments I received about that post challenged the relevance of the backward-induction argument. The challenges were mainly of two types: a) the end state, when everyone will certainly stop accepting a Bitcoin in exchange, is very, very far into the future and its date is unknown, and b) the backward-induction argument applies equally to every fiat currency, so my own reasoning, according to my critics, implies that the value of every fiat currency is just as much a bubble phenomenon as the value of a Bitcoin.

My response to the first objection is that even if the strict logic of the backward-induction argument is inconclusive, because of the long and uncertain duration of the time elapse between now and the end state, the argument nevertheless suggests that the value of a Bitcoin is potentially very unsteady and vulnerable to sudden collapse. Those are not generally thought to be desirable attributes in a medium of exchange.

My response to the second objection is that fiat currencies are actually quite different from Bitcoins, because fiat currencies are accepted by governments in discharging the tax liabilities due to them. The discharge of a tax liability is a real (i.e. non-monetary) service, creating a distinct non-monetary demand for fiat currencies, thereby ensuring that fiat currencies retain value, even apart from being accepted as a medium of exchange.

That, at any rate, is my view, which I first heard from Earl Thompson (see his unpublished paper, “A Reformulation of Macroeconomic Theory” pp. 23-25 for a derivation of the value of fiat money when tax liability is a fixed proportion of income). Some other pretty good economists have also held that view, like Abba Lerner, P. H. Wicksteed, and Adam Smith. Georg Friedrich Knapp also held that view, and, in his day, he was certainly well known, but I am unable to pass judgment on whether he was or wasn’t a good economist. But I do know that his views about money were famously misrepresented and caricatured by Ludwig von Mises. However, there are other good economists (Hal Varian for one), apparently unaware of, or untroubled by, the backward induction argument, who don’t think that acceptability in discharging tax liability is required to explain the value of fiat money.

Nor do I think that Thompson’s tax-acceptability theory of the value of money can stand entirely on its own, because it implies a kind of saw-tooth time profile of the price level, so that a fiat currency, earning no liquidity premium, would actually be appreciating between peak tax collection dates, and depreciating immediately following those dates, a pattern not obviously consistent with observed price data, though I do recall that Thompson used to claim that there is a lot of evidence that prices fall just before peak tax-collection dates. I don’t think that anyone has ever tried to combine the tax-acceptability theory with the empirical premise that currency (or base money) does in fact provide significant liquidity services. That, it seems to me, would be a worthwhile endeavor for any eager young researcher to undertake.

What does all of this have to do with the Neo-Fisherite Rebellion? Well, if we don’t have a satisfactory theory of the value of fiat money at hand, which is what another very smart economist Fischer Black – who, to my knowledge never mentioned the tax-liability theory — thought, then the only explanation of the value of fiat money is that, like the value of a Bitcoin, it is whatever people expect it to be. And the rate of inflation is equally inexplicable, being just whatever it is expected to be. So in a Neo-Fisherite world, if the central bank announces that it is reducing its interest-rate target, the effect of the announcement depends entirely on what “the market” reads into the announcement. And that is exactly what Fischer Black believed. See his paper “Active and Passive Monetary Policy in a Neoclassical Model.”

I don’t say that Williamson and his Neo-Fisherite colleagues are correct. Nor have they, to my knowledge, related their arguments to Fischer Black’s work. What I do say (indeed this is a problem I raised almost three years ago in one of my first posts on this blog) is that existing monetary theories of the price level are unable to rule out his result, because the behavior of the price level and inflation seems to depend, more than anything else, on expectations. And it is far from clear to me that there are any fundamentals in which these expectations can be grounded. If you impose the rational expectations assumption, which is almost certainly wrong empirically, maybe you can argue that the central bank provides a focal point for expectations to converge on. The problem, of course, is that in the real world, expectations are all over the place, there being no fundamentals to force the convergence of expectations to a stable equilibrium value.

In other words, it’s just a mess, a bloody mess, and I do not like it, not one little bit.

Who’s Afraid of Say’s Law?

There’s been a lot of discussion about Say’s Law in the blogosphere lately, some of it finding its way into the comments section of my recent post “What Does Keynesisan Mean,” in which I made passing reference to Keynes’s misdirected tirade against Say’s Law in the General Theory. Keynes wasn’t the first economist to make a fuss over Say’s Law. It was a big deal in the nineteenth century when Say advanced what was then called the Law of the Markets, pointing out that the object of all production is, in the end, consumption, so that all productive activity ultimately constitutes a demand for other products. There were extended debates about whether Say’s Law was really true, with Say, Ricardo, James and John Stuart Mill all weighing on in favor of the Law, and Malthus and the French economist J. C. L. de Sismondi arguing against it. A bit later, Karl Marx also wrote at length about Say’s Law, heaping his ample supply of scorn upon Say and his Law. Thomas Sowell’s first book, I believe drawn from the doctoral dissertation he wrote under George Stigler, was about the classical debates about Say’s Law.

The literature about Say’s Law is too vast to summarize in a blog post. Here’s my own selective take on it.

Say was trying to refute a certain kind of explanation of economic crises, and what we now would call cyclical or involuntary unemployment, an explanation attributing such unemployment to excess production for which income earners don’t have enough purchasing power in their pockets to buy. Say responded that the reason why income earners had supplied the services necessary to produce the available output was to earn enough income to purchase the output. This is the basic insight behind the famous paraphrase (I don’t know if it was Keynes’s paraphrase or someone else’s) of Say’s Law — supply creates its own demand. If it were instead stated as products or services are supplied only because the suppliers want to buy other products or services, I think that it would be more in sync than the standard formulation with Say’s intent. Another way to think about Say’s Law is as a kind of conservation law.

There were two famous objections made to Say’s Law: first, current supply might be offered in order to save for future consumption, and, second, current supply might be offered in order to add to holdings of cash. In either case, there could be current supply that is not matched by current demand for output, so that total current demand would be insufficient to generate full employment. Both these objections are associated with Keynes, but he wasn’t the first to make either of them. The savings argument goes back to the nineteenth century, and the typical response was that if there was insufficient current demand, because the desire to save had increased, the public deciding to reduce current expenditures on consumption, the shortfall in consumption demand would lead to an increase in investment demand driven by falling interest rates and rising asset prices. In the General Theory, Keynes proposed an argument about liquidity preference and a potential liquidity trap, suggesting a reason why the necessary adjustment in the rate of interest would not necessarily occur.

Keynes’s argument about a liquidity trap was and remains controversial, but the argument that the existence of money implies that Say’s Law can be violated was widely accepted. Indeed, in his early works on business-cycle theory, F. A. Hayek made the point, seemingly without embarrassment or feeling any need to justify it at length, that the existence of money implied a disconnect between overall supply and overall demand, describing money as a kind of loose joint in the economic system. This argument, apparently viewed as so trivial or commonplace by Hayek that he didn’t bother proving it or citing authority for it, was eventually formalized by the famous market-socialist economist (who, for a number of years was a tenured professor at that famous bastion of left-wing economics the University of Chicago) Oskar Lange who introduced a distinction between Walras’s Law and Say’s Law (“Say’s Law: A Restatement and Criticism”).

Walras’s Law says that the sum of all excess demands and excess supplies, evaluated at any given price vector, must identically equal zero. The existence of a budget constraint makes this true for each individual, and so, by the laws of arithmetic, it must be true for the entire economy. Essentially, this was a formalization of the logic of Say’s Law. However, Lange showed that Walras’s Law reduces to Say’s Law only in an economy without money. In an economy with money, Walras’s Law means that there could be an aggregate excess supply of all goods at some price vector, and the excess supply of goods would be matched by an equal excess demand for money. Aggregate demand would be deficient, and the result would be involuntary unemployment. Thus, according to Lange’s analysis, Say’s Law holds, as a matter of necessity, only in a barter economy. But in an economy with money, an excess supply of all real commodities was a logical possibility, which means that there could be a role for some type – the choice is yours — of stabilization policy to ensure that aggregate demand is sufficient to generate full employment. One of my regular commenters, Tom Brown, asked me recently whether I agreed with Nick Rowe’s statement: “the goal of good monetary policy is to try to make Say’s Law true.” I said that I wasn’t sure what the statement meant, thereby avoiding the need to go into a lengthy explanation about why I am not quite satisfied with that way of describing the goal of monetary policy.

There are at least two problems with Lange’s formulation of Say’s Law. The first was pointed out by Clower and Leijonhufvud in their wonderful paper (“Say’s Principle: What It Means and Doesn’t Mean” reprinted here and here) on what they called Say’s Principle in which they accepted Lange’s definition of Say’s Law, while introducing the alternative concept of Say’s Principle as the supply-side analogue of the Keynesian multiplier. The key point was to note that Lange’s analysis was based on the absence of trading at disequilibrium prices. If there is no trading at disequilibrium prices, because the Walrasian auctioneer or clearinghouse only processes information in a trial-and-error exercise aimed at discovering the equilibrium price vector, no trades being executed until the equilibrium price vector has been discovered (a discovery which, even if an equilibrium price vector exists, may not be made under any price-adjustment rule adopted by the auctioneer, rational expectations being required to “guarantee” that an equilibrium price vector is actually arrived at, sans auctioneer), then, indeed, Say’s Law need not obtain in notional disequilibrium states (corresponding to trial price vectors announced by the Walrasian auctioneer or clearinghouse). The insight of Clower and Leijonhufvud was that in a real-time economy in which trading is routinely executed at disequilibrium prices, transactors may be unable to execute the trades that they planned to execute at the prevailing prices. But when planned trades cannot be executed, trading and output contract, because the volume of trade is constrained by the lesser of the amount supplied and the amount demanded.

This is where Say’s Principle kicks in; If transactors do not succeed in supplying as much as they planned to supply at prevailing prices, then, depending on the condition of their balances sheets, and the condition of credit markets, transactors may have to curtail their demands in subsequent periods; a failure to supply as much as had been planned last period will tend reduce demand in this period. If the “distance” from equilibrium is large enough, the demand failure may even be amplified in subsequent periods, rather than damped. Thus, Clower and Leijonhufvud showed that the Keynesian multiplier was, at a deep level, really just another way of expressing the insight embodied in Say’s Law (or Say’s Principle, if you insist on distinguishing what Say meant from Lange’s reformulation of it in terms of Walrasian equilibrium).

I should add that, as I have mentioned in an earlier post, W. H. Hutt, in a remarkable little book, clarified and elaborated on the Clower-Leijonhufvud analysis, explaining how Say’s Principle was really implicit in many earlier treatments of business-cycle phenomena. The only reservation I have about Hutt’s book is that he used it to wage an unnecessary polemical battle against Keynes.

At about the same time that Clower and Leijonhufvud were expounding their enlarged view of the meaning and significance of Say’s Law, Earl Thompson showed that under “classical” conditions, i.e., a competitive supply of privately produced bank money (notes and deposits) convertible into gold, Say’s Law in Lange’s narrow sense, could also be derived in a straightforward fashion. The demonstration followed from the insight that when bank money is competitively issued, it is accomplished by an exchange of assets and liabilities between the bank and the bank’s customer. In contrast to the naïve assumption of Lange (adopted as well by his student Don Patinkin in a number of important articles and a classic treatise) that there is just one market in the monetary sector, there are really two markets in the monetary sector: a market for money supplied by banks and a market for money-backing assets. Thus, any excess demand for money would be offset not, as in the Lange schema, by an excess supply of goods, but by an excess supply of money-backing services. In other words, the public can increase their holdings of cash by giving their IOUs to banks in exchange for the IOUs of the banks, the difference being that the IOUs of the banks are money and the IOUs of customers are not money, but do provide backing for the money created by banks. The market is equilibrated by adjustments in the quantity of bank money and the interest paid on bank money, with no spillover on the real sector. With no spillover from the monetary sector onto the real sector, Say’s Law holds by necessity, just as it would in a barter economy.

A full exposition can be found in Thompson’s original article. I summarized and restated its analysis of Say’s Law in my 1978 1985 article on classical monetary theory and in my book Free Banking and Monetary Reform. Regrettably, I did not incorporate the analysis of Clower and Leijonhufvud and Hutt into my discussion of Say’s Law either in my article or in my book. But in a world of temporary equilibrium, in which future prices are not correctly foreseen by all transactors, there are no strict intertemporal budget constraints that force excess demands and excess supplies to add up to zero. In short, in such a world, things can get really messy, which is where the Clower-Leijonhufvud-Hutt analysis can be really helpful in sorting things out.

About Me

David Glasner
Washington, DC

I am an economist in the Washington DC area. My research and writing has been mostly on monetary economics and policy and the history of economics. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey's unduly neglected contributions to the attention of a wider audience.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 1,299 other followers

Follow Uneasy Money on