"Inflation is always and everywhere a monetary phenomenon" was Milton Friedman's slogan. It was revolutionary (or counter-revolutionary) when he said it in 1970, but it's now very widely accepted. After all, we make central banks responsible for keeping inflation on target. We do not make fiscal authorities responsible for targeting inflation (unless they happen to control monetary policy too). And we certainly don't give the Competition Bureau or Industry Canada responsibility for targeting inflation (which would make perfect sense if economists believed, as many did in 1970, that inflation was caused by monopoly power).
But nobody strictly believes the full quote from Milton Friedman. I bet hardly anybody even knows the full quote. "Inflation is always and everywhere a monetary phenomenon in the sense that it is and can be produced only by a more rapid increase in the quantity of money than in output." Nobody believes that last bit is strictly true.
What about my own slogan (though it might easily have been Friedman's): are recessions always and everywhere a monetary phenomena? And if so, in what sense?
Robinson Crusoe doesn't use money. If Robinson Crusoe has a bad harvest, his GDP will drop. It's conceivable he might even work less, if there is less to harvest. But is that a recession?
Well, you could call that a recession if you like. But it is missing one very distinctive feature that we normally associate with recessions.
In a recession, it gets harder to sell stuff, and easier to buy stuff. It takes more effort to find a buyer, if you want to sell goods or labour; and it takes less effort to find a seller, if you want to buy goods or labour. Most economists look at an economy like that and say that goods and labour are in excess supply. In the olden days, we would say there's a "general glut".
It's the excess supply of goods and labour that makes it a recession. The fall in output and employment is merely a typical symptom of that excess supply. Less stuff usually gets sold when there's less demand than supply.
In fact, I can imagine an economy having a recession even while output was rising, even rising faster than normal. An economy could experience a general glut, and at the same time have a massive oil discovery that causes its output to rise. Unlikely maybe, but I can't see why it couldn't happen. Just have the central bank halve the money supply at the same time as oil production skyrockets. That should do it. If the discovery is big enough, the increased oil production and oil exports could be quite enough to offset the decline in the rest of GDP.
Inflation is a rising price of goods in terms of money. Because money, understood as the medium of account, is what we price goods in. I don't see how anyone could sensibly talk about inflation without mentioning money.
A recession is an excess supply of goods in terms of money. Because money, understood as the medium of exchange, is what we buy goods with. I don't see how anyone could sensibly talk about recessions without mentioning money.
We live in a monetary exchange economy. Sure, some parts of the economy are handled by barter, like production within the household, and exchanges with close neighbours. But the barter parts of the economy seem to do OK during a recession. Maybe even expand when the monetary economy falls, so people are forced to rely on their own production of vegetables, or that of their family, friends, and neighbours.
It's the monetary exchange economy that suffers during a recession. It gets harder to sell stuff for money. It gets easier to buy stuff with money. There's an excess supply of other (non-money) goods, and there's its flip-side: there's an excess demand for money.
I don't see how anyone could sensibly talk about inflation without mentioning money. I don't see how anyone could sensibly talk about recessions without mentioning money.
When it comes to inflation, we are (nearly) all monetarists now. At least in accepting the shortened version of Milton Friedman's slogan. But very few (if any) economists would accept the full version of Milton Friedman's slogan. We turn a blind eye to the bit where he adds "… in the sense that it is and can be produced only by a more rapid increase in the quantity of money than in output". OK, we say, he was wrong on that bit, but still right nevertheless. We still say that central banks' monetary policy can be given full responsibility for targeting inflation. We recognise that a change in the money supply will change the price level, other things equal. We recognise that a change in some of those other things can also change the price level, holding the money supply constant. That's OK, we say, because it is the job of the central bank to do whatever is necessary by changing the money supply to offset any changes in any of those other things that would otherwise cause inflation to vary from target.
Similarly, if we accept that recessions are always and everywhere a monetary phenomena, that does not mean that changes in the money supply are the only thing that can cause a recession. That's OK, we should say, because it is the job of the central bank to do whatever is necessary by changing the money supply to offset any changes in any of those other things that would otherwise cause a recession.
There ought to be a symmetry here. We are all (quasi-)monetarists when it comes to inflation. We should all be (quasi-)monetarists when it comes to recessions.
Worthwhile Canadian Initiative 
You know, there are non-disequilibrium theories of business cycles… just sayin’.
david: yep. That’s what my bit about Robinson Crusoe was about. Those theories fail to address what to me is the essential (defining) feature of a recession. In what we call a “recession”, it gets harder to sell stuff and easier to buy stuff. We buy stuff with money and sell stuff for money.
And our talking about “excess supply” is just a crude way of saying “it’s hard to sell stuff (find a buyer) and easy to buy stuff (find a seller)”.
If you had phrased it slightly differently, you could provoke a horde of angry Austrians to attack that assertion!
But I am not an Austrian. I shall leave it to them (where is Greg Ransom when he is needed?).
In the meanwhile, I think I should again point out that disequilibrium theories don’t necessarily need money; there are a number of commodities which occupy prominent positions in the economy without goods being generally priced in their units. Oil, for instance. And one can model sticky prices without money by fixing the ratio of oil-to-units-of-representative-GDP (insert your choice of microeconomic rationalization here). Real relative price stickiness rather than nominal price stickiness, perhaps (compare real wage rigidity).
There are a number of disequilibrium theorists to whom this distinction has no impact on policy recommendations – if you’re New Keynesian, and there is an oil shock, you can nonetheless make it go away via monetary adjustment because eventually you can overwhelm menu costs and BAM everything adjusts. Obviously, the crucial detail is not whether there is a disequilibrium, which can happen in any number of ways, but the precise proposed mechanism by which agents are thought to (limited-rationally) set prices.
david: you can have disequilibrium without money. You can have sticky prices without money. Agreed. But you can’t have a generalised excess supply without money. You can’t have it hard to sell stuff, and easy to buy stuff, without money.
In a barter economy, an offer to sell is an offer to buy. If it’s hard to sell one good it must be hard to buy another. If there’s an excess supply of one good, there must be an excess demand for whatever it is they are trying to sell that good for. We live in a monetary exchange economy, and one of the 2 goods in every market is money. When we try to sell stuff, and can’t, that means we are trying to buy money, and can’t. The excess supply of goods that we observe in a recession is an excess demand for money.
A couple of points:
– take the concept that a pure barter economy has a market between any pair of goods, and a monetary economy really only has a subset of all possible markets between any pair of goods, because we specify that money must be one of the goods. Fine. But the alternative to a monetary economy is not necessarily a pure barter economy with markets between every pair of goods – incomplete markets can occur for real, non-money-related reasons, like real transaction costs for certain pairs of goods. An economy with apples, pears, bananas, and oranges can simply not have a market between apples and pears or bananas and oranges and presumably it is easy to see how such non-gross-substitutability can cause the price mechanism to jam up along the way.
In the limit of incomplete markets by removing markets not involving money, one achieves the monetary economy, so to speak. But one doesn’t need to go all the way to prevent excess demands and excess supplies from finding each other.
– Take a non-monetary pure exchange barter economy with no transaction costs and complete markets – just sticky prices, with no other artefacts. Say that everyone is identical and starts with an endowment of ten apples and ten pears. Say that everyone prefers apples to pears at the margin, but fix the sticky price at one apple to one pear. Therefore: excess supply of pears, excess demand for apples, no trade occurs.
If you think “wait, is it really sensible to have a sticky price if no trade actually occurs at that price?”, alter one consumer’s endowment so that at the margin she prefers pears to apples; she will trade with the others until she is indifferent. For sufficiently many other consumers, this will not remove the excess demand/supply. One could to arrange preferences for this consumer so that one good is inferior and the other superior, presumably; then tweaking her endowment will achieve the desired example.
– if your objection is that neither of these represent a general excess supply of “stuff”, well, “stuff” is heterogeneous. Trivially one cannot have both an excess supply and demand for the same good. But why not for two goods, like “stuff” and labour? Or real estate? Or oil?
(I should point out that I do not disagree with the empirical characterization of the ongoing recession as a disequilibrium one, nor even the argument that observed and future recessions are likely disequilibrium phenomena; I do object to the assertion that it is the only theoretically sensible outlook.)
Didn’t Blanchard say that the price of turnips is always and everywhere a vegetable phenomenon? His point was that Friedman’s pronouncement (minus the erroneous bit) is simply vacuous. Of course inflation is monetary – it’s the rate of change in the value of money.
In a recession, it gets harder to sell stuff, and easier to buy stuff.
Actually, it’s quite easy to sell AAA bonds. The real problem is that it’s hard to sell products and most inputs to production. So perhaps the important distinction is between demand for liquid assets and the demand for output. Or is it between ‘flexprice’ asset markets and ‘fixprice’ goods markets?
But we don’t seem to get anywhere with those questions, so let me put a different question to you: what do I have to believe in order to be a quasi-monetarist in your eyes?
Or here’s another way of putting the same question. I’m currently spelling my way through an excellent textbook on macroeconomics, which briefly discusses Old Keynesian economics in Chapter 2, with a more extended discussion of quantity signals and the short-side rule in Chapter 5. Money is dealt with in a way which would, I’m sure, gladden your heart — except for one thing: it doesn’t get into the act until Chapter 11. Could the author possibly be a quasi-monetarist? Just what are the entry requirements for this club?
Nick – thanks for another educational post – you must be a great teacher.
Could you consider making a post about “money” as well?
I think the term “money” is a bitch, and always have a lot of problems in trying to fill it with content.
If I look at myself, I can´t remember a single time during at least the last two years where I used paper money – and how much money do I have in my account? Is it the number that I already own or is it that number plus the limit of my credit card (given that the banks can lend more and that lending, as now, probably doesn’t have any effect on interest rates)? Are my assets money? Sure, everyone can´t liquidate their assets at the same time but doesn’t the existence of assets make a fixed money supply go much further. Without assets – if you sit on money that you don’t want to use – we are kind of stuck. If, on the other hand, I got assets – I might change my assets against your money and thus we would get them into circulation. Etc. etc. etc.
As you see – I am confused – and I don´t think I´m the only one.
David:
I expect there to be expanding and contracting sectors of the economy all the time. That is how resources are reallocated. This process involves surpluses, shortages, and structural unemployment.
So, telling me that sticky prices in a two fruit barter economy results in fewer transactions is nothing like a recession.
It is more like imagining that reallocation could conceivably occur without unemployment if all prices and wages were perfectly flexible. Sectors losing demand just lower prices and wages so that quantity demanded is maintained. And workers get job offers in the sectors with higher demand, give their notice, and start at the new place of employment. It could happen that way. No structural unemployment involved.
Comparing that world to the real world isn’t the same thing as saying there is a recession in the real world always. Or even that if more economic change is occuring, there is a recession. It isn’t a recession because there are areas of the economy with growing demand that match (or in an economy with growing capacity, overmatch) the sectors that are shrinking.
As for your oil and GDP business–I am less sure. But I think here Nick’s frequent appeal to barter causes the problem. You imagine that there is a shortage of oil (lines at the gas station,) and there are surpluses of everything else. This is still a recession, in your mind, because the shortage of oil is just one market. It is close to a general glut of “stuff.” They see the lines at the gas station with people holding various sorts of stuff waiting. They reduce their production of the stuff they would have brought to the gas station.
In a monetary economy, we ask, “what do the frustrated oil buyers do with the money they cannot spend on oil?” The implicit assumption in your story is that they hold it. And quasi-monetarists, say, see… the demand for money has risen. You are assuming a constant quantity of money. There is an excess demand for money matching the general glut of other goods.
I favor GDP targeting, and so in the scenario where there is an increase in the demand for oil and neither price nor quantity increases, the demands for other stuff must be maintained. If the price of oil or quantity of oil rises, (or if it was a decrease in supply, the quantity of oil falls less than in proportion to the price,) then demand in the rest of the economy contracts.
If we imagine price level targeting, and the price and quantity of oil remains unchanged, and there is a shortage, then demand in the rest of the economy is maintained so that the price level doesn’t fall. Only when the price of oil rises does demand fall for stuff, so that its price(s) can fall.
I am not so sure what happens if the quantity of money is held constant or worse, some short term interset rate is held constant. Probably something bad with interest rate targeting. Like usual.
“Similarly, if we accept that recessions are always and everywhere a monetary phenomena, that does not mean that changes in the money supply are the only thing that can cause a recession. That’s OK, we should say, because it is the job of the central bank to do whatever is necessary by changing the money supply to offset any changes in any of those other things that would otherwise cause a recession.
There ought to be a symmetry here. We are all (quasi-)monetarists when it comes to inflation. We should all be (quasi-)monetarists when it comes to recessions.”
How about we change terminology a little? Recessions have to do with Relative Prices. Depressions have to do with Demand.
This way you can still sensibly talk about a recession when there is a supply shock (oil prices for example), that ‘merely’ requires an adjustment of relative prices. And depressions, when the increased demand for money is not offset and there is a lack of demand?
Both can occur simultaneously, but it would help if we’d make a distinction in terminology. Depressions are always and everywhere a monetary phenomena. Recessions are always and everywhere about re-allocation ;)?
Or in terms of the New Classicals, Recessions are RBC, Depressions are Lucas 1972?
@Bill Woolsey
But telling you that sticky prices in a money/”stuff” economy results in fewer transactions is acceptably a recession?
Central to both is some kind of postulated mechanism for rational price stickiness; I don’t see why real and nominal wage rigidity is a respectable concept whereas real real-estate price rigidity or whatever should not be. Yes, it is nothing like the real world, with its stunning history of non-neutral money. My point is merely that it is wholly coherent.
I think Kevin Donoghue has nailed the crucial point. Most people are going to agree that recessions have got something or other to do with money, but is that enough to make them monetarists, or even quasi-monetarists? Nobody (but you) can say until you define those terms.
Nick, you’re free to specialize terms as you wish but that just shifts the terms of the question to Are we in a recession now?
Whereas before we might have asked is this recession caused by monetary or real factors.
Just because there is reported unemployment does not mean there is a general glut. There could well be reservation wages (in real or nominal terms) that cause people to select leisure over work.
If it’s a nominal issue we can go back to monetary policy. If it’s a real wage issue, well not without an interest rate transmission mechanism or a way to shape expectations of output.
If it’s an issue of hoarding accommodating money demand will be good enough but only in that situation.
Oh it’s nice to wake up in the morning to see some good-quality comments. Not sure the quality of my responses will be as good, but oh well.
david: “In the limit of incomplete markets by removing markets not involving money, one achieves the monetary economy, so to speak. But one doesn’t need to go all the way to prevent excess demands and excess supplies from finding each other.”
Interesting. Lets imagine going in the other direction. Start in a pure monetary economy, then move a little bit away from that limit by introducing a very small number of barter markets. The economy should presumably still behave approximately like a pure monetary economy. (I would be embarrasses if it didn’t; if my (implicit) model were “fragile in the limit”!)
Unemployed workers have an excess supply of labour. They want to sell their labour for money, and then sell that money (or most of it) to buy consumption goods. But since they can’t sell their labour, their notional excess demand for consumption goods never shows up in the market for those goods. Now suppose workers’ wages can get paid partly in their output. Brewery workers get some beer in their pay packets. Bakers get some bread, etc. If each worker consumes 100 different goods, but produces only one, I don’t think this would make much difference to employment and output, though it would help things a little. There would still be an excess supply of all types of labour, and all consumption goods. But in the limit, as each firm produced a greater and greater number of goods that workers consume, the excess supply would be eliminated by those barter markets.
I’m a bit lost by your apples and pears example though. Is there any good (except money) that is typically in excess demand in a recession? (Maybe job-search services, or those people whose job is to help you sell stuff?). If recessions empirically had an excess supply of (say) labour, but a roughly equivalent excess demand for (say) output goods, I think I would stop being a quasi-monetarist (at least as far as saying “recessions are always and everywhere a monetary phenomenon”), and start saying instead “recessions are caused by real wages being stuck too high”, or something.
(That last point I think also applies to Kevin’s very good question.)
Nick:
The recession symptoms you describe can also be explained as the effects of a shortage of money. When money was silver coins, the coins would wear out over time. New coins would weigh 20 grams while old coins weighed maybe 19 grams. The economy needs more coins, but Gresham’s law takes hold and new coins are hoarded rather than circulating. The resulting shortage of coins makes it hard to sell stuff, but easy to buy stuff if you’re lucky enough to have coins.
Now replace the coins with bits of paper, each of which is convertible into a coin. If we have free banking, then banks can freely issue new bits of paper on fractional reserve principles and there is no shortage of money. Likewise there is no inflation because the bits of paper are backed by enough assets to make them worth one coin each. But if there are restrictions on the issuance of bits of paper, then there can be a shortage of money. This causes a recession, but since each bit of paper is worth 1 coin, the value of a bit of paper does not change.
Kevin: “Actually, it’s quite easy to sell AAA bonds.”
Yep. For two reasons:
1. they are very liquid. OK, that’s a bit of a tautology. What I mean is that they have all the characteristics that tend to make goods very liquid: thick markets, organised markets, low-information required (unlike used cars they are all the same), etc.
2. Yes, prices that adjust very quickly.
Remember my post on the peanut theory of recessions? Suppose just one good, peanuts, had perfectly flexible prices. We would never see peanuts in excess supply (or demand). It would be very tempting to attribute recessions to something screwing around with the relative price of peanuts. It would be very tempting so say that central banks should control the relative price of peanuts. It would be very tempting to say that we there cannot be an excess demand for money, because you can always get more money by selling peanuts.
Peanuts were an allegory(?) for bonds.
http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/07/the-peanut-theory-of-recessions.html
Kevin: “Didn’t Blanchard say that the price of turnips is always and everywhere a vegetable phenomenon? His point was that Friedman’s pronouncement (minus the erroneous bit) is simply vacuous. Of course inflation is monetary – it’s the rate of change in the value of money.”
If we are trying to explain things, we look for what they have in common with other things, and what they don’t have in common with other things. If we saw the price of vegetables rising, and other prices not, we would start explaining the rise in the price of turnips by noting that fact. It must be a vegetable phenomenon.
Back around 1970, many economists really did put forward explanations of inflation that had nothing to do with money. (The past really was another world, that is hard to believe once existed). Now, I agree, there is nothing logically wrong with such explanations. But they ought at least have some sort of answer to the question: “But if your theory is right, what is happening to the demand and supply of money, and why doesn’t that matter? How does the fact that we are talking about a falling value of money fit into your picture?”
Now, right at the beginning of my post, I addressed the question “what does it really mean to say that we all now accept Milton Friedman’s slogan?” What does it mean to say that we think it is something more than vacuous? And my answer is that we assign monetary policy reponsibility for inflation. That was what was controversial about what Friedman was saying. That’s where the rubber hit the road. Many economists really did believe that that was wrong back in 1970. I can remember it. I was there (just). (OK, I was in the UK, which was further away from monetarism than the US, but read some old stuff by JK Galbraith, or look up “cost-push inflation” to get a flavour.)
This whole post is an argument from symmetry. What does it mean to be a quasi monetarist? When it comes to inflation, someone who does not assign responsibility to monetary policy is not a QM. When it comes to recessions, someone who does not assign responsibility to monetary policy is not a QM.
OK, that’s a negative definition of quasi-monetarism, (which makes it necessary, but not sufficient), but it’s the best I can do this morning. (And see my point above at the end of my response to david).
Up until 3 years ago, I thought we were all (OK, leaving RBC types aside, because they never had any policy-relevance) quasi monetarists in that sense when it came to recessions. But now there’s a crisis, and most of us are off worshipping Baal.
Nick, why doesn’t RBC have any policy relevance? Isn’t it just as important to show what you can’t do as what you can and how an economy looks when you can’t?
Nomi: Thanks! (Not as good as i should be. I’m feeling old and tired.)
“I think the term “money” is a bitch, and always have a lot of problems in trying to fill it with content.”
Yep. You are not alone! Is “money” M1, M2, M3, etc.?
We have to think in terms of a “functional” definition of money. “Money” is whatever we use as money.
There are only 2 functions that matter. (Forget all the textbooks that say there are 3, because money is a store of value. Almost everything is a store of value, sometimes better and sometimes worse than money. Houses are a store of value, but that doesn’t make houses money).
1. Medium of account. That good in which prices are quoted is “money”.
2. Medium of exchange. That good we buy and sell all other goods for is “money”.
(Nearly always, but not always, those two functions refer to the same good).
I’m focussing here on medium of exchange. Currency definitely. What’s in your chequeing account definitely. From there it gets more controversial, and so I’m going to stop there, because I would go too far off-topic.
Bill: as you might suspect, I tend to agree with your comment, and have nothing useful to add.
Phil: agreed. I have done my best (for now) above. Not good enough, clearly.
Martin: “How about we change terminology a little? Recessions have to do with Relative Prices. Depressions have to do with Demand.”
Well, we could do that. But it would mean we would have to go back and say that a lot of things we had previously called “recessions” (like the 1982) weren’t really recessions, but depressions. In fact, no obvious example of such a “recession” comes immediately to mind.
Jon: my response would be: “Is it harder than normal to sell stuff, and easier than normal to buy stuff (“stuff” defined very broadly, to include labour and assets, not just output)?” If the answer is “yes”, then I would call it a “recession”. And I would answer “yes” for the US right now. And Canada a qualified “yes” too.
Mike: I’m with you except for (no surprise) this bit: “Likewise there is no inflation because the bits of paper are backed by enough assets to make them worth one coin each.”
But if there’s a shortage, so they are not exactly convertible on demand (or an excess), then why aren’t bits of paper like shares in a closed-end mutual fund, trading at a premium or discount to the backing assets, even though there is 100% backing? Maybe because shares in closed-end mutual funds are not the medium of account, so have flexible prices relative to the backing assets?
Martin: OK. I should have been clearer. What I meant was that RBC theorists have had (as far as I can tell) no impact on economic policy.
“After all, we make central banks responsible for keeping inflation on target. We do not make fiscal authorities responsible for targeting inflation ”
I would take this with a grain of salt.
First, we already have an extensive system of income support payments, which are now supplying 1 Trillion of deficit spending per year to the private sector.
Do you really believe that if all of these supports were removed — no unemployment, no foodstamps (1/6 of households are now on foodstamps), no disability payments, no welfare payments. That inflation would be the same? That the CB could just handle it? There are many examples of the CB utterly failing to control inflation and keep it on target — e.g. look at the US Fed.
Second, Bernanke immediately called for Fiscal policy in the wake of this crisis. First in the form of TARP, and then with the stimulus. Again, what do you think would have happened to prices if the all the major money center banks failed, if there was no deposit insurance, and if there were no capital injections — note that none of these activities are monetary policy.
Fiscal policy has done much more to stabilize the price level in this current crisis.
I don’t think so… I think it is fragile in the other limit, in that a little move away from a pure barter market can behave approximately like a pure monetary economy, in the sense of removing a very small number of goods from most of their markets. You don’t have to shut down most markets to achieve the effect, because money is not alone in being a good that turns up often on one side of a transaction. Oil-derived goods turn up on the other side. Most people do not buy crude oil; if there is an excess supply of labour and an excess demand for crude oil, even with a cleared market for money we are still dependent on a network of markets to sort out a series of intermediary prices. The adjustment process is non-obvious, even in modern literature. Pick your choice of microfoundations and one can easily theorize why this might happen only slowly, even in a rational universe (and all we have done is remove a mass of markets related to just two goods: labour and oil).
I reiterate that I have no idea how relevant this possibility might be to Real Life; while I have read several commenters on your previous blog posts mention oil (in the obvious context of the 70s oil crisis), my knowledge of past empirics is not great enough to intelligently comment. If Scott Sumner asserts that the quantity of money & money-substitutes is the key economic actor in modern history, well I have no reason to disagree.
Safe assets? Other stores of value? Where the safety of money is inversely related to its inflation rate, and so right now it is remarkably safe in most developed nations.
Considering recent, post-Bretton Woods non-US countries. Are US dollars nonetheless in excess demand during a domestic recession? This would demonstrate that the medium-of-exchange account is not relevant; what agents are really fleeing to is safety. Many countries seek to influence exchange rates, so the reason for sticky prices there is obvious. Perhaps someone else may be able to answer this.
Conversely, Scott Sumner blames past stagflation on supply-side issues, so perhaps there are some recessions not caused by an excess demand for money.
Nick:
The bits of paper might be convertible into coins on demand, but still in short supply (e.g., if banks face restrictions). And the bits of paper would be valued based on the assets backing them, just like any liability, including closed-end funds. But I don’t like the idea of fund shares (or bits of paper) trading at a premium or a discount, since I like to do my theorizing in a no-arbitrage world. Obviously mis-pricing does happen, but not in any systematic way.
Also remember that there are many kinds of convertibility: instant, delayed, certain, uncertain, at the buyer’s option, at the seller’s option, physical, financial, etc. The suspension of one kind of convertibility (instant and certain physical convertibility at the customer’s option) still leaves the bits of paper convertible in many other ways.
DAvid: they are two types of what we call recession , types that are extremely different beasts.
The monoetary one , where we are afraid and want to hold money,whatever it is, for safety reasons.
And the real ones, like the 73-oil shock.. There is a real physical shortage, various stickiness prevent an immediate reallocarion of resources. In both cases we measure unemplyment and concludes they are the same thing. That confusion in 73 , when we tried to solve by monetary expansion what was an oil shortage ( politically motivated, not even a physical one)led to the “Keynesianism no longer works trope.” Though the resulting inflation brought back to real price of oil to its previous level. HAd we adhered to pure monetarism we would have kept the cartel price.
Nick, are you familiar with Keynes’ definition of money in ‘A Treatise on Money’?
Do you think it is a good/useful definition?
It has been easier to sell government bonds than to buy them — hence the falling yields. In fact, you can replace every occurrence of “money” with treasury security in this blog, and still have it be an accurate description of what happens during demand led recessions.
But alas, the central bank has very limited capacity to cause the stock of privately treasuries to increase.
@Jacques René Giguère
No, the policy responses need not necessarily differ – say you are a New Keynesian and blame the failure to adjust to clear labour markets on assorted price-adjustment costs; then as noted earlier, a sufficiently large monetary shock will overcome this cost, even if the root cause of the earlier shock was real.
Fundamentally, to any disequilibrium theory the focus is on finding the least costly way to provoke price or quantity adjustment. In a purely monetary recession, adjusting the quantity of money is least costly. But one can generalize.
rsj: good critique. My guess is you have articulated what many would/should say in response.
“There are many examples of the CB utterly failing to control inflation and keep it on target — e.g. look at the US Fed.”
This would be a more telling critique if the Fed really did have some sort of explicit inflation target (or price-level path, or NGDP-level path, target). It doesn’t. If it did, the Fed would be able to communicate what it is trying to do, and the credibility of that communication could itself have helped massively to prevent AD falling, and prevent both inflation falling as well as the recession. Given the amount of disagreement within the Fed, we really don’t know what the Fed is trying to do. It’s not as bad as citing the failure of the Weimar central bank’s failure to prevent hyperinflation. Or the failure of central banks generally to prevent inflation when they don’t think it has anything to do with monetary policy. But it’s a bit that way.
How much of the financial crisis was itself a consequence, rather than a cause, of the failure of monetary policy? Scott Sumner argues it was mostly the latter. (Unlike me, he also is better at providing empirical evidence in support). If monetary policy fails to prevent expected inflation and expected real growth from falling, that will cause asset prices to fall, and put a lot of banks in queer street.
david: suppose, just suppose, that cows were money. An epidemic that killed off all the goats (or made them go dry) would cause a recession. Unable to drink goats’ milk, people would want to hold more cows. Nobody would spend their cows, because they didn’t want to lose the milk.
Sure, the goats are the original cause of the problem. But goats only cause a recession because they impact the demand for money. If we could print cows, or even make the cows go dry too, so people started to spend their cows again, the recession would end.
And, the price of goats would rise, relative to cows, in this case. In a normal recession, where some of the cows die off, the price of goats would fall relative to cows. But in both cases the proximate cause of the recession is an excess demand for cows. If we didn’t use cows as money, the goat problem wouldn’t cause an excess supply everything else.
goats = bonds, or “safe assets”.
On stability of GE and gross substitutibility. That is not an area of economics I can get my head around easily. IIRC though, from a long time back, there’s a very big difference between the conditions required for stability in a monetary exchange economy than in a Walrasian economy (which is different again from a true barter economy, because in Walras there is 1 big market where all n goods are traded, while in barter there are (n-1)n/2 markets where 2 goods are traded in each.
In a recession though, most stuff is in excess supply and/or its (relative) price falls. Goats (bonds) are the exception in this case. And maybe the price of some inferior goods would rise (or they be in excess demand). But I can’t think of any examples. Kraft dinner, because the unemployed demand a lot more??
David: in the fall of 1973, the disequilibrium was outside “economics”. And no amount of either money or credit could haveeither stop the arab boycott nor brought new oil productions in non-arab-non-OPEC countries on line faster.Policy responses were not inadequate or slower than we wished. They were irrelevant.
JCE: It’s ages since I read the Treatise. I don’t remember the definition.
Nick:
“In fact, no obvious example of such a “recession” comes immediately to mind.”
The oil shock in ’73 mentioned by Jacques?
David:
“No, the policy responses need not necessarily differ – say you are a New Keynesian and blame the failure to adjust to clear labour markets on assorted price-adjustment costs; then as noted earlier, a sufficiently large monetary shock will overcome this cost, even if the root cause of the earlier shock was real.”
You are assuming that monetary interventions have zero costs. If there is no excess demand for money, you will merely create inflation and inflation is generally a ‘bad’ thing. Adjusting to a real shock is difficult enough as it is without moving the general price level some more.
Not to mention that this isn’t exactly an ethical thing to do. The people living on a fixed income, already facing higher costs of living due to the real shock will now face a general increase in their cost of living. Money belongs to us all. If you are going to use government interference to reduce money wages, reduce those wages that are too high, instead of reducing all wages and all incomes.
Elaborating on the ‘bad’ thing: recall the Friedman-Phelps critique that led to the NAIRU? Instituting such a policy is repeating the mistakes of the past.
Martin, Jacques: why would a fall in the supply of oil cause an excess supply of everything else, if it didn’t have some monetary impact.
This goes back to one of my recurring thought-experiments. Suppose there is suddenly an increased desire to save, in the form of antique furniture. The paradox of thrift says that could cause a recession, unless the price of antiques rises sufficiently to reduce desired saving back to its original level. No way. Assume the price of antiques is fixed. So there’s an excess demand for antiques. So what? People can’t buy antiques, so they reformulate their plans in all the other markets, taking that quantity constraint in the antique market into account, and (probably) go on doing exactly what they were doing before. So, there’s an excess demand for antiques, and no excess supply of anything else. (And yes, sod Walras’ Law, which is false).
Antiques = bonds = oil.
Sure, it’s not quite the same, because we use oil to produce other stuff. But that should simply reduce the supply of other stuff, creating an excess demand for other stuff, not a recession.
Ah Nick, you misunderstand me there. I don’t see a recession as an excess supply of everything, I see it as a period of negative growth (that’s how it is currently defined, isn’t it?). That’s why I wanted to make a distinction in terminology.
Assume a Meteor hits the earth and the CB is capable and willing to offset the increased demand for money. There is no general glut in such a situation, however there is a substantial contraction in economic activity for a longer period of time as the economy adjusts.
Countercyclical assets include guns, or gold, and so on. But being countercyclical is not enough to demonstrate excess demand – excess demand would only occur if the price fails to adjust correctly.
Which is the point. For excess demands/supplies to occur, prices and quantities have to be wrong; otherwise there is a real-business-cycle effect but no surge in involuntary unemployment, etc. So the proximate cause of the recession is a failure of the price mechanism; if people want to hold more cows relative to goats, the price of cows relative to goats should rise enough to reflect this. That cows are the medium of exchange is an irrelevancy in this example, as it happens. The problem is the sticky price (and sticky quantity) for whatever reason.
Is the price of bonds restricted? Conceptually one could have a group of goods that are all readily substitutable with each other, and another group of goods which are ready substitutes, but which don’t substitute with the former group – then intuitively within each group prices ‘clear’ its internal market, but there can be an excess demand for one whole group and and excess supply for the other. Only one member of the group may be the proximate cause. Likewise bonds, the price of which adjusts readily enough. Your earlier peanuts blog comes to mind.
But one can distinguish a flight to safety from a flight to liquidity by assessing safe but not-really-adjusting assets, presumably. Are there any? I did mention US dollars for a non-US state that manipulates its interest rate. How might one conceptualize Mundell-Fleming for a small open economy in the context of a disequilibrium approach, does “excess supply of domestic labour” vs. “excess demand for foreign goods” work for you? Because of the presumed substitutability of money for foreign goods, perhaps.
On stability: my reading of the literature is that a sufficiently imaginative choice of example can make any damn thing happen even under implausibly restrictive conditions, never mind plausible conditions. So we are stuck with handwaving and hoping that the market can sort it out. My intuition here is no better than yours.
On Walras’ Law: it behooves me to point out that Walras’ Law works if you actually apply the individual budget constraint that it is derived from; if you model people as reformulating their excess supplies, then you should also allow people to reformulate their excess demands and then such consistency makes Walras’ Law work just fine. There would be no excess demand for antiques and no excess supply of anything else, as expected.
That should be “US dollars for a non-US state that manipulates its exchange rate”.
david: “On Walras’ Law: it behooves me to point out that Walras’ Law works if you actually apply the individual budget constraint that it is derived from; if you model people as reformulating their excess supplies, then you should also allow people to reformulate their excess demands and then such consistency makes Walras’ Law work just fine. There would be no excess demand for antiques and no excess supply of anything else, as expected.”
OK. But do you see what you have done if you do reformulate Walras’ Law in that way? You have converted an interesting (if false) statement “the sum of the values of the excess demands = 0” to a perfectly useless statement “0+0+0+0+0=0”. Because you are saying that the excess demand for antiques is now 0, because people realise they can’t actually buy any, so stop asking for them. Similarly, if there’s an excess supply of labour, unemployed workers will realise they can’t sell any, so stop offering it. Full-bore “discouraged worker” effect, applied to every single market, so that all excess demands and all excess supplies are zero, at all possible price vectors.
Back later.
Here is another thought experiment with some roots in the real world. Say I found pets.com and spend tens of millions of dollars amassing an infrastructure to resell pet food. Investors pour money into my company with what they feel is a plausible business case. I pay employees and consume resources in warehouses and IT infrastructure. Sales fail to materialize and I run out of money, everyone goes home. Investors write off to pennies on the dollar.
In this particular case I produced nothing of value and the recession comes when I lay off my staff and investors write off their losses. So there is a gap there that is real for the investors and employees that is a net detriment to the economy if I instead invested in a business model that worked. I think in that sense it should be no surprise that recessions happen when we make bad decisions. I don’t see how the money supply can change that at all.
I would rather not pick yet another fight about Walras’ Law, honestly… what about the rest of my comment 😛
But ok. Excess demands are unmet demands at the prevailing price, and excess supplies are unmet supplies at the prevailing price. This isn’t actually well-defined, though, since planned demands and supplies can change. So: you have consumers react to excess demand by moving their excess supply to other markets. But holding the prices at which they do so fixed, if new antique chairs popped into existence, these consumers would simply substitute back along those prices and then swap their excess supply for their antique chairs. Hence: there actually was an excess supply of everything else.
Alternatively, if you define excess supply as requiring that people hold resources directly swappable for antique chairs idle (aka money), then there is no excess supply, because they’ve all adjusted their plans to take the limited supply into account. But there’s also no excess demand, for the same reason that they now no longer have resources directly swappable with antique chairs. People do indeed realize that they can’t buy any, so they no longer hold money idle waiting for them. Whether this can still qualify as ‘unmet demand’ depends on your choice of definition, but then you must be consistent.
In the context of unemployment, if an unemployed work uses their time to productively spring-clean their own house whilst seeking jobs at the prevailing wage, does this qualify as excess supply of labor to you? They can’t find a job, so they reformulate their plans and move on. That 24 hour budget has to be allocated somehow! But they would readily undo said reformulation if they could find a job.
Martin: I sometimes hear people say that a recession is defined as 2 consecutive quarters of negative real GDP growth. “By who?”. On what authority can they write the dictionary? And is that a typical symptom of a recession, or a definition? Even those guys at NBER who are said to “officially” say whether the US economy is (i.e. was) in recession look at more things than just GDP. Meaning is use (Wittgenstein). Do we say that Germany and Japan had recessions in 1945? Maybe, but those were weird sorts of recessions. Why not just say that saturation bombing caused GDP to drop?
Sure, I’m putting forward my own definition. You can define it differently. But I would argue that my definition is both: more in accord with common usage; and a more useful definition. But I’m not really hung up on it. (Just a little, I suppose).
david: you know I can’t resist a fight over Walras’ Law 😉
(Especially since, I think (but I’m not sure), a lot of that stability literature I don’t understand assumes Walras’ Law!)
“But holding the prices at which they do so fixed, if new antique chairs popped into existence, these consumers would simply substitute back along those prices and then swap their excess supply for their antique chairs. Hence: there actually was an excess supply of everything else.”
No there wouldn’t be, in aggregate.
1. Suppose the new (LOL, sorry) antique chairs come as a helicopter drop. The excess supply of everything else vanishes, since they have now satisfied their demand for antique chairs by picking up the new ones off the street.
2. Suppose that new antique chairs are produced. The sellers of those new antique chairs will presumably demand goods with the proceeds. It’s only if they demand to hoard money that we get a recession.
A simple model: people take their normal income, and their normal holdings of cash, to the antique market. “Got any antique chairs?”. “Nope, try next week.”. Then they go to all the other markets and spend their normal amount of cash as normal. (If they decided to hold abnormal amounts of cash, just on the off-chance they saw an antique chair they could snap up quick, that would cause a recession.)
Sure, as an accounting identity quantities bought and sold must add up to equal change in stock of money. But, if we visit each market sequentially, we are maximising U in that market subject to the budget constraint and to quantity constraints in every other market. That determines excess demand or supply in that one market. Because the set of markets that are other markets changes as they visit each market in turn, each excess demand/supply is determined by a different set of constraints. So they don’t add up. We buy and sell sequentially, not simultanously.
Here’s how I think about excess demands and supplies: what would we observe? Would we observe people making more than normal efforts to try to buy some? Would we observe people making less than normal effort trying to sell some? If we asked people “would you like to buy an antique chair” would they reply “yes! do you have one?”. We have a sense that unemployed people want jobs in a recession, even if they have given up looking, and so don’t technically meet the Stats Canada definition of “unemployed”. What would happen if new antique chairs did arrive on the market? Would there be a bunch of eager buyers? What would happen if new job offers suddenly appeared? Would there be a bunch of eager sellers? If “yes”, it’s excess demand (or supply).
“In the context of unemployment, if an unemployed work uses their time to productively spring-clean their own house whilst seeking jobs at the prevailing wage, does this qualify as excess supply of labor to you?”
Yes, absolutely.
I’m falling behind.
Nick, I don’t necessarily disagree with your definition, what you’d call a recession, I’d call a depression under the scheme mentioned earlier; you’d probably call it a more severe recession. I’d just call negative growth that is not caused by a not-offset increase in the demand for money a recession.
“Meaning is use”, how do you think the public calls a period of negative growth? Even if we re-define/define recession according to your scheme, the fact remains that you need a word to denote a period of lower or negative growth that is due to non-monetary causes. What will they write in the newspapers? What will they say on the news? What if the pundits disagree on whether it is structural or nominal?
You’re a teacher, how do you call the 73-75 episode when explaining what happened to your students? Do you or did you ever use the r-word when referring to a supply shock? I am open to any scheme and I think you have a very persuasive case in that a recession can be nothing else than a general glut. However how do we call a period of lower/negative growth then, a stagnation/contraction?
Martin: “However how do we call a period of lower/negative growth then, a stagnation/contraction?” That sounds good to me.
73-75 was both a recession and a contraction. Oil was at least partly responsible for the contraction.
david: “Which is the point. For excess demands/supplies to occur, prices and quantities have to be wrong; otherwise there is a real-business-cycle effect but no surge in involuntary unemployment, etc. So the proximate cause of the recession is a failure of the price mechanism; if people want to hold more cows relative to goats, the price of cows relative to goats should rise enough to reflect this. That cows are the medium of exchange is an irrelevancy in this example, as it happens. The problem is the sticky price (and sticky quantity) for whatever reason.”
Yep. Agreed. If all prices were perfectly flexible, then you couldn’t get an excess demand for money lasting longer than an instant. (I don’t follow you on the “sticky quantity” bit though.)
You are losing me on the rest though. Maybe because my brain is overloaded. Sorry.
jesse: if you make a bad (ex post) investment decision, you cause GDP to fall below what it would have been. Agreed, we don’t need to talk about money to explain that fall in GDP. But is it a recession? Where’s the generalised excess supply? Wouldn’t there be a fall in supply, and excess demand, because you have taken productive resources and produced nothing?
Nick, let me ask you this.
Suppose the government passed a law that said as soon as anyone received a dollar, they must immediately use that dollar to purchase something that isn’t a dollar.
In that case, would recessions be impossible?
rsj: short answer “yes”. Long answer “no”, because:
1. You can’t make it immediate. That would make velocity infinite. Can’t be done.
2. One of the whole points of wanting to use money anyway is the same reason we hold inventories: because it’s just too dams hard to do everything at once. (That’s why God invented time).
3. Because the supply of real output might increase, causing a recession, even if M and V were fixed.
4. Because there might be a change in the degree of vertical integration over time, so the same flow of income would require a different flow of transactions.
5. People would figure out a way to get around the law.
6. Other stuff I haven’t thought of.
But, that said, I want to answer “yes”.