There is nothing new in the substance of this post; it is an exposition of standard monetary theory. (Though some might find the exposition interesting.)
It is an open invitation for people to tell me where standard monetary theory of the effects of inflation might be wrong.
It's an invitation open to all, but I do have Austrian economists especially in mind. As far as I can tell, one of the main points of disagreement between Austrians and other monetary theorists concerns non-super-neutralities of money. Both believe money is not super-neutral, but Austrians seem to argue for an additional non-super-neutrality that is not present in non-Austrian monetary theory.
This matters, because I interpret Austrians to argue that the seeds of recessions like the current one are sown by monetary inflation.
First some terminology. A "nominal" variable is usually defined as one which has $ in the units; a "real" variable is one measured in physical (or non-monetary) units. (This doesn't work exactly, because both nominal and real interest rates have the units 1/time, but I am going to ignore this flaw in the definitions.)
Money is "neutral" if a change in the level of the stock of money has no effect on real variables (it only affects nominal variables).
Money is "super-neutral" if a change in the growth rate of the stock of money has no effect on real variables (it only affects the growth rate of nominal variables).
[Note to Post-Keynesians: I have defined neutrality and super-neutrality implicitly assuming "money" is exogenous; but it would be easy for me to redefine them taking any nominal variable as the exogenous control instrument or target. Indeed, my thought-experiment below makes no distinction at all between endogenous and exogenous variables.]
Standard theory says that both neutrality and super-neutrality are false in the short run. The main reason is that prices do not adjust instantly.
Standard theory says that neutrality is true in the long run (with careful caveats), where prices are assumed perfectly flexible in the long run.
Standard theory recognises that super-neutrality is not precisely true in the long run. There are some non-trivial real effects of an increase in the growth rate of the money supply. Inflation does matter in the long run.
Let us see what those real effects might be. But first, let me explain my approach.
There are two ways to solve simultaneous equations. The first way is the way (or ways) they taught you in school. The second way is to guess the answer, then check to see if your guess is correct. I'm going to use that second way. This is often how economists solve for an equilibrium. They guess what the equilibrium might look like, then check to see if their guess makes sense.
My initial guess is that money is super-neutral.
There are two economies, A and B, running side-by-side in parallel universes. At any point in time, the real variables are exactly the same in A and B. They have the same real resources, technology, population, preferences, produce the same quantities of all goods, and have the same employment, relative prices, real incomes, real savings and investment, real growth rates, etc. But B has higher inflation than A. Let's say inflation is 10% (10 percentage points) higher in B than in A. And the money supply (defined in all possible ways) is also growing 10% faster in B than in A. And it's not just average inflation that is 10% higher in B than in A; each individual price (or wage) is also growing 10% faster in B than in A. And nominal interest rates are also 10% higher in B than in A. But real interest rates (nominal interest rates minus inflation, however defined) are the same in A and B.
That's my initial guess. If it were true, money would be super-neutral. We wouldn't care about inflation, since we only care about real things.
Now, my initial guess is wrong. Here are the standard reasons why we know it's wrong (the non-super-neutralities of money):
1. If we assume that currency pays no interest, then the real interest rate on currency must be 10% lower in B than in A. This is the inflation tax on money (or currency). If we assume that the demand for money is a (negative) function of the real rate of return on money, then the real demand for money must be lower in B than in A, so in equilibrium the real stock of money M/P must be lower in B than in A.
2. If the money supply is growing 10% faster in B than in A, the central bank must be printing more money in B than in A, and will be earning higher real profits from seigniorage. (I'm assuming we are still on the right side of the Laffer curve for the inflation tax, so that the smaller M/P effect does not offset the bigger Mdot/M effect). And the higher profits will add to government revenue. This means government expenditure can be higher, or other taxes lower. To get a rough estimate of the size of this seigniorage, assume currency is 5% of GDP (in the ballpark for Canada), so a 10% higher money growth rate would create at most (ignoring the downward slope of the money demand curve) extra seigniorage of 0.5% of GDP. (And the Bank of Canada would see its annual profit fall from the current $2 billion to about $0.5 billion if it cut the inflation target from 2% to 0%.)
3. If nominal interest rates cannot go negative, and the same shock hit economies A and B, that shock might cause some nominal interest rates to hit the 0% lower bound in A but would leave them above the 0% lower bound in B. (Remember that nominal interest rates in B are guessed to be 10% higher than in A).
4. If nominal prices (and wages) are short run sticky, then nominal prices must either change more frequently in B than in A, or else change by bigger amounts when they do change. This may cause bigger costs of changing prices (menu costs) and/or bigger (inefficient) fluctuations in relative prices in B than in A. (The same applies to adjusting tax rates, minimum wages, etc..)
5. People may just get confused by a changing value of money as a device to measure values of goods, and never adjust to inflation. At the very least, when deciding whether this is a good price for eggs they will need to remember when exactly it was they last saw eggs at $2 a dozen.
6. I've probably forgotten some non-super-neutralities.
So, tell me where this list is incomplete, or wrong.
One at a time please!
Worthwhile Canadian Initiative 
Nick,
I’m by no means an expert, but it seems to me that you have entirely left out one of, if not the most important considerations for the neutrality of money.
Assume no central bank and, to start, no increase in the supply of money. Then assume an alcoholic counterfeiter, making perfect money from his basement in St. Louis. As he spends his artwork, the first effect will likely be an increase in the market price of alcohol in the St. Louis area, before the new ‘money’ diffuses over both products and geographies. This is entirely different that just adding a zero to all the existing money in place. In general, new money distorts relative prices primarily because of its non-uniform flow, not because of its level of increase.
Regards, Don
Don: OK. A and B are exactly the same, except on of the people in B is a counterfeiter, who prints money. This means that he collects an inflation tax off all the other people. This changes the distribution of income in B compared to A. If Bob the counterfeiter has exactly the same preferences as the average person, the pattern of demand for different goods is exactly the same as in A. It’s just that Bob gets to consume a lot more goods, and everyone else a tiny bit less. If Bob the counterfeiter really likes Scotch, more so than the average person, the total amount of Scotch produced will be higher, and the relative price of Scotch higher, and the total amount of other goods and their relative prices will be a tiny bit lower.
The non-super-neutrality will be exactly the same as a 0.5% tax on income, transferred to Bob. That’s small beer compared to the other re-distributions of income going on with the tax/transfer system.
These redistributions of income will change the pattern of demand and may change relative prices. But we don’t normally call that a “distortion”.
In my example, that 0.5% redistribution goes to the government, not to Bob the counterfeiter. But I can’t see how that makes any serious difference. Just that the government probably will spend it on different goods than Bob.
I don’t see why it should sow the seeds of disaster.
I have read some of the Austrian literature on the business cycle but wouldn’t consider myself as an authority on Austrian thought. But I will give it a go regarding what an Austrian might think could be incomplete regarding your essay on the effects of inflation.
I don’t think an Austrian would blame inflation for the business cycle in the manner you have described. What I find missing is the injection mechanism of inflation in the economy. If inflation is 10% higher in economy B than economy A, how did it get that way? Is Ben Bernanke flying over economy B in a C-5 Galaxy dropping currency out the back of the plane and people are picking up that money and spending it directly on consumer goods? If this is the case, I wouldn’t suppose this would cause a boom bust cycle as described by the Austrians.
In the case where the money supply is expanding through the banking system by increasing the supply of loanable funds this, on the other hand, would create an unsustainable boom Austrians I think would argue. A question for the standard theory is would it make any difference for the economy if the supply of loanable funds increased due to people saving more; or if the supply of loanable funds increased due to an increase in credit created out of thin air by the banking system? The Austrians would say yes it make a difference to the economy whether there is an actual increase in savings due to the people becoming more future oriented that then becomes available for lending rather than if more loans are available to due the central bank increasing the amount of credit by pushing a few buttons on a computer.
And if standard theory would happen to agree with the Austrians that this in fact does matter, what then would standard theory say would be the impact on the economy regarding the difference between and actual increase in savings versus an artificial increase in credit created by the Federal Reserve.
On the other hand if standard theory did not agree with this why not? Would saving not be necessary and people preferences regarding consumption and savings not matter, since after all, the Fed could simply create artificial savings with the push of a button.
So, the Austrians are asking different questions then standard theory. I think why Austrians ask different questions are that Austrian theory is a capital based theory macroeconomics where as standard theory is a labor based macroeconomics. Austrians think about the stage of the production process and how the production process gets distorted through credit expansion. What is missing in standard theory is how people’s savings preferences affect the economy and what happens in an economy when those signals regarding saving preferences are distorted by the Fed.
Tom: new base money is printed and introduced to the economy either by the central bank (my example), or by Bob the counterfeiter (Don’s example). This inflation tax means that people holding money get taxed an extra 10% of their currency holdings. The revenue from that tax gets handed over to the government, which can do three things with it: spend it; cut other taxes; or save it. (In Don’s example, Bob the counterfeiter spends it on booze). This is covered by my non-super-neutrality #2.
Why should a 10% inflation tax on base money (or currency, if reserves at the central bank pay interest) cause such disasterous effects? Why is it different from any other tax (except for 1,3,4, and 5)?
If the government saves all the revenue from the inflation tax/seigniorage, so that national savings increase, this seems no different from the effects of a small tax increase that generates a government budget surplus. Would an increase in the government budget surplus of around 0.5% of GDP (for a 10 percentage point increase in the inflation rate) cause disasterous consequences?
In any case, the government cannot run a surplus forever. In the very long run, the government must either spend the inflation tax by increasing G, or else cut other taxes to compensate. So national savings is not different in the very long run anyway.
And remember, I guessed that the real interest rate was the same in A and B, and this guess seems to check out. With the real interest rate the same, desired consumption will be the same for the same disposable income, and desired investment will be the same. Except for the fact that the government may either spend or save the revenue from the inflation tax, C+I+G is the same between A and B, so my guess here checks out OK.
Nick,
Government revenue from seigniorage is so small that it can safely be ignored and plays no role in the Austrian business cycle theory. But I get the feeling that most macroeconomists are working on such an aggregated level such as national savings that you are missing what is happening on a more micro level.
I will try another example to get across a point of interest in the Austrian theory of the business cycle. Let’s draw the supply and demand of loanable funds for commercial businesses. Now suppose the supply of funds is made up of saving from individuals and the demand for funds are from business wanting to make some sort of investment in machinery or working capital. In the first case, many people become more future oriented and the supply of saving increases and the supply curve moves to the right, which decreases the nominal interest rate for those funds. In the second case, people’s time preferences remain unchanged and the supply of saving does not increase, but the Federal Reserve steps in with open market purchases that increase the amount of money and credit available to the banking system for lending. The supply of loanable funds increases, the supply curve moves to the right, and the nominal interest rate for those funds has fallen. Question. According to standard theory is the effect on the economy the same in both cases? The Austrians would say no.
Tom: I am saying that the seigniorage revenues are exactly the same magnitude of any potential extra supply of loanable funds from printing money, because their source is the same. If the seigniorage revenues are too small to matter, then any extra supply of loanable funds is too small to matter too.
According to standard theory an increased desire to save has different effects from increased money growth rate (though it matters what the government does with the revenues from printing money).
I understood Austrians to be arguing that any positive money growth rate would lead to very bad consequences. Comparing my two economies A and B, I just don’t see it (except for the non-super-neutralities I have already listed.
The simplest case is where we assume that the government uses the seigniorage revenues to cut other taxes by the same amount. If we then ignore my 1,3,4, and 5, money will be super-neutral (unless the other taxes were more distorting).
To see this, note that real disposable income will be the same in both A and B. Let us guess that the real interest rate will be the same. So consumption demand is the same, Investment demand is the same, government spending is the same, so C+I+G=Y is the same in both A and B. So my guess checks out. No contradiction or disequilibrium. In other words, printing money does not affect the supply of loanable funds, or the demand, or the real rate of interest.
Nick: I think you’re dispensing with the time path argument too quickly.
A critical point of ABCT is to explain why money creation tends to be inflationary. Just on the basis of MV=PQ, a stable price-level is possible in the face of inflationary pressure. Yet we know from experience that predominantly P rises with M and Q falls with M. Thus given a price-level target where fluctuations above trend are met by fluctuations below, Q will periodically recess. Why is this so?
ABCT argues that this arises in part from the time-path of money: money is created by banks and lent to producers (this is true even for home mortgages). Certain forms of production–those where consumption is relatively more deferred–are more sensitive to the interest-rate. That is, the relative structure of production shifts. Given scarce resources, longer production cycles now bid away inputs from shorter production cycles. Yet people’s time-preferences remain unchanged. Given that the prior arrangement was efficient, this new arrangement will likely result in higher prices for immediate consumption goods that more than offset the lower prices that arise from the stimulated supply of deferred consumption goods. Q may even fall but in either case, inflation results.
Also, because the underlying time-preference has been unchanged, the endowment of society is depleted–it is consumed away i.e., the production possibility frontier may even contract during this time driving the spread between the actual and the natural rate further. This leads the inflation-rate to accelerate over time. Should this occur (essentially evolving into stagflation), the only way to achieve of the price-level target is by setting the actual rate above the natural-rate (policy of getting ahead of inflation as under Volcker).
IMO, this latter inflation argument is superfluous to the theory. The critical part is that the distribution (structure) of production changes, but the Austrian’s predicted the 1970s which should be a credit to their theory.
Last we need to consider the final element, why does Q recess when ‘M’ is used to average out the price-level to a trend-line rather than merely causing a decline in ‘P’? Austrians argue that roundabout production being interest-rate sensitive is curtailed quickly (EMH), but structural limitations prevent immediate-good production from immediately benefiting from the reduced competition for resources (some factors of production are rendered unavailable for any purpose during the transition and some may be lost). Thus Q recesses but the price-level sticks.
Nick,
I am not so sure that the seigniorage revenue is the same magnitude as the increase in money and credit by the Federal Reserve, but even if it were it is not the seigniorage revenue that is the problem. It is the distortion of the interest rates caused by the injection of credit in to the loanable funds markets. It would be analogous to distortions that are caused in product markets by price controls. When the government set the price below the market price this leads to excess demand for a product. When the government drives the interest rate below the natural rate of interest this also leads an excess demand for actual saving. However, the Fed papers this excess demand for actual saving over with artificially created credit. Food for thought for standard theory: does driving the interest rate below the natural rate have any effect on the economy other than inflation?
Also, Austrian thought on the business cycle is not monolithic. So while some Austrians may have argued that any monetary growth is bad, many have argued that money growth that keeps the nominal spending relatively constant is ideal, that is, any changes in V should be offset by M in order to keep Py constant.
My final observation is that while Austrians think the economy is always tending towards equilibrium it is not alway in an equilibrium state. For example the supply of saving and demand for saving (investment) is not in equilibrium with the Fed distorting the loanable funds market with artificially created credit. I certainly hope you will get some more comments from others who might be able to illuminate the Austrian position better than I can.
Suppose that the government spends its seniorage revenue on tanks. Because of short run price stickiness, it can purchase 10,000 tanks. Now, rather than a 10% inflation rate “goal,” the goal is to create the amount of money necessary to purchase 10 tanks. It is possible that there is an inflation tax high enough to purchase 10,000 tanks in the long run, after the prices are flexible. But, it is possible that this is not feasible. The maximum revenue from the inflation tax cannot purchase 10,000 tanks in the long run. But, they try and try. Either they destroy the money in a hyperinflationary disaster or they give up. If they give up, then the tank manufacturing industry must contract!!! The former tank workers will have to find jobs producing other things. A recession.
Change the argument a bit so that the revenue goes not to the govenrment but to subsidize bank lending. In the short run, the subsidy is more effective because prices are sticky. Once prices become flexible, their is less revenue and a smaller subsidy for bank loans. But, the real interest rate is what is being targetting. And so, the inflation rate is increased more and more to try to keep the subsidy the same. To keep real interest rates low or purchases of capital goods as high. If the long run laffer for seniorage will not allow for enough revenue for this subsidy for loans, then either they futilly try and raise money growth and inflation more and more until there is a hyperinflationary diaster, or else they give up, and there must be a contraction in those industries where the borrowers were spending their money.
I think the Austrian Business Cycle theory balls up this story (with the assumption that the goal is a near zero real interest rate that is not feasible in the long run,) and the highly temporary impact of lending money into existence that is accepted with the intention of spending. The excess supply of money (before prices adjust) is matched by an increase in the supply of credit.
The assumption that all people have equal preferences is an assumption too far for me. The difference between an inflation tax and some other tax (in terms of its effect on prices and the economy) is precisely that the distribution is not uniform nor predictable. It depends on many factors, including who gets to spend it first. It tends to accumulate in the fancy trends of the day: high tech stocks then real estate then commodities then bonds then probably back to commodities and gold next. This is where I start to sympathize with the behavioral economists (ie. psychology matters, if all the new money is buried in a giant hole forever than none of this matters). When the money circulates (unevenly and possibly in fits and spurts) the boom in prices represent bad information that misleads entrepreneurs to mal-invest (ie. borrow the cheap cash to create what eventually becomes over-capacity for goods that have no real sustainable demand when the circus leaves town). Inevitably that condition leads to a bust that corrects the imbalances by re-structuring the economy and liquidated the bad debts, freeing the assets of bankrupt firms to be acquired by people doing more productive things. So booms are bad, busts are necessary evils that nevertheless produce huge real losses — and who knows what could have done achieved were that money properly invested from the beginning.
That’s my understanding.
If we assume that with higher inflation, nominal interest rates and other returns on capital are higher, and if we assume there is no inflation tax credit, then part of the income tax paid on interest, dividends and capital gains is in effect a wealth tax, not an income tax.
Alex: yes. Agreed. Take a simple example. Suppose nominal interest rates are 4%, inflation is 2%, and you are in a 50% marginal tax bracket. Your after tax real rate of interest is 0%. With numbers like that (and they are not totally imaginary), it is easy to see why people who have the cash, and don’t need a mortgage, would buy a house, where you pay no tax on the imputed rent or “capital gains” (which is really mostly just inflation). The opportunity cost of their own funds is zero. Buying and living in a mansion costs you nothing, except property taxes, heating, and maintenance.
That’s a massive non-super-neutrality, that could have very serious effects in terms of house price bubbles, even at low inflation rates. And it’s easy to understand how it works. (It could be fixed easily by indexing the tax system properly, of course, and is maybe fixed partially in Canada by the lower tax on some forms of investment income). I wish the Austrian argument were as straightforward.
Bill: I understand what you are saying, and it makes perfect sense to me. (If you print money to try to get more seigniorage than the Laffer curve allows, or try to target some real variable above or below it’s natural rate, the result will be a hyperinflationary disaster).
But is THAT what the Austrians are trying to say? I thought they were saying (or one of the things they were saying) was that printing money, even if it the amount required to target a steady (say) 5% inflation, would inevitably lead to disaster. I didn’t understand your last bit, about “I think the Austrian Business Cycle theory balls up this story…”. Are you perhaps saying that the Austrians get confused between your story, as I have re-described it above, and the idea that printing money always leads to disaster, even if it’s just enough for steady inflation? If that’s what you are saying, that makes a lot of sense (as what you say generally does).
Jon and Tom: So your interpretation of the Austrian argument is that printing money to cause a steady (say) 10% inflation will increase the supply of loanable funds, and lower real interest rates. In the case where we assume the government cuts other taxes to compensate for the increased seigniorage revenue, so the budget stays in balance, and government savings is unchanged, I disagree. There will be no effect on the real interest rate (ignoring any other non-super-neutralities).
There are two ways to see this:
In the output market, let’s guess that real output and the real interest rate stays the same, and then check our guess. Since total taxes (including the inflation tax) stay the same, disposable income stays the same, so C stays the same, and I and G stays the same, so C+I+G stays the same, so Y stays the same. So our guess checks out. If we were in equilibrium in the output market before, we stay in equilibrium.
In the loanable funds market. Again, my guess is that Y and r stay the same. I and S stay the same. Now you (and the Austrians) are thinking: “what about the new money that is being printed and lent out by the central bank? Doesn’t that increase the supply of loanable funds for investment?”. NO, it doesn’t. Here’s why. If real money is M/P, then the new money being printed is (say) 10%(M/P). Assume zero real income growth and constant velocity for simplicity (my argument does not depend on this, but it makes it easier for me), then inflation is also 10%, so people will need to borrow 10%(M/P) each year just to keep their real money stocks constant over time. So the extra 10%(M/P) supply of extra loanable funds is exactly matched by an extra 10%(M/P) demand for loanable funds so people can keep their real money balances constant despite inflation.
In other words, the extra money printed and lent out by the central bank does NOT go to fund investment. It goes into people’s pockets.
I think, in a nutshell, that is what is wrong with the Austrian argument. In a steady inflation, flexible prices, actual equal expected inflation, printing money to increase the supply of loanable funds creates inflation and that creates an equal increase in demand for loanable funds so people can replenish their real money holdings. No effect on the real interest rate or investment.
pointbite. I don’t really need the assumption that preferences are the same. If they are different, then there will be a non-super-neutrality, but it is no different from the effects of any tax that changes the after-tax distribution of income.
Yep.
In the short run, assume inflation is fixed. Printing money and lending it out does not cause inflation (by assumption). So the supply curve of loanable funds shifts right by (Mdot/M).(M/P)=Mdot/P. So the real interest rate falls and investment increases.
In the long run, prices are flexible, and every percentage point increase in the money growth rate causes a percentage point increase in the inflation rate. So the demand for loanable funds also increases by (Pdot/P).(M/P)=(Mdot/M).(M/P)=Mdot/P as people seek to replenish their real money balances. So both the demand and supply curves for (real) loanable funds shift right by the same amount. No effect in the long run on real interest rates and investments.
The Austrians include the newly-printed money in the supply of loanable funds, but forget to include the demand for newly-printed money to compensate for inflation in the demand for loanable funds. They “balls up” the loanable funds market diagram.
I think there is an ideological or philosophical angle to this. Austrians are almost always libertarians and my read is that they think it is immoral/dangerous for government or a central bank to control the money supply and they would prefer money have axiomatic underpinnings (i.e. gold or silver) independent of government. To a hard core Austrian/Libertarian the B universe demonstrates 10% more immorality and so they view it with revulsion; “that universe is going to hell in a hand basket”.
Well said Patrick.
I was going to mention the tax angle, but Alex beat me to it. It is by far the most important violation of super-neutrality. I don’t agree that it is “easily” fixed through indexation. Every single financial flow, including money going in an out of your bank account, would have to be adjusted to reflect the price level at the time. The US income tax system is already a monstrosity, with indexation of all financial transactions it would be 10 times worse. The best solution is to eliminate taxes on capital. Just tax labor (progressively if you prefer) and consumption (which is actually a disguised labor tax.)
Bob Murphy has a post up where he claims Rothbard didn’t believe in the Fisher effect, because he didn’t believe there was such a thing as “expected inflation.” If so, it would explain why Austrians have trouble dealing with super-neutrality, but I don’t think this is the view of more sophisticated Austrians.
Here it is, with Bob’s correction:
http://consultingbyrpm.com/blog/2009/06/toward-new-theory-of-inflation.html
Nick:
Mises (for instance) argues rather explicitly that eventually what you say is so. That inflation builds until the real-rates stabilize. I don’t think he’s forgotten to include the rising price-level. He merely assumes that it isn’t instantaneous.
I.e. it is precisely the point you mention at which the recession occurs unless interest-rates are lowered further. This is why the austrian’s argue that this process is harmless for short-durations (the price distortions do not result in structural changes and harmful for long-durations (if the Chinese retard changes in the price-level).
Here are some quotes I posted at Scott’s blog a while back:
Mises:
Hayek writes:
Hayek continues, perhaps more contentiously:
Jon,
Everything that Hayek and Mises are saying agrees virtually 100% with the various strands of mainstream theory, it agrees with the monetarists and the New-Keynsians.
Is that the point you’re trying to make or have you somehow got the impression that this is different?
Like Adam, I can interpret those passages from Mises and Hayek in a way that is compatible with what is now the standard view. In particular, they say that the attempt to keep the interest rate permanently below the natural rate by printing money will lead to (or require) ever-accelerating inflation. And I infer from that that they believe a permanently higher rate of inflation does not lower the (real) market rate of interest.
They just focus more (and a bit differently) on the transition, in real time, from one inflation rate to a higher one.
Maybe my beef is with some of their followers, rather than Mises and Hayek themselves. In particular, those that argue that a permanent target for inflation that requires any printing of money must have disasterous non-super-neutralities. The ones who draw a loanable funds market diagram, with savings and investment, than add the supply of newly-printed money to the supply of loanable funds (while ignoring the demand for newly-printed money in the demand for loanable funds), then say how this “distorts” the rate of interest, etc. and use this to explain why you must allow deflation during periods when the supply of output is growing, and why you must not target 2% inflation.
They take what might be a useful model of the short run effects of printing money (before the increase in inflation has affected the flow demand for nominal money) and use it for long-run analysis of steady state inflation targets.
In fact (this is not new, and not new to me, but it needs saying anyway) you have to give Mises and Hayek credit for saying something way before the rest of the profession figured in out after Friedman said it again in ’68.
Just I think Friedman said it more clearly, and tied it in with the neutrality insight, just slipping one extra time derivative in. But that’s being picky.
What makes it clearer in Friedman is that he treated the growth rate of the money supply as the control instrument, while Mises and Hayek treated the rate of interest as the control instrument. It’s even clearer if you treat the target rate of inflation as the control instrument.
Nick:
I agree that I find most restatements of the Austrian argument wanting. The cartoon description is accessible to the layman and gets repeated. Practitioners on the other hand seem to begin with a mental bias against the Austrian argument. They ‘know’ apriori that the Austrians ‘lost’ the argument, and they approach the analysis from that perspective. Part of the trouble is that Austrians lost. Their pool of acolytes is limited by the tendency of the smartest to perceive which battles are worth fighting–being an Austrian isn’t useful career wise. The result I think is a failure to adequately distill the Austrian argument into its salvageable components.
Much is made of the fact that Hayek and Mises failed to distill the argument themselves, but I view this as akin to Einstein’s failure to unify quantum mechanics and relativity: it doesn’t make his earlier work wrong just incomplete.
Jon: That makes sense to me. I mostly agree. But I don’t think the problem is that they “lost”, so much as that many of the bits they got right got incorporated into standard economics. So when we now think of Austrian economics, we think only of the remainder. The hegemony steals all their best stuff, then defines them in the ways they differ from the hegemony. It can’t be any other way.
More substantive commentary:
Mises never believed that a policy of inflation targeting was feasible in the same sense that a stable (and predictable) inflation plays a role in the foregoing analysis.
Although I believe he would concede that a CB could stabilize a metric (such as BLS CPI-U) that the metric was not the same ‘inflation’ just discussed. It is an imperfect proxy of such that depends upon some weighted market basket. So ‘price inflation’ in an abstract sense is not controlled and not predicable even in an inflation targeting regime.
If we take the Euro and ECB policy as a proxy, we can see something akin to the Austrian’s antecedents in years immediately preceding the bust. The ECB had a target of mild inflation. According to the ECBs proxy measure, the policy was successful; however, commodity price inflation was quite high. There is a mapping of the Austrian argument to the notion that commodities are ‘immediate goods’ and finished goods are the fruits of an extended production cycle. Thus, commodity prices experience inflation whereas finished goods benefit from the depressed interest rate. Because the ECBs metric essentially ignores commodity-prices but people ‘believe’ in the ECB, expectations will be wrong and a sustained spread in the real-rates will be maintained for some time.
Jon: interesting. To translate that into the language of the hegemony: the choice of which nominal variable (or price index) to target, and which monetary policy instrument to use to hit that target, will have implications for how relative prices fluctuate in the face of shocks, and so will have real effects.
“In particular, those that argue that a permanent target for inflation that requires any printing of money must have disasterous non-super-neutralities”
Nick, with regards to the demand for money bit — does it matter if that demand is temporary? Unless I’m missing something (probably, I haven’t read everything carefully) if demand disappears as quickly as it arrived (ie. before anybody could do anything about it) then what, where do those dollars go? If you’re assuming the central bank will mop up liquidity, given they were unable to control events on the way down, why is it unreasonable to assume they will be unable to control events on the way up?
Someone made the point about Austrians being Libertarians… maybe a fundamental mis-trust of government is not irrelevant. Even Friedman argued any system that requires a man to “work it right” is a bad system.
The Austrian story is more one of rising than steady inflation, leading to either deflation or hyperinflation, and that steady inflation would be ineffective and therefore unnecessary for the purposes of government. They view this as theft, like other taxation, just more deceptive. In addition, antique Austrians seem to assume V is constant so maintaining a constant M is sufficient to prevent changes in PQ, thus assertions business cycles can be eliminated following their policies. Personally, I think many just want riskless real return.
Isn’t there a foreign sector? Surely the market for lendable funds is a world market and the main effects of money supply changes are non the exchange. Hey, I thought Canada was a small country!
oops
… the main effects are on the exchange rate…
Jon,
I find Mises incomprehensible to modern ears (he makes no clear distinction between real and nominal interest rates), but Hayek uncontroversial (how does his position differ from that of Friedman for instance).
What strikes me about all this is the underlying assumption that the economy is buzzing along in perfect balance (so no pre-existing imbalances) and then for some reason the (presumably bored) monetary authorities decide to pep things up a bit. Seems an unlikely scenario to me.
I’m also puzzled that there is no discussion of total debt and its significance (given what is happening at the moment). Part of the problem I think is the ambiguity of the word money. Is money M1, M2, M3, … Mn?
Oops
of course I didn’t mean that Canada was a small country geographically, I meant that it was a relatively small economy in a global sense.
reason at 4:56, very well said.
RBC models have the same problem, no nominal frictions means no good can possibly come of monetary policy. Thus monetary policy only does harm… That said, the models are not useless since real frictions matter as well.
However, I’ve read a fair amount of Hayek (he’s actually one of my intellectual heroes, Krugman is another so go figure) and it’s clear he didn’t mean much of what modern “Austrians” understand (perhaps the same could be said of Keynes). Further, the real problem with the Austrians as represented on these blogs is they don’t have the even slightest understanding of mainstream economics yet are somehow sure it’s wrong.
Nick Rowe:
“Jon: interesting. To translate that into the language of the hegemony: the choice of which nominal variable (or price index) to target, and which monetary policy instrument to use to hit that target, will have implications for how relative prices fluctuate in the face of shocks, and so will have real effects.”
This just translates to “history matters” doesn’t it. So are we post-keynesian now?
Nick:
One important issue is “what version of Austrian economics?”
Selgin, White, Horwitz, and Garrison claim that having the money supply vary to offset changes in velocity is a good thing, and that the deflation rate should be the negative of productivity growth. Increases in the quantity of money not offset by decreases in velocity have injection effects. These can be quite persistent, but must eventually be reserved, and every downturn is interpreted to begin as just this sort of reversal. Fluctuations of nominal expenditures also have undersirable effects, in their view in a standard “new Keynesian,” old monetarist way. But there is an insistent that modest deflation (reflecting productivity growth) is optimal and not harmful in any way. There is also a notion that private issue of hand-to-hand currency and the absense of reserve requirements will result in a banking system that has that effect.
The “Rothbardians” claim that the quantity of money should change with gold mining of gold, and the price level should chance with the relative price of gold. Any increase in the quantity of money has injection effects that can be quite persistent, but must eventually be reversed. This is associated with a view that banks should keep 100% gold reserves.
Both in this thread, and also in the discussion of Austrian economics on Sumner’s blog, someone mentioned the claim that inflation must accelerate because otherwise the government would not acheive the purpose of inflation. The assumption is that the purpose of inflation is to distort relative prices and the allocation of production. In particular, lower real interest rates. But that has some purpose too… more capital accumulation?…. higher real wages?
Anyway, the equilibrium processes that involve earmarking the seniorage to subsidize credit market (not the governent budget like you assume,) are ignored. All that is left is the “injection effect.” This is an effect of prices haven’t adjusted yet leading the the offsetting increase in the nominal demand for credit and the nominal demand for money. This
effect (it seems to me) is temporary. And to the degree it does impact relative prices (real
interest rates) and the allocation of recourses (capital accumulation) it cannot last.
Anyway, if the government is trying to keep real interest rates low and capital accumulation high, inflation gets progressively worse. But NONE OF THIS APPLIES IF THE GOAL OF THE POLICY ISN”T TO MANIPULATE RELATIVE PRICES! If there is a money growth goal or an inflation target, it is possible that when it is introduced there would be injection effects and a distortion of the alloation of resources. Then, the policy just continues on. Any “malinvestments” get liquidated early on. The process by which the malivestments persist requires that the policy is aimed at maintaining them at the expense of worsening inflation.
For example, suppose the Fed made the health of the housing industry the goal of monetary policy and expanded the growth rate of the money supply until new home production matched the peak of the boom. My own view is that monetary policy should aim to return to the old growth path for nominal income. If that means that the housing industry must contract–so be it. But, there is this notion in Austrian circles that the whole point of the policy was to maintain housing production at high levels.
Again, because “superneutrality” is not true, it is possible that targetting seniorage, for example, would actually allow for an impact on the allocation of resources, including real interest rates and capital accumulation. But, Austrians typically ignore this process. These
are persistent, equilirium effects.
Economists like Cowen and Caplan (and me) were trained as Austrians, more or less, along the lines of Selgin, White, etc. One reason why these “Austrian skeptics” are skeptical is that
when you think about policy as above–money growth rule or price level targetting, then any malivestment from the termporary impact on the allocation of resources is due to entrepreneural error. Making investmnets based upon a termporary pattern of expenditure that only are profitable if it is permanent, would be a mistake.
Jon,
“There is a mapping of the Austrian argument to the notion that commodities are ‘immediate goods’ and finished goods are the fruits of an extended production cycle.”
I’m sorry but such arguments just make me angry – they are absolute nonsense. Some commodities have very long production cycles and some finished goods very short production cycles.
pointbite: If there is steady inflation, of (say) 10%, then every year people will need to get hold of 10% more money, just to keep their stock of money holdings constant in real terms. So, every percentage point increase in the growth rate of money supply translates into a percentage point increase in inflation, and a percentage point increase in the growth rate of nominal money demand. It’s permanent, not temporary.
reason: Yep. I should have added to exchange rate to the list of prices. The nominal exchange rate will be falling (depreciating) 10% faster in B than in A, but the real exchange rate (adjusted for inflation) will be the same in B as in A, so exports and imports will be the same in B as in A, etc. This is for small, medium, or large open economies.
M1, M2, M3, nominal debt, are all rising 10% faster in B than in A. The first guess is that real M1, M2, M3 and debt are the same in B and A.
But non-super-neutrality 1. says that real currency will be lower, so any monetary aggregate that includes currency will be lower in real terms. If there are reserve requirements against demand deposits, and those reserves do not pay interest, then inflation is also a tax on demand deposits, so real DD will also be lower. There will possibly be some spillover from the inflation tax on currency and reserves into debt as well, depending on the exact specifications of the model. But my view is that these effects will be small, at any reasonable inflation rates. And they would be swamped by other tax effects from a non-indexed tax system, like the ones that Alex and Scott emphasise above.
reason@4.56: Yes, as Adam notes, you are touching on some key questions here.
I have compared two parallel universes; one with a (say) 10% inflation target (B), and another with a (say) 0% inflation target. Depending on the shocks hitting them, and how the central bank responds, they could both be buzzing along quite smoothly, or both in a total mess. But my guess is that, in real terms, the two economies could be identical. Is there any reason to suppose my guess must be wrong?
My thought experiment totally ignores the question of the transition from one target rate of inflation to another. These are two parallel universes, that never meet. There is no fork in the road, where we decide which path to take. We have two parallel paths. Each economy has its own separate history. Comparative dynamics, if you like. Austrians focus a lot on that transition. Sometimes that makes it difficult to distinguish whether they are talking about transition effects or long run effects.
There are really two separate questions:
1. What is the best inflation target to have?
2. If the answer to 1 is different to the current inflation target, are the long run benefits of getting there bigger than the short run costs of the transition?
In my case, since I believe the current 2% target is roughly optimal, I am not worried about 2.
Adam: Yep. Agreed. Me too. I got an awful lot out of reading Hayek and Mises, and other Austrians.
Bill. Yep. Good comment.
If super-neutrality were true, it wouldn’t matter what inflation rate we targeted. And the attempt to use inflation to target anything real be end in disaster.
In my view, the main purpose of inflation (why I would advocate 2% rather than 0% or negative), is non-super-neutrality #3. I want to stay clear of the zero lower bound. All the other non-super-neutralities argue for as low a rate of inflation as possible. And #2 argues, as in Friedman’s optimum quantity of money, for deflation, so that nominal interest rates on govt bonds are zero (ignoring the real costs of paper and ink etc.)
Since we have already hit the zero lower bound with a 2% target, we certainly dare not lower the target. If anything, I would want to raise it a little. But I think that switching to a 2% price level path target could be enough to keep us off the lower bound in future. (That’s another story).
Jon,
the ECB has little to do with commodity inflation – are you suggesting that it should target import prices? Relative price changes happen, but as commodity prices usually eventually feed through into consumer prices, I don’t really understand what you are getting at here.
Nick, I was referring to the current situation. Do you think the current demand for money is permanent at these levels?
pointbite, do you think the current supply of money is permanent at these levels?
Nick, also regarding the question of steady inflation at 0% or something other number, I’m not sure the number is really the point (at least it’s not for me). The point is whether you trust the central bank or manage the economy effectively (which history proves nobody can do for very long) and whether you believe it’s “fair” that some people benefit from inflation while others suffer the consequences of inflation. Yes your point about about it not being much different from any redistributive tax is well taken, how is that a defense? It’s like the Republicans saying Lincoln suspended habeas corpus during the civil war, so we can do it now. Maybe it’s wrong in both cases? Or maybe the comparison is just different enough to be meaningful, at least with taxes the rules are transparent, with a managed inflation rate people with inside knowledge can benefit disproportionately. But not we’re now talking politics, not economics. So the difference (perhaps) is not an economic argument, it becomes more a morality argument. That’s why the feelings are so intense, in my opinion, on both sides. Everyone sees the other as stupid because HIS number add up, but that was never the debate.
Sorry for the million grammatical errors… if I sign in can I edit my comments?
Adam P, no of course not. Read my comment from 5:15 yesterday for context regarding that question.
Perhaps because I just read the Edward Barnays’ book “Propaganda” and another from a disgruntled school teacher ranting about the evils of forced schooling, but I’m increasingly becoming concerned about this era of PHD’s scientifically managing the population with continuously refined (ie. flawed) models. Even if people think it’s for our own good. That doesn’t mean it can’t work, or won’t work consistently for short periods of time (I’ve read too many behavioral economics and psychology books to have much respect for the rationality of average people) but I can’t stand it nevertheless. Ignorance may be bliss, but that doesn’t mean you won’t get angry when the deception is revealed. You can walk away from a car dealing feeling good about your purchase, but if you figure out the tricks they used to influence your decisions it may still make you angry after the fact. This is probably the wrong place for this rant, but here goes — the “better” economists (or perhaps just the ones who perceive themselves as the “better” economists) will acknowledge publicly they never have complete confidence in their models, but nevertheless recommend (with condescension) sweeping policy changes that affect millions of people — on want amounts to little more than an educated guess. What should a non professional economist make of such things? Am I a guinea pig to be experimented on? Imagine if engineers built skyscrapers that way. It bothers me. The appeal of Austrian economics to non-economists, I believe, is the political recognition of the fallibility of managers, it’s more honest to say “we don’t know so we won’t impose a decision on your behalf”. Sometimes I wonder if that’s why mainstream economist are conditioned (in school) to hate them, because it makes their profession less “sexy” and elitist. I’m not surprised that people are surprised by what Austrians actually believe when they actually read their material, that’s quite revealing in itself.
(duck)
pointbite: I think the demand for money will fall as nominal interest rates rise and fear lessens, then slowly begin to rise again as real income growth and a rising price level returns. A lot of empirical evidence (and theory) of the demand for money backs that prediction. But there might be technological changes in banking and payments systems that could upset it.
On the more general point: whether we like it or not the Bank of Canada is in charge of monetary policy. (In my view, they have generally done a good job, and I would need convincing that the job could be done better by some other body.) The question I focus on is whether we want the Bank of Canada to continue to try to hit 2% inflation, or some other number. If we choose 0% inflation instead, there will be a lot of large and arbitrary redistributions of wealth in the transition, but in the steady state the only result will be a reduction in the approx $2 billion tax on people who hold cash (which will require other taxes to be increased to offset them).
I don’t like that $2 billion tax on currency. It’s probably slightly regressive, if we assume, reasonably, that the poor have a higher currency/income ratio than the rich. It’s also an inefficient tax, and creates bigger distortions per unit of revenue than an equivalent income tax or GST. But it is a way to tax illegal activity, and tax evaders. In any case, a $2 billion tax is small beer, and I don’t lose a lot of sleep over it. If that’s the price we have to pay to reduce other costs of non-super neutrality, it’s fairly cheap.
My more general response to your “rant” is to mostly agree. That’s one of the reasons we want to leave decisions as much as possible to individuals, so they can make their own best guess, and take responsibility for their own successes and failures.
But if the government is in the business of producing money, and it is, then someone has to decide how much money it should produce. We can’t duck this question. Yep, we (monetary economists) don’t know as much about this as we should. But the thing that always restores my faith in what we do know is listening to non-economists. They (generally) know even less than we do!
Wow, not a single projectile! Now I know why I keep coming back.
“But the thing that always restores my faith in what we do know is listening to non-economists. They (generally) know even less than we do!”
Hmm… you know this reminds me of something, I think it was a section of “Blink” by Malcolm Gladwell, in which he discusses the problems associated with too much knowledge and how sometimes it leads us to ignore our intuition and make bad decisions. The book wasn’t as interesting as his first, but it was still worth the read.
And on the more general — this is how the system works whether we like it not — point, systems can change. The problem with your position is that we will never know if something else could have worked better. And the reality, unfortunately, is that Canada is such a small country changing our system probably wouldn’t matter that much in the aggregate anyway.
I support targeting a growth path for nominal income that increases 3%. And so, zero inflation in the long run.
I think we have already shown that accepting inflation to guarantee the wonders of conventional monetary policy (the great moderation) was an error. Using open market operations to target the overnight lending rate is something that cannot be depended upon. Sometimes the central bank needs to purchase longer term and riskier assets and not worry what is happening to overnight lending rates.
And so, I don’t think protecting short term interest rate targeting is worth accepting any inflation at all.
Fundamentally, I think there is a problem in treating tangible, zero-interest currency as the epitomy of money. Instead, deposit accounts with positive or, possibly, negative nominal rates should count as the core idea of money. Then, perhaps, issuing zero interest currency for small, hand-to-hand transactions, should be understood as a minor appendage.
Think about that for a while. Think about how many payments are made by transfer of deposit. Think about how central banks have for years passively adjusted currency to meet the demands of trade.
Embrace it. Long ago, money was gold. Then the tangible paper was added to the gold. And finally, those accounting entries at banks were included too. Gold is gone. Paper exists. But stop letting the tail wag the dog.
Anyway, it is true that if currency is issued, and it has zero interest (or at least there is no way to charge for its use with “negative” interest,) and it is to be kept at par, then it is the quantity that must be adjusted to meet demand. And, if other assets that have similar risk/maturity characteristics (like T-bills) would otherwise have a negative nominal yield, but only don’t because poeple can hold this zero interest currency, then issuing currency in sufficiently large quantities will be unprofitable. The issuer is going to have to do intermediation that no one else wants to– issue something with short maturity and “no” risk (the currency) while holding a portfolio of with more interest rate or credit risk. And maybe they will have to do a lot of that to keep the price level (or nominal income) on target.
Personally, I favor private currency issue. This would mean that currency wouldn’t be an especially good credit risk. And if it became unprofitable to issue, then there would be currency shortages. But that wouldn’t spill over into a shortage of the medium of exchange generally. Perhaps it is just too hard to imagine, but there would be no redeemability of bank deposits into banknotes unless the banks found it profitable to do. And so, there could be shortages of hand to hand currency, but people would still be able to buy and sell using deposit accounts. (Redeemability would still apply through interbank clearings, perhaps using deposit balances at the central bank, but, like other deposit balances, they could have positive or negative interest rates.)
Mises clearly understands the distinction between real and nominal rates. I grant you that he does not use the language explicitly but considering the text is dated 1912, I’m willing to infer those distinctions from the logic of what he is saying rather than insist he qualify each use of the term.
Considering that Hayek won a nobel prize and is frequently cited as the leading economist of the 20th century… one expects his formulations to be largely part of the modern orthodoxy.
As Nick rightly suggested; its too harsh to say that the Austrians lost. Hayek’s nobel traces the lineage of its thought to the Austrians, but that sort of thing is not thought as ‘Austrian economics’ but rather just ‘economics’
I think I made a fairly weak statement ‘some-mapping’, so don’t be angry but I do concede to the broad-brush charge. IMO, you do touch on something which that I think the Austrian argument is poorly distilled as to which prices are which.
The abstract notions they present make sense but I sometimes struggle to classify–perhaps because its a dichotomous classification of a continuous variable…
Nonetheless, let confront what you say directly. Most metals and fuels are indeed fruits of lengthy production cycles but they are also inputs to even longer ones.