Response to Sumner on Keynes and Market Monetarism

Yesterday Scott Sumner made the claim that market monetarists are the true heirs of Keynes. Sumner is a terrific scholar who I am much indebted to for my own intellectual growth and the trajectory of my own research. I do not believe that his claim is  altogether unfounded, but I do believe that it is contradicted by intellectual history. I commented that his claim was not exactly true because 1) Hawtrey was the first to formulate the relationship between deficiencies in aggregate demand and money and 2) Keynes posited that, in order to be effective, the increase in the money stock must be channeled through the labor market via fiscal expansion, at least in times of depression. Keynes's support of fiscal expansion as necessary to make monetary policy effective, which,as I have discussed previously, was attacked by Hawtrey, fundamentally shifts him away from the tradition of the market monetarists. As I noted in my comment, market monetarists are the heirs of Hawtrey, not Keynes.

Consider this passage from a lecture by Keynes in 1931 (quoted by Don Patinkin):

I am not confident, however, that on this occasion money the cheap money phase will be sufficient by itself to bring about an adequate recovery of new investment. Cheap money means that the riskless or supposedly riskless, rate of interest will be low. But actual enterprise always involves some degree of risk. It may still be the case that the lender, with his confidence shattered by his experiences, will continue to ask for new enterprise rates of interest which the borrower cannot expect to earn. Indeed this was already the case in the moderately cheap money phase which preceded the financial crisis of last autumn.

If this proves to be so, there will be no means of escape from prolonged and perhaps interminable depression except by direct State intervention to promote and subsidize new investment. Formerly there was no expenditure out of the proceeds of borrowing, which it was though proper for the state to incur, except war. In the past, therefore, we have not infrequently had to wait for a war to terminate a major depression. I hope that in the future we shall not adhere to this purist financial attitude, and that we shall be ready to spend on the enterprises of peace what the financial maxims of the past would only allow us to spend on the devastations of war. At any rate I predict with an assured confidence that the only way out is for us to discover some object which is admitted even by the deadheads to be a legitimate excuse for largely increasing the expenditure of someone on something! (211)

One of these deadheads was Hawtrey, with whom Keynes clashed at the Macmillan Commission. Hawtrey wrote an article decrying this position before the Keynes wrote his General Theory. 

Keynes believed monetary policy to be inadequate to lift the economy from deep depression. He was not referring to only the Great Depression amidst which he was writing, but depressions more generally as he said explicitly that "we have not infrequently had to wait for a war to terminate a major depression." In other words, Keynes believed that his posited economic scenario was relatively common. This aligns with much of what he wrote in The General Theory.

Market monetarists like Scott Sumner - in some ways I consider myself in this camp -  promote NGDP targeting because it can offset deficiencies in aggregate demand. Given that Keynes believed that this policy could not successfully lift the economy out of depression, precisely the time when it would be most useful, I find any claim to the Keynesian legacy far-reaching. Most of what was Keynes said that was useful had already been stated by Hawtrey long before Keynes said it (for example, see Glasner's working paper ).

I should close by noting that any claim to a Keynes's legacy is difficult to substantiate because Keynes's ideas across his academic career were not wholly consistent. This is typically true of anyone. But if Keynes's legacy is wrapped up in his General Theory, which is no great claim to make, then it is appropriate to accept the doctrine set forth in that work as most accurately representing his legacy. I would say much the same for Hayek at the end of his career with regard to his work on social orders and currencies. Sumner is right that market monetarists have some agreement with Keynes, and even much overlap, but I believe that the few discrepancies between Hawtrey and Keynes left a chasm between the programs of these two thinkers. This gap leaves market monetarism squarely within the tradition of Hawtrey contra Keynes.

Late Thought: Scott definitely agrees that Hawtrey is a better representative of market monetarism. 1359 EST

Right or Wrong on Cantillon Effect, Richman is Right

Over the last week there has been a flurry of exchange between Austrian economists Murphy and Horwitz and market monetarists Sumner, Rowe, and Woolsey over the Cantillon Effect as mentioned in Sheldon Richman's article in The American Conservative. The good news is that progress has been made in the argument . The Austrians have done a good job of recognizing certain deficiencies in the simple, vulgar reading of the Cantillon Effect, and some of the opposition has acknowledged the refinement of the position. I'm specifically referring to one response from Sumner and the post that I linked to for Woolsey. This is a postitive development, but it has overshadowed Richman's main argument: elites use government to increase their own wealth and power and in the process make class structure more rigid.

Richman concentrates on government as a mechanism for wealth reallocation. Such involvement opens up substantial rent-seeking opportunities. As he describes in his article, "It was these [19th century, French] laissez faire radicals who pointed out that two more or less rigid classes arise as soon at the state starts distributing the fruits of labor through taxation: taxpayers and tax-consumers." While rents can be distributed directly - i.e., handouts gathered through taxation - they can also be distributed indirectly through inflation and the ensuing relative price changes. Investors who correctly hedge against expected inflation receive a portion of these rents. In this way, the Cantillon Effect is pertinent to the discussion. So far, the debate has done little to address this point.

Within a market, actors have incentives to increase their wealth holdings. This is complicated when government interferes through monetary manipulation, trade regulations, granting of labor monopolies, and any other action that transfers wealth from group A to group B by the use or threat of force. Impediments to collective action among large groups provide an opportunity for small groups to engage in rent-seeking behavior. Whether or not intentional, this leads to oppression of the poor by the rich. I need only cite a few examples. Young and unskilled workers - often those who are from the most disadvantaged of social groups - are kept out of the work force by labor union supported minimum wages and other labor market restrictions. Recently, a large number of middle and lower class Americans were economically crippled when, after buying houses that were overpriced due to politically driven incentives, the housing market collapsed. And, as Richman points out, wealth is redistributed from the poor to the rich due to inflation as the wealthy "are far better positioned to hedge and recover than workers who are laid off from their jobs." Some of the elite might face major losses as the effects of the policies work against their interests, but the general effect is to disempower and even confuse ordinary men and woman as the wealthy collect economic and political power by the generation of these crises small and large.

There is an important question that needs to be considered in analyzing the redistribution of wealth and power. How much of this is intentional? Are the wealthy overtly attempting to politically and economically castrate the middle and lower classes? There might be some elites out there that think in these terms, but I am inclined to take into account the incentives of the system rather than begin with accusations of conspiracy. If social systems encourage voluntary exchange, parties will attempt to better themselves through voluntary exchange. If it encourages the use of force to redistribute wealth, individuals will attempt to gain by taking advantage of government coordinated wealth transfers. The latter implies the use of force and is, is at best a zero sum situation. More often, the costs of the transfer lead to a net loss of wealth.  The more powerful the government, the greater the incentive for individuals and organizations to use the latter strategy to accumulate wealth and power. This is the message that has been overlooked. This is a cry against tyranny that should not be ignored. I am glad that the aforementioned economists have partaken in a fruitful discussion concerning the Cantillon Effect, but lets not forgot the context within which it was mentioned.

Sumner and Christensen Harangue Incompetent Technocrats

At EconLog, Scott Sumner is being witty. Not only did he title his post, Governments don't create problems, they solve them. He closes with this:

For you math jocks, here is a mathematical translation of the absurdity:

NGDP = M*V
NGDP = C + I + G + NX

Increases in the money supply obviously cannot boost NGDP, because the real problem is excessively low consumption, investment, government spending, and net exports. Don't think top government officials would ever be so blind as to think this way? Think again.

PS. Here's the equation that they should put in principles textbooks:

M*V = C + I + G + NX

It will never happen---as it might get students thinking dangerous thoughts.

The post stems from a comment about low inflation in Europe and the inability of central bank officials to take responsibility for the effects of their policies. Lars Christensen covers the same point today. 

Scott Sumner On Kahan Tea Party Intelligence Distribution

Insightful. Check it out.

I apologize for picking on Dan Kahan, because he seems like a good guy.  And he’s no worse than the typical Yale academic.  But he really should be embarrassed. How could an academic expect people who identify with the Tea Party to be below average in any sort of intelligence/education metric? It boggles the mind. 

Average people pay little attention to public affairs.  Following public policy is not normal behavior; it’s what smart people do.  People like to talk about how popular Fox News is, but compare its ratings to professional wrestling, or some other non-intellectual show.  You will be surprised by how few people watch Fox.  The only reason the Tea Party didn’t do better is that the group included those who merely sympathize–if you took actual members the score would have been far higher.

Hawtrey on the Weakness of the Gold Standard: Gold, per se, was not the Problem

As I continue my study of the Ralph Hawtrey's analysis of the classical gold standard in The Gold Standard in Theory and Practice, I notice that he sets forward in his narrative an argument that implies the problem that I am currently extrapolating upon in an upcoming paper. (I hope to get it up on SSRN in the next week or two.) He writes:

The immediate effect of the suspension of the free coinage of silver in Europe was to concentrate the whole demand for additional metallic currency upon the gold supply of the world.

As I have argued before, the establishment of the gold standard eliminated metallic substitutes for gold as base money. By definition, this made demand for gold more inelastic, thus creating an environment that encouraged price volatility.

As the price of any good becomes more expensive, individuals tend to substitute away from it. By preventing the employment of substitutes for gold, gold standard countries – meaning, in practice, all western nations after 1879 – made more fragile the international monetary system. This is a fact too little appreciated in the literature concerning the gold standard - with the exception research from Bordo and Reddish that I posted recently, and probably David Glasner, Scott Sumner, and other interested market monetarists. The restraint provided by the classical gold standard appears to garner support for it among libertarian leaning economists. As a result, the effects of intervention in the classical gold standard have gone on generally ignored as arguments concerning it have become polarized. i.e., in debate gold becomes either a barbarous relic or a beacon of growth and economic stability. 

Researchers should be asking: “How did the gold standard change when silver was demonetized?” and “Why did it fail?” Hawtrey baldly explains the problem:

Since there is nothing in the circumstances of either metal [gold or silver] to make it more stable in value than the other, are we to be driven to the conclusion that the precious metals are inherently defective for that purpose? That would be a mistake. The true moral of the nineteenth-century monetary experience is rather that the defects in gold and silver as standards of value have been attributed to causes within human control. Governments have been too prone to modify their currency systems without regard to the reactions they might cause in the world markets for the precious metals, and therefore in the currency systems of their neighbors.

Conflicting policies from independent central banks destroyed the stability provided by the gold standard. Hawtrey preferred that central banks might cooperate to avoid the problems associated with the monometallic standard, but such hopes were dashed by political reality.

Arguments concerning the gold standard in history too often devolve into a fight about the merits of the gold standard per se. Consider George Selgin’s “The Rise and Fall of the Gold Standard in the United States” (which, despite my qualms, I still recommend for anyone attempting to gain familiarity with the gold standard. His discussion of silver demonetization quite informative!). Although he clarifies that a gold standard does not depend on “’legal tender’ status”, the complications associated with the adoption of a monometallic standard under a legal tender regime breeds complications that are ignored. In defending the gold standard against the claim that it is inherently deflationary and therefore suppresses economic growth, he writes:

…actual statistics for the [deflationary] interval in question reveal healthy average growth rates for both total and per capita real income … with declining prices reflecting, not flagging demand (as they did in the 1930s) but robust growth.”

The gold standard itself was not itself exceptionally deflationary, but a monometallic regime enforced by law was deflationary. This was not due to an increase in production. The growth of the gold stock could not keep pace with demand for gold after silver was demonetized. Deflation was more a result of the elimination of metallic substitutes than of increases in production. (See my earlier post.) It is for this reason that we see a strong downtrend in gold denominated prices between 1873, around the time that most major nations demonetized silver, and 1896.

Surely the debate can be improved. If the gold standard was destroyed by “causes within human control,” by governments that were “too prone to modify their currency systems without regard to the reactions they might cause,” then the interesting story to be told concerns political economy. In this story, the gold standard is more of a bystander than a system of instability or inherent promoter of deflation. 

While researchers like Barry Eichengreen and Peter Temin suggest that the gold standard was overly constrained monetary policy, I suggest that the classical gold standard overly constrained markets. The limitations of a monometallic legal tender monopoly impeded the formation of expectations in regard to future prices as substitution away from gold could no longer limit swings in prices. (For insight, see Barsky and De Long on inflation expectations under the classical gold standard.) If we are to discuss the gold standard, we must first ask "before or after silver was demonetized?"

Central Banks and the Interwar Gold Standard

I just uploaded my paper to SSRN:

In the last several decades David Glasner, Douglas Irwin, Ronald Batchelder, and Scott Sumner have revived Hawtrey and Cassel’s explanation of the Great Depression. According to them, the intensity of the Great Depression can be explained by a dramatic increase in demand for gold by central banks, principally in the United States and France, which forced down prices internationally. This paper expands their analysis with a model to test the impact of changes in aggregate gold reserves on the gold price level throughout much of the interwar era.

Two Roads?: Theoretical Case and Historical Precedent for a Nominal Anchor (Part IV)

We have left to consider the a defense of an endogenous fiat money base. Like the gold standard, but more responsive, a nominal income standard will alleviate excess demand for money. Sumner argues that nominal income targeting provides a nominal anchor, meaning that it prevents dramatic swings in prices that would otherwise result from an unstable demand for money (2012, 152). He notes that “the current price level and current NGDP are far more affected by the future expected money supply than they are by the current money supply (144).” If prices reflect expectations, then it is critical that growth rate of one of the major determinants of prices, the money stock, be predictable. Much in the way that “dynamic equilibrium requires consistency of plans which . . . depends on a flexible price system,” a rule that makes the expansion of the stock of base money stock both responsive to prices and predictable will help facilitate coordination of plans among economic actors (Lachmann 1978, 116). As convergent expectations are requisite for Say’s principle to hold, and therefore, for markets to clear, a clearly defined rule that predictably governs growth of the base money stock will reduce policy uncertainty and thereby increase the efficacy of the price system to convey tacit information (Hayek 1943).

The benefits of a nominal income level target are especially of significance for the loanable funds market as it should stabilize inflation expectations. By this effect, it will also better enable credit markets to clear as interest rates, the price of money in the future, will not suffer from distortions stemming from nominal factors such as volatile demand for base money. A nominal income target can also ensure that the rate of inflation remain positive. The significance of this becomes clear upon considering a scenario where the real rate, r, is less than the rate of deflation, -π. The Fisher equation denotes that in the long run the nominal – observed – interest rate as the sum of the real interest rate and the inflation rate. This is formally stated as:

i = r + π

We can imagine a scenario where the market clearing nominal rate of interest is negative. That is the abovementioned case where the real interest rate is less than the rate of deflation:

               r < -π

This special case was recognized in 1913 by Ralph Hawtrey:

What if the rate of depreciation of prices is actually greater than the natural rate of interest? If that is so nothing that the bankers can do will make borrowing sufficiently attractive. Business will be revolving in a vicious circle; the dealers unwilling to buy in a falling market, the manufacturers unable to maintain their output in face of ever-diminishing order, dealers and manufacturers alike cutting don their borrowings in proportion to the decline of business, demand falling in proportion to the shrinkage in credit money and with the falling demand, the dealers more unwilling to buy than ever. (186)

A reduction in the quantity of credit demanded leads to a surplus of savings. If, due to an abnormally high rate of deflation, the equilibrium nominal rate necessary to clear the loanable funds market is negative, then this surplus is inevitable. Since credit doubles as money, a fall in credit outstanding leads to an equivalent drop in total output. A nominal income target stops this problem in its tracks and will allow markets to more efficiently liquidate inventory. Instead of a fall in output, markets will respond by more quickly reallocating previously overvalued goods as a general expansion of the money stock is unlikely to save from bankruptcy those firms that made the most egregious mistakes during the boom. A base money stock that responds to changes in demand for money will help facilitate the reallocation of resource toward their highest valued use. Compare this to a general fall in prices whose end date is unknown.

The danger of tremendous deflation is not a new concern as bankers have long been concerned about financial destabilization due to deflation. Even before the establishment of the Federal Reserve, private banks innovated the means to mitigate the damaging effects of an elevated demand for money during recessions and depressions. Banks would issue temporary currency to stem a deflationary impulse (Timberlake 1984, 6). Money demand shocks were endogenously offset by temporary increases in the money stock. Those who issued the currency provided it to banks that appeared to be illiquid, but were unavailable to those banks that were clearly insolvent (7). Like endogenous gold and credit stocks, the increase in temporary currency arose due to the deflationary impulse set off by a sudden increase in demand for money. This temporary currency was elastic enough to prevent a crisis from turning systemic. Absent activist central bank policies, the gold standard, was able to serve as a nominal anchor. Only as a result of gold hoarding policies from the Bank of France and the Federal Reserve did the gold standard prevent a runaway deflation like occurred due to central bank intervention in 1929 (Eichengreen 1996; Irwin 2012). It does not take much imagination to see that a nominal income level target will serve a similar role to what temporary currency played in moderating crises.

Glasner Reviews my Paper on the Interwar Gold Standard and Central Bank Gold Demand

See it here. And if you haven't already, check out the paper here.

I was pleasantly surprised to receive an email a couple of weeks ago from someone I don’t know, a graduate student in economics at George Mason University, James Caton. He sent me a link to a paper (“Good as Gold?: A Quantitative Analysis of Hawtrey and Cassel’s Theory of Gold Demand and the Gold Price Level During the Interwar Period”) that he recently posted on SSRN. Caton was kind enough to credit me and my co-author Ron Batchelder, as well as Doug Irwin (here and here) and Scott Sumner, for reviving interest in the seminal work of Ralph Hawtrey and Gustav Cassel on the interwar gold standard and the key role in causing the Great Depression played by the process of restoring the gold standard after it had been effectively suspended after World War I began.

Krugman, Gold, and 1 Trillion Dollars (Lift Pinky to Corner of Lip)

Paul Krugman posted on the trillion dollar coin and used the opportunity to talk about the "barbarous relic" and the benefits of state-managed money:

For people like me, on the other hand, the economy is a social system, created by and for people. Money is a social contrivance and convenience that makes this social system work better — and should be adjusted, both in quantity and in characteristics, whenever there is compelling evidence that this would lead to better outcomes. It often makes sense to put constraints on our actions, e.g. by pegging to another currency or granting the central bank a high degree of independence, but these are things done for operational convenience or to improve policy credibility, not moral commitments — and they are always up for reconsideration when circumstances change.

Now, the money morality types try to have it both ways; they want us to believe that monetary blasphemy will produce disastrous results in practical terms too. But events have proved them wrong.

There are better arguments for Krugman to confront than that of the "money morality types." Most free market economists don't wish for a state-managed gold standard. Even Hayek was weary of a wholesale return to gold by central banks because of the potential of discoordination (see Prices and Production). The danger of a return of state managed gold standards would inherently bring volatility in the purchasing power of gold unless central banks could coordinate their monetary policies, which is unlikely. Serious economists do not suggest a return to this gold standard. They do not propose that gold has intrinsic value. Krugman is confronting a political argument, not a legitimate argument from economists.

Krugman reveals the crux of the argument concerning money when he refers to changes in the money stock. But the question to confront is not whether the "quantity and characteristics" of money "should be adjusted" to suit economic circumstances, but by what mechanism this should be done. Free banking theorists suggest that the market can handle these adjustments either in whole or in part. Some like Scott Sumner, Nick Rowe, and David Glasner stress income targeting by central banks and argue that this has a stabilizing effect. Glasner especially stresses the endogeneity of the broader money stock and draws the conclusion that central banks simply need to adjust the monetary base to target nominal wages (similar to NGDPLT - to understand the nuance read this). Others like White and Selgin theorize about competitive banking systems where the unit of account is determined by the market and outside money is commodity money (for a nice theory of the evolution money, whether managed by markets or the state, read this). The latter free banking theorists consider gold, and sometimes silver, as money in their analysis because these commodities have historically been chosen by the market as outside money and the unit of account. It is not inconceivable that the market would once again adopt gold as the unit of account. It could serve this purpose as other commodities and assets serve as a store of wealth (probably via ETFs and other instruments), and thereby alleviate upward pressure on the purchasing power of gold. It is also possible that the market uses something other than gold as the unit of account.The point is that the market process can work out the details.

Krugman's argument assumes that the unit of account is managed by the state. This assumption is not untenable under current circumstances, but if he is going to consider an alternative like a gold standard, he ought to confront a robust theory, not a political talking point.  Markets, by trial and error, can determine the unit of account and the quantity of outside and inside money endogenously. While some unsophisticated commentators might suggest that nations return to the gold standard, this is not the proposition of serious theorists.

*Late thought*

I would like to see Krugman discuss the interrelatedness of fiscal and monetary policy and consider the impact of this on central bank policy and macroeconomic phenomena. I'll have to save these thoughts for a later post.