Arnold Kling  

The Wisdom of CPK

Hummel on Krugman, Keynes, and... Question for Conservatives...

If I were writing a blurb for the fifth edition of Charles P. Kindleberger's classic Manias, Panics, and Crashes, I would say that "This is the best narrative that has been written about the 2008 financial crisis, and it was written in 2005."

Of course, that might be superseded by a blurb that reads, "This is the best narrative that has been written about the collapse of the eurozone." Stay tuned.

I'm now 1/3 of my way through reading this edition (I read the first edition back in the late 1970s). There is much here, including an opening to re-engage in a debate with Scott Sumner, which I will put below the fold.

p. 14:

When asset prices tumble sharply, the surge in the demand for liquidity may drive many individuals and firms into bankruptcy, and the sale of assets in these distressed circumstances may induce further declines in asset prices. At such times a lender of last resort can provide financial stability or attenuate financial instability. The dilemma is that if investors knew in advance that government support would be forthcoming under generous dispensations when asset prices fall sharply, markets might break down somewhat more frequently because investors will be less cautious in their purchases of asset and of securities.

This succinctly anticipates the intellectual debate about TARP. I am quite sure that CPK would come out on the "pro" side. I was always on the "anti" side and remain so, but I still respect CPK's wisdom.

p. 15:

--at least one monetarist view--is that the mania would not occur if the rate of growth of the money supply were stabilized or constant. Many of the manias are associated with the surge in the growth of credit, but some are not. a constant money supply growth rate might reduce the frequency of manias but is unlikely to consign them to the dustbins of history. The rate of increase in U.S. stock prices in the second half of the 1920s was exceptionally high relative to the rate of growth of the money supply, and the second half of the 1990s

p. 18:

When government produces one quantity of the public good, money, the public may proceed to produce many close subsidies for money...The evolution of money from coins to bank notes, bills of exchange, bank deposits and finance paper illustrates this point.

And I would add the evolution of what people refer to as the "shadow banking system."

Another passage on this point, from p. 70

The stylized historical fact is that every time the monetary authorities stabilize or control some quantity of money, M, either in absolute volume or at a predetermined rate of growth, more of the money and near-money substitutes will be produced in periods of euphoria.

On this issue, I am with CPK. I remain loyal to the MIT view, which is that the textbook model of the money supply founders when confronted by a real world of financial innovation. CPK summons a roster of famous classical economists in support of the MIT view. On p. 71, he gives us John Stuart Mill:

The purchasing power of any individual...consists, first, of the money in his possession; secondly, of the money at his banker's, and all the other money due him and payable on demand; thirdly of whatever credit he happens to possess.

Now, Scott Sumner has an answer for the MIT view, which is to have the central bank focus on nominal GDP. The central bank does not need to be able to define M; it merely has to know the difference between actions that are expansionary and actions that are contractionary, and move in the appropriate direction when forecast nominal GDP deviates from target.

CPK prefers the "credit" view to the "money" view. (See the tenth of my macro lectures). He would call these the "banking" view and the "currency" view, respectively.

The "credit" view is that all sorts of credit instruments matter in the economy. Moreover, the central bank has little control over how the use of credit expands or contracts. But CPK never had to suffer with Scott Sumner rubbing his sandpaper over your arguments, pointing out that nominal GDP is nominal GDP, and who cares about what credit instruments the market uses?

My response is in terms of PSST. Credit expansions allow unsustainable patterns of specialization and trade to develop. Any entrepreneur who can borrow money can stay in business, even if the business is unsound. Or, anybody who can refinance a house at a higher appraised price can keep the house, even if they cannot make the payments. When credit collapses, the entire trade pattern has to be reconfigured.

But none of this gives me a solid argument for why the central bank cannot hit a GDP target. Yet, as Sumner would point out, our central bank can't, or won't (he would say won't). The case for PSST would be a lot cleaner if nominal GDP were on target and we still had ridiculously high unemployment.

Combining the credit view with PSST comes close to the standard Austrian story. But it is also close to CPK's exposition of the Minsky story. In fact, CPK is a better exponent of the Minsky model than Minsky, and there are those who will speak of the Kindleberger/Minsky model.

For a monetarist, steady monetary growth will lead to steady economic growth, forever. For CPK and Minsky, steady economic growth will eventually give way to euphoria, fueled by credit innovation that outmaneuvers the steady monetary growth. Hence, the Minsky epigram "stability leads to instability."

p. 83:

The paradox is that the role of the instability of credit began to be neglected about the time of the Great Depression of the 1930s.

The lead-up to that sentence is that folks like Mill, Marshall, Pigou, and Fisher had talked about credit, but that talk died down in the 1930s. I am not sure how Keynes fits in. Relative to the flame of the "credit view," was Keynes a keeper or a douser? Hicks was definitely a douser, I would say.

As Scott Sumner has pointed out, the relationship between the intellectual cycle and the business cycle is interesting. CPK points out that some economists reacted to the Great Depression by arguing for radical moves to curtail risk in banking. For example, Henry Simons proposed 100 percent reserve requirements. p. 86:

Simons advocated a system in which all financial wealth would be held in equity form with no fixed money contracts so that no institution that was not a bank could create effective money substitutes.

This is an extreme form of what I call the "easy to fix rather than hard to break" approach to financial reform. In engineering terms, equity degrades more gracefully than debt. I would like to encourage shifts away from debt finance by removing all of the tax and regulatory provisions that today "nudge" in the opposite direction: the mortgage interest deduction, other mortgage subsidies, and the corporate income tax are the chief culprits I have in mind.

Probably more CPK wisdom to come as I delve further.

COMMENTS (10 to date)
david nh writes:

Perhaps you know this already but a sixth edition is due out 5 Aug 2011.

Lewis writes:

Fireworks. I grew up in Alabama, where nearly all fireworks are legal. Now I live in Pennsylvania, where only the very lamest fireworks are legal. What bothers me is the inconsistency of gun rights and firework rights. If I buy a very small amount of explosive powder in a paper packaging, it's probably illegal. But if I buy a stronger powder and a metal machine designed to kill people/animals, it's definitely legal and even encouraged by many conservatives.

For consistency's sake, legalize fireworks. I hate hearing the arguments against fireworks: they shoot eyes out. Guns kill children every day. So do cars.

Lewis writes:

For some reason my fireworks comment appeared on this post instead of Caplan's post about things we would like to see legalized. Looks kind of funny hehe.

Scott Sumner writes:

A few comments:

1. I'd guess Simons was trying to come up with a plan that would be consistent with a gold standard regime (or Bretton Woods.) In that case you need to stabilize money demand, which requires lots of regulation. Under a fiat regime we don't face that problem.

2. It's possible that even with steady NGDP growth you could have the sort of crisis that Kindleberger discusses. But 1987 was about as close to a controlled experiment as we'll ever see. Let's review the facts:

a. The 1920s saw a big run-up in stock prices, followed by a 40% crash in Sept/Oct 1929. Then NGDP fell 50% and Depression resulted.

b. The 1980s saw a big run-up in stock prices, followed by an almost identical crash to 1929 (in Sept/Oct 1987). This time the Fed kept NGDP growing at a steady rate. Not even an itsy-bitsy teeny-weeny slowdown followed this stock crash. I'd say that refutes Kindleberger and supports my views.

(Was that too much sandpaper?)

ThomasL writes:
When government produces one quantity of the public good, money, the public may proceed to produce many close subsidies for money...The evolution of money from coins to bank notes, bills of exchange, bank deposits and finance paper illustrates this point.

And I would add the evolution of what people refer to as the "shadow banking system."

This is very interesting. Is this something of an "endogenous money substitutes" viewpoint? Where the quantity of money proper is exogenous, but that the demand for additional money cannot be denied, and so substitutes are created endogenously? (ie, money+substitutes is endogenous, even if money itself is exogenous)

If so, what might this mean for inflation? The Fed has dramatically increased the money supply, but if the demand for money+substitutes has fallen, increasing the supply of money proper would serve primarily to change the ratio of money to substitutes. Only by driving out the demand for substitutes completely could you truly increase the money supply.

Related to inflation in a supply sense, would be what role the money substitutes have in prices. The more nearly perfect a substitute, the greater the role, I would think, but I'm not certain.

Estupidante writes:

You are coming close to one interesting point in your post - that Minsky/Kindelberger and Austrian view are very, very close to each other. The only slight, but decisive difference is that while Austrian think markets can self correct themselves and are not suspect to endogenous cycles, Minsky and Kindelberger claim oposite. Thus origins of cycles are either endogenous or exogenous (central bank). For me the this is most clear from Cooper's book Origins of Financial Crises.

Arnold Kling writes:

re Scott Sumner's point 2b. I think that CPK would say that the difference between 1929 and 1987 is that in 1929 you also had a real estate boom end. He points out that bad real estate loans are what killed the banks.

Shayne Cook writes:

Arnold, Scott:

"re Scott Sumner's point 2b. I think that CPK would say that the difference between 1929 and 1987 is that in 1929 you also had a real estate boom end. He points out that bad real estate loans are what killed the banks."

In addition to Arnold's point, it is my understanding that the 1920's stock market bid-up was largely debt financed, and less debt-financed (margin financed) in the 1980's bid-up. Is that not true?

Certainly there was a lesser understanding of the implications of debt versus equity financing in the 1920's than the 1980's, particularly as applied to investment asset pricing (homes or stocks, in these cases). The 1929 stock market crash, and an improved understanding (at the time) of the differences between debt and equity financing led to the Glass-Steagall separation of "Commercial" and "Investment" banks in 1933.

effem writes:

I believe actual events back up your PSST theory. From my perspective the US has already been running a lite version of NGDP targeting since Greenspan took the helm of the Fed by ensuring short recessions rather than true deflationary events. This falls short of full NGDP targeting but is pretty close.

In this environment the ideal business is one which: 1) has revenues that grow at the rate of NGDP, 2) has semi-fixed costs, and 3) can attract cheap leverage. In other words, banks are the ideal business in an NGDP-targeted world.

Not surprisingly, the financial sector has grown massively. NGDP-linked revenue, semi-fixed costs, and cheap leverage have created a perpetual money machine provided that there is never a deflationary event (government backstops make the case even stronger but are not strictly necessary).

So an NGDP-linked world would create an enormous finance industry until the point when something drastic means the NGDP target cannot be hit (political activism, world crisis, natural disaster, etc.). At which point the bloated banking system will blow up - wiping out all the "progress" from the NGDP-linked economy.

Is this not exactly what has happened? Seems like PSST at work to me. Sure, if we commit more fully to hitting NGDP we could perhaps delay the blow-up of the banking system but eventually something will force us to miss the target at which point the losses will be the mirror image of the "gains."

Scott Sumner writes:

Arnold, After 1987 a real estate boom also ended in some important parts of the US, such as Boston, where I was living at the time. Prices fell about 20% in Boston. And that was with Greenspan increasing NGDP at a brisk rate. Imagine what would have happened if Greenspan let NGDP fall in half after 1987!

The big surprise in the Great Depression is how little real estate prices fell, given the income of Americans fell in half. And how few big banks failed. That suggests there was no real estate bubble in the 1920s (except obviously the 1926 bubble in Florida.)

Comments for this entry have been closed
Return to top