Arnold Kling  

Scott Sumner Cries Out

Jim Hamilton Says He's Happy O... Morning Commentary...

He wants the Fed to do whatever it can to try to get prices headed up rather than down. His best argument:

What do we have to lose?: We can get rid of interest on bank reserves (and consider a penalty rate), set an explicit nominal target, and engage in quite substantial quantitative easing using indexed bonds (and perhaps a few foreign government bonds) without incurring much risk at all. And even if we have to eventually move more heavily into assets more exposed to U.S. inflation risk (long term T-bonds) I don't see how those risks are any worse that what we are now doing at the Fed. Isn't a risky policy that has a good chance to boost AD superior to a risky policy that has little chance of achieving that goal?

Pointer from Tyler Cowen, who is much more enthusiastic than I am about Sumner's ideas.

I don't buy the monetarist story here. Sumner's thesis is that causality runs from slow money growth to slow GDP growth to financial collapse. I believe the causality runs in exactly the opposite direction.

In the old days, a monetarist would pick a definition of the money supply, plot its growth rate, and predict GDP a few quarters ahead. Those of us who were skeptical were always on the lookout for folks who would change their choice of money measure every few years to demonstrate better fits with the data.

But the real problem came when the movements in money were contemporaneous with or lagged behind the movements in GDP. Milton Friedman famously said that there were "long and variable lags" between changes in money growth and changes in the growth rate of nominal GDP. I remember at some point one of his opponents (I want to say Franco Modigliani, but I would not swear to it) asked sarcastically if the long and variable lags had become long and variable leads? That latter quip is the one that I would throw back at Sumner.

I'm describing myself these days as the world's only libertarian Keynesian. In any case, I don't go for the notion that you can just pour money all over the economy and generate a recovery. I think that we have shifted out of a regime with a low risk premium and into a regime with a high risk premium. I think the only way out is for profits to increase, which would lead gradually to business expansion, rehiring, and more overall economic activity. See Would Keynes Have Supported the Stimulus Bill?

Comments and Sharing

COMMENTS (13 to date)
Troy Camplin writes:

"Sumner's thesis is that causality runs from slow money growth to slow GDP growth to financial collapse." This is the Keynesian approach of putting the inflationary cart before the GDP horse. Haven't we learned from this yet? It is economic growth that causes natural inflation, and economic contraction that causes natural deflation. Keynes saw the correlation, but reverse the causation. The result was the stagflation of the 70's that nearly crippled the U.S. and Western Europe's economies, and did cripple many economies around the world. The reason is that companies eventually realized the monetary signals were false, and reacted as though they were in a recession. Why has nobody recognized that the opposite of Keynes' formulation is true? If this is Sumner's thesis, be prepared for 70's style stagflation.

Why are these idiots in charge, but I can't get a job? That's what I can't understand.

Tyler Cowen writes:

I don't agree with Sumner's causal account either but nonetheless I think he has the best plan for reversing AD declines as one secondary effect of the ongoing crisis.

Greg Ransom writes:

If you read Roger Garrison, you'll realize that Friedman as a macroeconomist was at his core working with an essentially Keynesian model -- see Garrison's _Time or Money_ or his essay asking Was Friedman a Keynesian?

So you're in the Friedman tradition of "libertarian Keynesians".

I realize that post-Lucas Chicago macroeconomists are not working in the same paradigm as Friedman.

Bill Woolsey writes:


Sumner insists on claiming that the Fed "caused" the crisis by an act of ommission. He claims that the Fed failed to increase the money supply enough to offset the drop in velocity and so failed to keep nominal income on the proper growth path.

Sumner's view is that the Fed should ajust the money supply whatever amount needed to keep nominal income on some stable growth path. Failue to make the needed change in the money supply causes nominal income to deviate from its path.

He is claiming that this failure of the Fed to do its duty has created a worse financial crisis than before. He grants that velocity dropped because of the initial financial crisis. Then, the act of ommission "caused" the drop in nominal income, which created a further financial crisis.

Your discussion about the problems with forecasting nominal GDP using a past growth rates of the money supply is all beside the point. Whatever does that have to to with Sumner? Sumner isn't claiming that velocity is stable. He is saying that the Fed should change the money supply to offset changes in velocity and keep nominal income on target. He is not advocating a target for some measure of the money supply.

Anyway, if an increase in the economy risk premium has resulted in a decrease in velocity, this is irrelevant to Sumner's position. It doesn't matter why velocity fell. The Federal Reserve should (in his view) raise the money supply whatever amount needed to get nominal income back on target.

There are two sensible lines of criticism of Sumner:

1. Increases in the money supply cannot raise nominal income back to its previous growth path. Any effort to raise the money supply will cause velocity to fall even more, so that nominal income will remain below target. In other words, a liqidity trap.

2. Nominal income isn't important. The "high risk premium" economy will have less production and employment. Nominal income only impacts the price level. While the Fed could get nominal income back to its previous growth path, this will just mean a higher price level with output and employment continuing on the same depressed path they would have had anyway. In other words, "market clearing" and some version of real business cycle theory.

In reading your commentary on current events, I think you would continue to try to think about what would happen if nominal income stopped dropping and instead rose back to its initial growth path.

What would happen to consumption, investment, and profit? How would a high risk premium be reflected?

In my view, it will be eflected in less real risk. More consumption out of income. More investment out of profit. Less risky investment projects. In the long run, this would result in smaller increases production and smaller increases real income and so more inflation. But in the short run, there is no reason it should result in lower spending across the board.

Vasco writes:

I'm not going to comment on the substance of your post, because I'm not sure what to say, but I wish you'd stop referring to hypothesis you disagree with as "stories". It's just condescending and unbecoming of scientific debate.

Bill Woolsey writes:

One more note--

Charging banks to hold reserve balances at the Fed aligns very well with the notion that the problem is an increase in the risk premium. The Fed is providing banks with a low risk investment. Why not make them pay for the safe harbour?

The reduction in the willingness to take risk is refelected in a rush to T-bills, FDIC insured deposits, and balances held by banks at the Federal Reserve. There is more demand for these things than a willingness to supply them, so why shouldn't those who want to hold them and avoid risk pay?

If they don't want to pay, but rather earn, then they should hold riskier assets. BAA corporates are paying 8%.

If they don't want to pay to hold safe assets, and won't take risk to earn on risky assets, then they should consume.

Quantitative easing is going to be, sooner or later, the Fed providing safe assets for investors and taking the risk itself. It should work in the usual way, meeting the demand for money/offsetting the drop in velocity. But it does also involve the Fed taking risk for people who want safety. Hopefully, if the Fed doesn't immediately jump to the riskiest assets, the result will be shift in production to lower risk activities.

Charging banks for holding reserve balances (and perhaps generating slightly negative yields on FDIC insured deposits and T-bills) has a similar effect, while reducing the amount of risk the Fed must take. Rather that having the Fed increase the quantity of money to offset the lower velocity, it lowers the demand for money/raises velocity.

In the end, people may accumulate zero interest currency and the Fed must accomodate the demand, and accumulate risky and longer term assets anyway. But charging banks for holding reserve balances is a good start.

And paying banks not to hold reserves, that is, to not make loans, is a travesty.

Niccolo writes:

Yeah, not particularly convinced that a fall in the money supply was ever really the problem in the first place.

Nathan Smith writes:

For better or worse, the means of exchange is a government monopoly.

The value of money should be stable. To keep it so is the Fed's job.

In the past year, house prices have fallen. This is deflation, although it doesn't get called that because of the way the Consumer Price Index is constructed.

When deflation occurs, some people who bought homes on credit may find that they owe more than the homes that serve as collateral for their mortgage loans are worth. So they have an incentive to walk away.

Thus a monetary problem led directly to an institutional, law-and-order problem.

Raise house prices, and either people will pay their mortgages, or banks can foreclose and recover what is owed them. Either way the toxic assets are revalued, and the banks are saved. This can be accomplished through inflation by quantitative easing. If the banks are too shell-shocked to act as effective conduits for money creation, give the money directly to the people. Crisis goes away.

Then change the price index for good, make it the Fed's business to keep house prices stable (no big ups as well as no big downs) so it doesn't happen again.

Adam writes:

You win on this round, Arnold. What is Tyler thinking? Say yes to Zimbabwean economic policy? First, kill the managers, innovators, and entrepreneurs through taxes and regulatory. Then add inflation into the mix is only going to make things unimaginably worse. That is one way to destroy an economy. It's not the way to economic bliss.

We've just come out of a prolonged period of malinvestment in housing, finance, and capital goods in general. No policy is going to get us back on that track of bubbles. Resources need to be reallocated out of housing, finance, and capital goods. All the bailouts and economic policy tricks just keep resources in those bloated industries and prolong the crisis.

The best economic policies would be to encourage the reallocation. Provide internet clearinghouses for jobs and reemployment. Offer incentives for retraining and reeducation for skills in growing industries--tech, med care, etc. Also, strengthen the safety net for unemployed workers in a way that encourages adjustment, rather the more unemployment.



Scott Sumner writes:

Arnold, Thanks for the comments. Let me clarify that I am not really a monetarist, at least in the sense of someone who believes that the various monetary aggregates are useful indicators of the stance of monetary policy (except during hyperinflation.) On the other hand my views do coincide with those of the monetarists in other areas--such as transmission mechanisms.
My causality story is complex and not easy to explain. I believe the Fed is constantly sending signals about its intentions for the future path of monetary policy. Current changes in AD reflect changes in the future expected path of policy. My Warren post is an example of how I apply that concept. When policy has credibility, markets expect the Fed to do whatever is necessary to the money supply (or interest rates if you prefer) to get back quickly to target inflation (say 2%.) This worked pretty well in 1982-2007. When interest rates fall to near-zero levels (as in the fall of 2008), traditional instruments (monetary base and interest rates) work much less effectively. To preserve a credible 2% inflation target extraordinarily aggressive policy is needed. The Fed's passivity on monetary policy as it fought the banking crisis in September-October signaled to markets that the necessary activism would not be forthcoming. Markets crashed.

Bill Woolsey writes:


Resources should be realocated from housing, finance, etc. Where should they be going? Why aren't there shortages in the sectors that have been starved for resources as too many went into the bloated sectors.

Monetary policy isn't supposed to reverse desirable resource adjusments.

But what we are observing is surpluses of just about everything. Not a shift from one sort of production to other sorts, but less production across the board.

Monetary policy can correct a generalized decrease in spending caused by an increase in the demand to hold money. It doesn't matter why there is an increase in the demand for money.

If money demand is growing, an increase in the money supply doesn't result in higher prices--nothing like Zimbabwe. It instead just prevents falling prices.

Charlie Heath writes:

I'd say the US has to stabilize some asset prices - home values, equity - or else face further consequences of having gambled on equity over debt while borrowing heavily and happily inflating equity prices in a bubble.

In terms of US aggregate demand, it seems to me that at least at the consumer level, we're finally headed towards a sustainable path, that is, we have a positive savings rate, though not by much. Why should we be boosting consumer demand? And for capital spending, where is any justification for spending more, until private sector capital aligns with demand?

As for keynes and govt demand, I think the two best arguments in favor are (a) the federal government can still borrow at a rate better than anyone else, so might as well at least take advantage of that, if we can think of anything with a payoff at all, (b) we should do something to try to make a glide path to a sustainable aggregate demand within the global context, comparative savings rates, our currency hegemony,etc; (c) ought to try to hit some of Obama's infrastructure goals where there is a long term payoff, to take advangage of the current comparative sanity in terms of long term energy supply/costs. The plan since 1950 hasn't prepared US well for a competitive (or comfortable) future.

Barkley Rosser writes:

What I find a bit fantastic about this whole discussion is that the Fed has already gone way beyond what Sumner proposes, and not much has happened. Take a look at the Fed's balance sheet, which has massively expanded since September, quite few times over. It is now cluttered with unbelievable junk such as failed SIVs from Europe to the tune of more than $600 billion dollars. They took those on to help prop up the ECB, which could not handle that stuff. The Fed has gone far far beyond inflation-indexed bonds into an absolute netherworld, and with its credibility as shredded as it is now, even more publicly pushing its reported 2% inflation target is not going to do much either.

Comments for this entry have been closed
Return to top