Scott Sumner  

Look past the labels

Moving Across My Monitor... In defense of Johnson's brain ...

Commenter CA directed me to a recent post by Roger Farmer:

We are stuck in a low inflation liquidity trap, caused by the fact that money and short term securities are currently perfect substitutes. The way out of the liquidity trap is to raise the interest rate; an argument that has been called neo-Fisherian in recent blog posts by Stephen Williamson and Noah Smith. Raising the interest rate however, and doing nothing else, will generate a recession; possibly a large and persistent recession. To prevent that from happening, the Treasury must engage in a simultaneous fiscal expansion. That fiscal expansion could be achieved through a money financed transfer to households; it could also be more efficiently achieved through a government guarantee to support asset prices.
At first glance this didn't look very promising. Those who follow my writing on monetary policy know that I am not a fan of either Neo-Fisherism or any sort of Keynesianism. Farmer seems to be trying to combine the two.

It's always important to look beyond labels, however, and on closer inspection Farmer's views are quite close to mine---closer than that of almost any other non-market monetarist.

Let's start with my views on Neo-Fisherism. I've noted that they do have a point---inflationary monetary policies are generally associated with higher interest rates, and vice versa. My criticism of Neo-Fisherians focused on their tendency to reason from a price change---to discuss the impact of an increase in interest rates, without first asking whether the change was generated by easier money or tighter money. After all, there are multiple ways to boost interest rates.

Farmer notices that simply raising rates would push the economy into recession. I interpret this as Farmer assuming that raising rates via a tight money policy would trigger a recession. And Farmer is correct on that point. We both think higher rates are needed, but only if generated by an expansionary policy mix.

What about Farmer's call for fiscal stimulus? After all, I'm pretty outspoken in my opposition to fiscal stimulus. Here's what he recommends:

Do not build roads and bridges as a temporary stimulus. A better way to prevent the recession that might otherwise occur when the Bank raises the Bank Rate would be an explicit commitment by the Financial Policy Committee of the Bank of England, to support the value of the stock market. This could be achieved by offering to buy or sell shares in an Exchange Traded Fund at a value linked to the performance of the unemployment rate.
Just to be clear, this is not my specific policy recommendation. Nonetheless, he's thinking about the issue in much the same way as I do. Farmer recognizes that a healthy economy would have higher interest rates, but also that an attempt by the central bank to arbitrarily raise rates via a tight money policy would be contractionary. So he wants to combine higher rates with another policy, which boosts aggregate demand. And he doesn't want that "fiscal stimulus" to consist of more government spending or tax cuts, instead preferring asset purchases. These would be used to target asset prices.

My view is that we can combine higher interest rates with a more expansionary monetary policy if we adopt a higher NGDP target and commit to do whatever it takes to achieve that target. While Farmer wants to target something like a stock market index, I prefer an alternative asset price, preferably NGDP futures. But if I can't have NGDP futures, then targeting another asset price (exchange rates, commodity price index, stock price index, etc.) would be vastly superior to targeting interest rates.

There is no reliable relationship between interest rates and aggregate demand. We don't even know whether higher interest rates would be associated with more demand, or less. In contrast, a policy that pushes up the price of NGDP futures, or the price of foreign exchange, or the S&P500, is very likely to be expansionary. Farmer recognizes that our current interest rate-oriented approach to policy has led us to a cul de sac. Promising low rates for the foreseeable future is seen by conventional Keynesians as an expansionary policy. Farmer and I recognize that it might just as well be interpreted by the markets as a promise to "become Japan."

I've advocated a "whatever it takes" approach to monetary policy, combined with targeting the forecast. Thus the central bank buys as many assets as are required to boost NGDP growth expectations up to the target. You start with T-bonds, then if they are exhausted you switch to other safe assets such as AAA bonds, and if necessary you buy riskier assets such as common stocks. Where Farmer and I differ is that I don't think it would ever be necessary to buy common stock, unless the NGDP growth target were set at a very low level. Farmer is more pessimistic about liquidity traps, and doesn't think T-bond purchases alone would be enough to do the job.

To conclude, both Farmer and I agree on:

1. Targeting NGDP
2. Higher interest rates are likely to accompany a successful expansionary policy
3. Conventional fiscal policy is unwise
4. We both would prefer the government purchases assets, rather than build infrastructure, if this were needed to hit the target.
5. We both think of monetary policy in terms of targeting a price such as NGDP futures, commodity prices indices, forex prices, stock indices---not interest rates.

We differ on:

1. He calls central bank purchases of assets "fiscal policy", while I call it monetary policy. That's just a semantic difference.
2. He prefers to target stock price indices during a liquidity trap; I prefer NGDP futures prices.
3. He thinks the central bank would have to go beyond T-security purchases to escape a liquidity trap. I don't think they would have to unless the NGDP target were set too low.

This is an example of two economists who travelled very different paths and ended up in almost the same place. Sometimes you need to look beyond the labels to understand what someone is actually proposing.

Screen Shot 2016-09-30 at 9.59.21 PM.png

Comments and Sharing

COMMENTS (14 to date)
Jose Romeu Robazzi writes:

Great post.
Perhaps, countries that are highly indebted could profit most from an expansionary policy that buys private assets (stocks, real estate securities), since that policy would not be seem as public debt monetization and perhaps be more effective.

Also, I have studied under Jeremy Siegel from the Wharton School, and his book "Stocks for the Long Run" has a very interesting discovery: for longer terms, above 20 year holding periods, stocks are very low risk, in fact, if one takes consecutive 20 year returns on a broad stock index such as the S&P 500, there is never a real (inflation adjusted) loss, as opposed to holding bonds, that may show a real loss on 20 year holding periods depending on the time window one chooses. I think that this fact is important because mitigates a lot the criticism that there may exist negative fiscal impact of monetary policy decisions that disctate buying "risky" assets such as stocks. As an ongoning concern, the central bank has infinite holding period to worry about, therefore the fiscal impact will be that from time to time the central bank will distribute financial gains to the treasury ...

Market Fiscalist writes:

Why would targeting a stock market index be a good idea ? The stock market is very volatile left to its own devices, and for a CB to adjust the money supply to eliminate this volatility would lead to the volatility being transferred to inflation and NGDP, wouldn't it ? For example if the CB had tried to back out the 1987 stock market crash via monetary policy it would have led to a very bad outcomes.

Am I missing the point perhaps ?

Morgan writes:

This might be somewhat off topic, but goes to the question of the mechanism by which we counter slow growth due to plodding NGDP growth.

So I'll be honest about this, even if it outs me as an economic naïf. But why isn't the obvious way to boost NGDP simply to print money? I'm not talking about Zimbabwe levels of printing money, I'm just saying - if you want NGDP to grow by 4% a year, and it's currently growing by 1.4%, why not increase the money available to spend on domestic output until the actual increase in spend on domestic output is 4%?

Why use interest rates or bond sales or index funds as an intermediary mechanism?

mbka writes:

Very interesting. This proposal would be monetary offset in reverse, right? Brilliant. Details to be worked out for sure...

Disclaimer - I don't really understand monetary policy - that's why I read your blogs. And I keep on adding little nuggets like these to start making sense of it.

Will writes:

Reminds me of the FTPL view where raising rates is assumed to be "contractionary" if the higher interest rate on debt is assumed to be financed by more contractionary fiscal policy in the future.

Farmer's proposal seems to prevent this from happening as it ties fiscal policy down to a nominal indicator.

Bryan Atkins writes:

[Comment removed pending confirmation of email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

patrick k writes:

All I can say, thank god no one listens to this blackboard day dreaming. The problem with the economy is too much tinkering not too little.

Scott Sumner writes:

Thanks Jose.

Market, I certainly wouldn't favor pegging stock prices at a fixed level. Perhaps they could be used as an intermediate target. Even so, I'd prefer NGDP futures.

Morgan, I completely agree.

mbka, Can you be more specific as to how this is monetary offset in reverse?

Will. The key is to combine interest rate policy with something else, so that you pin down whether the move in interest rates is expansionary or contractionary.

mbka writes:


Can you be more specific as to how this is monetary offset in reverse?

feel free to shoot me down on this one - I am at the edge of my understanding here. So here it goes:

As I understand it, monetary offset is when a central bank produces monetary policies to offset the consequences of fiscal policies, in order to stabilize monetary economic variables.

The above discussed proposal is when a fiscal authority proposes fiscal policies to offset the consequences of monetary policies, in order to preserve non-monetary economic variables.

That's what I meant by "reverse".

Brent Buckner writes:

Prof. Sumner, you wrote:
"I am not a fan of either Neo-Fisherism or any sort of Keynesianism"

OTOH as I have read it you have often emphasized that your beliefs fit pretty well within a neo-Keynesian framework.

"It's always important to look beyond labels"


Scott Sumner writes:

mbka, Thanks, that makes sense. I missed it because I view the proposed policy as monetary, but you are right that Farmer (and many others) view it as fiscal.

Brent, Yes, I think it fits pretty well with pre-2008 New Keynesianism, but not the current version.

The current version seems to see interest rates as an indicator of the stance of monetary policy, and also seems to think monetary policy becomes ineffective at zero rates. That was not the NK view in 2007.

BC writes:

Suppose someone had a fetish for fiscal policy and infrastructure and they said that they agreed with Scott and Farmer but they think that, instead of purchasing T-Bonds or equities, the government should purchase securities of construction firms and other firms that build infrastructure. Does buying these securities at market prices avoid Cantillon effects and other distortions? Suppose they went a step further and suggested that asset purchases should include securities of special-purpose investment vehicles that use proceeds from securities sales to build infrastructure like roads and bridges and use the toll revenues to pay coupon payments on the securities. Suppose they went one more step and said that the government might as well purchase the underlying infrastructure assets themselves: buy raw materials and construction labor (at market prices) to build roads and bridges. The government would end up owning the actual roads and bridges rather than securities with payments linked to the roads and bridges.

At what point does monetary policy become fiscal policy, asset purchases become government spending? It seems like the nature of the securities and/or the size of the government's holdings relative to the overall market size should matter in some way.

Scott Sumner writes:

BC, Monetary policy does become a bit more fiscal policy-like each step of the way. I tend to view it as fiscal policy if it increases the future tax burden of the public. Buying ETFs does not do that, filling potholes (probably) does.

The question is whether the activity is costly. Monetary policy is not costly, fiscal policy is.

You said:

"Suppose someone had a fetish for fiscal policy and infrastructure"

Using infrastructure spending for fiscal policy doesn't get you any more infrastructure than not using it. Rather the infrastructure you do get is built in a more countercyclical fashion.

BC writes:

Good point on the infrastructure spending for fiscal policy being countercyclical but not higher on average.

For infrastructure spending to not increase future tax burden, I guess it would need to be paid for with user fees and tolls? That is a big difference between purchasing infrastructure-linked securities vs. directly building infrastructure without subsequently charging tolls and user fees. Taxpayers pick up the bill in the latter case.

Comments for this entry have been closed
Return to top