Scott Sumner  

Monetary offset isn't really a choice, it's built in to policy

Deficits always matter... Trump's Trade Trifecta...

James Alexander directed me to a very good Bloomberg article on "monetary offset":

Federal Reserve Chair Janet Yellen may well be thinking it. Jeffrey Lacker, president of the Richmond Fed, came close to saying it. "A more stimulative fiscal outlook usually warrants higher policy rates," Lacker said a week after Donald Trump won the presidency. Translation: Trump's plans to stimulate economic growth could inspire the Fed to move in the opposite direction. Such a move, which would put the U.S. president and its central bank at odds, is called a monetary offset.
In my view, monetary offset is less discretionary than most people assume. Of course in a literal sense the title of this post is wrong, the Fed can choose to engage in monetary offset, or not choose to. But in a deeper sense I don't really think they have much choice.

Right now the Fed has a very impressive economic research apparatus, which models the economy under alternative monetary policy options. FOMC members then produce forecasts of inflation under "appropriate monetary policy". When you look two years out, the Fed's inflation forecast is generally about 2%. That's because the Fed's inflation target is 2%, and they wish to set policy at a level where they achieve this target. An exception might occur during a deep slump like 2009, where the Fed might believe that it will take a bit more than 2 years to get back to 2% inflation. But of course that's not where we are today.

As long as the Fed decides policy using this deliberative procedure, monetary policy is somewhat automatic, or at least less discretionary than most people assume from looking at the FOMC ponder over rate changes. And this means that as long as they continually set the fed funds target at a level expected to produce 2% inflation a couple years in the future, they will be automatically adjusting interest rates to completely offset the demand-side effects of fiscal stimulus (of course supply-side effects are another story.)

How would we know if the Fed had decided not to sterilize the impact of fiscal stimulus from the Trump administration? The answer is simple, you'd see the Fed's 2018 inflation forecast rise to 2.5%, or 3%. But that's not what we see, and I think it's very unlikely that we will see this sort of inflation forecast in the near future. If they didn't raise the inflation target during the recession (when it would have been helpful for growth) why would they raise it now?

Perhaps people are confused about this because they think in terms of stimulus being aimed at boosting growth, not inflation. Since the Fed has no mandate to restrict growth, why would it offset fiscal stimulus? But the direct effect of fiscal stimulus is to boost NGDP, and any further impact on RGDP depends on the slope of the short run AS curve. (Or the Phillips curve, if you prefer.) Even some economists appear confused on this point. Here's Simon Wren-Lewis, discussing empirical studies of monetary offset:

As far as I know, no one had expressed a concern about fiscal austerity because of the impact this will have on nominal GDP. The issue is always the impact on real activity, for reasons that are obvious enough.
Here Wren-Lewis confuses the ultimate goal with the direct effect. Real GDP can rise for many reasons, including demand and supply-side shocks. Keynesian stimulus is generally assumed to work through demand-side effects, which means that both growth and inflation should rise if stimulus is boosting demand. If you see growth rise but no change in NGDP, then the cause of the growth is obviously not "more spending" it's an increase in the incentive to work, invest and innovate. That sort of growth would not validate the theories of John Maynard Keynes; it would validate the theories of Arthur Laffer.

Here's how Nick Rowe responded to Wren-Lewis:

I disagree on the NGDP vs RGDP thing. Sure, RGDP is what we care about, but we don't know whether a rightward shift in the AD curve will cause increased RGDP, increased price level, or (more likely) some mixture of the two. Using NGDP instead of RGDP is a crude way of acknowledging our ignorance about the slope of the Short Run Phillips Curve.

Now, if fiscal loosening (in countries with flexible exchange rates) causes RGDP to increase but does not cause NGDP to increase, which is what your results and Mark's [Sadowski] results seem to imply, when taken together, we have a very uncomfortable conclusion for all of us. It tells us that the Short Run Phillips Curve slopes the wrong way. It tells us that an expansionary fiscal policy causes inflation to *fall*. Which is a very strange result indeed, unless you want to talk about the supply-side benefits of bigger deficit spending.

Now the supply-side argument is not stupid. If you suddenly lay off thousands of government workers, that causes some real supply side issues, because those workers won't be able to reallocate instantly even if aggregate demand does not fall. The search theorists do have some sort of a point. Any sort of change in the composition of demand between private and government sectors, in either direction, will cause some temporary increase in frictional unemployment, shifting the SRPC in the bad direction.

I don't find my above "explanation" fully convincing. But it's better than assuming the SRPC slopes the wrong way, so that an expansionary fiscal (or monetary) policy *reduces* inflation in the short run.

I agree. But an even simpler explanation is that tax increases reduce output in exactly the way predicted by Art Laffer.

I'm not an enthusiastic supply-sider like Laffer or Larry Kudlow, but even a moderate supply-sider like me would predict some negative growth effects from tax increases.

To summarize, Wren-Lewis thought he was testing the Keynesian model, whereas it was actually a test of whether Keynesian or supply-side models are correct. His results suggest that the basic demand-based Keynesian model is wrong and the supply-side model is right.

PS. New Keynesian models actually do allow for some supply-side effects. Thus if you spend a lot on military goods, then workers will feel poorer, and therefore they will wish to work harder to maintain their living standards. Obviously this is not the sort of transmission mechanism that people like Paul Krugman and Wren-Lewis have in mind when advocating fiscal stimulus.

HT: Ramesh Ponnuru, Stephen Kirchner

Comments and Sharing

COMMENTS (7 to date)
Thaomas writes:

New Keynesian Models also give a "supply side" effect if governments fail to react to lower borrowing rates and marginal costs of many factors of production that fall below their market prices (sticky wages/rents/price of entrepreneurial effort) to invest in things that increase the productive capacity of the economy.

Do you consider HANK-SAM models

To vindicate your policy prescription for the Fed to target NDGP trend?

Scott Sumner writes:

Thaomas, My response is the same as for all the dozens of previous times you have made this argument---never reason from a price change.

Thaomas writes:


When isn't acting on a price change, a business opportunity as a recession is for a prudent government a good idea? I do not see what "reasoning from a price change" has to do with it.

Don Geddis writes:

@Thaomas: "When isn't acting on a price change ... a good idea?"

You have assumed that lower borrowing rates mean that there are now more positive opportunities for investment. Which means that you have assumed the lower rates were caused by a shift of the credit supply curve.

Sumner's "never reason from a price change" points out that lower borrowing rates can just as easily arise from the opposite reason, a shift of the demand curve for credit, basically because the economy (and future expectations) have gotten so much worse, that there are now many fewer positive investment opportunities than there used to be.

Obviously, in the latter case, it would be foolish to choose that moment (when there are fewer net positive investment opportunities) to decide to increase investment.

Since all you have identified is that "borrowing rates are lower" -- i.e., the price of credit has dropped -- you offer no evidence that the lower rates have any connection to increased investment opportunities. They are just as likely to be connected to decreased investment opportunities instead.

You need to know why the rates lowered.

Blair writes:

Scott, what did you make of Bob Hall's comments last week on Econtalk? Right at the end, he appears to state that there is an overwhelming consensus that monetary offset doesn't exist. Russ Roberts sounds momentarily confused but then they run out of time before getting to the bottom of it.

[link added--Econlib Ed.]

Blair writes:

Here is the transcript:

Guest: No. I think a lot of the ideas that have permeated the consensus macro modeling--first of all, the term 'Keynesian' is--it means so many different things to different people. If you look at how most macroeconomists build a model, it has a huge number of things in common. They--a parting point, if, what's called the New Keynesian Model, isn't related to this question that you mentioned at all, whether or not fiscal policy is effective. It all has to do with whether the economy reaches its equilibrium quickly or whether there's a time when prices are wrong--that is, prices are sticky and wages may be sticky. But then there's a lot of work--and this is, certainly on the labor market side, this is something that I've contributed to, is applying sort of more standard analysis: nothing sticky but other things that sort of mimic stickiness that represent equilibrium. So, that kind of work is, I think, has taken us quite a ways. And it's been in place for a while. You know, Lucas's ideas, which just seemed so dramatically different from the way we've been brought up. Became fully integrated. And now, our scene is, 'Well, that's sort of the old-fashioned way of thinking. Rational expectations, that's old-fashioned.' The idea is sweeping now that we should model expectations as beliefs that can be different across people, heterogeneous beliefs. And people at Chicago are just as enthusiastic as anybody about pursuing that kind of idea. There's--the idea of schools of macro has pretty much disappeared. Especially in the younger generation. It would be impossible find, you know, an economist in a top university today who could be reasonably identified to be--well, especially a monetarist. But even a Keynesian in the old-fashioned sense. We all build models in which if the government purchases more, there's higher GDP. It just has to be. And no one would disagree about that today. So if that's the--if that's the question about what Keynesian means, then Keynes is gone completely. We all agree that an important determinant of total activity is how much the government chooses to buy.

Russ Roberts: Well, I was going to end there. But I can't. I can't. Because I have to challenge that statement. Certainly there are many, many illustrious economists, far more illustrious than I am. Two that come to mind are Valerie Ramey and Robert Barro who have suggested that that relationship is not very reliable or doesn't hold at all. Or it could be negative.

Guest: No, no, no, no, no.

Russ Roberts: Do you think that's true?

Guest: No, no. You are talking about--first of all, Valerie was a student of mine. And I follow her work very carefully.

Russ Roberts: Doesn't she say that government spending has little effect on--that the multiplier is quite small?

Guest: Okay. So, her most recent paper says--well, first of all you have to decide whether you are [?] or at a lower bound or not. And if you aren't, then a billion dollars of government purchases adds about a billion dollars to GDP. So, the so-called multiplier is 1. And I think that's within the consensus range. It's certainly within my beliefs.

Russ Roberts: That doesn't stimulate anything. That's not offset--that's not worrying whether it's offset by monetary policy?

Guest: No. That's assuming that it is offset. Nonetheless, there's an effect. By historical pattern. This is poring over historical data.

Russ Roberts: Correct.

Guest: And some old examples of rapid change of government purchases are the main source of the information. So, since we haven't had much of it in the last 20 years, this is mostly evidence from more than 20 years ago. But it's the evidence we have. And, you know, studies from other countries have been supportive. There's a big literature on this topic. A multiplier of 0, which would mean no effect, is outside the range of almost all empirical work and almost all modeling. It would not be a viable position. And I know Valerie's work; and Barro's paper, QJE (Quarterly Journal of Economics) paper, gets positive multipliers. Not large, but positive. And there's pretty much agreement now that a 0-lower-bound where monetary policy's hands are tied, that multiplier is higher. And that's what's Valerie's most recent work shows. So I think it all fits together. But there's nobody, nobody who says that it makes no difference at all if the government buys more. And that's a substantive thing that's happening in the economy. Of course it's going to influence the economy.

[from --Econlib Ed.]

Scott Sumner writes:

Thaomas, Again, You cannot know how quantity should respond to a price change unless you know WHY the price changed. If it fell because demand fell, then quantity should fall. If it fell because supply increased, then quantity should rise.

Blair, I missed it, but monetary offset is standard macro theory. Even Paul Krugman buys it. So do monetarists. Not sure what "consensus" he is referring to.

Comments for this entry have been closed
Return to top