I think I will begin a discussion about “Nominal GDP targeting,” which seems to be a popular notion these days. I have not said much about this, except to mention that I think it is a bad idea.
At the outset, I will say a few things. First, my audience: this is mostly for generalists (for example, Congresspeople) who don’t necessarily know all that much about monetary economics. Second, it is for people who do know something about monetary economics, having followed my writings on this site, and in four books, over the years. As they are familiar with the concepts that I use, and their foundation, the discussion will make sense to them. It is not, for the most part, for academics, including those both in favor of and against NGDP Targeting. They will inevitably be disappointed, and grumble about this and that, such as not following the norms of academic society; norms which seem to produce exceedingly mediocre results, it seems to me.
Along the way, I will make some big generalizations, simplifications and interpretations. This is largely so that I have something to talk about that makes sense, even if I don’t agree with it. Unfortunately, most academic writing on these topics is such a mass of confusion that it is hard to address it directly at all. I have to “interpret” and “generalize” it just to talk about it. This is a condition that we ran into many times during our discussion of monetary interpretations of the Great Depression, in 2016 and 2017.
August 24, 2017: The Interwar Period #2: It’s Not That Complicated
I could try to address something — an academic paper, let’s say — directly, on a point-by-point basis. But, in general I have found this to be very laborious and rather unproductive. I could write 10,000 words on all the confusion and error in a single paragraph of a typical academic paper or book. And where would we be at the end of all that? Probably nowhere, since nobody would read such a thing anyway, and even if they did, probably nobody would get anything out of it.
First, some general points. NGDP Targeting is based on the idea that a currency manager (for example, central bank) should have a target of maintaining an unchanging growth rate of Nominal Gross Domestic Product. Since NGDP is commonly deconstructed as a “real” component, and an “inflationary” component (this deconstruction in itself has some problems), we can see that NGDP targeting will have certain tendencies:
- When “real GDP” is low and “inflation” is low, basically a recessionary situation with low inflation, monetary policy is “easy.”
- When “real GDP” is high and “inflation” is high, monetary policy is “tight.”
- When “real GDP” is high (but not too high) and “inflation” is low, this is presumably a happy situation.
- When “real GDP” is low but “inflation” is a little high, this is also presumably a happy situation, since we have achieved the “monetary ease” that we desired.
Another common way to think about it is that the “money” supplied is adequate to maintain an evenly growing “aggregate demand” (basically, this is NGDP). This is basically the same thing as above, but in somewhat Keynesian terms.
You could have “soft form” NGDP targeting, in which NGDP serves as a primary guideline for central bank policy, which is nevertheless undertaken in a somewhat ad-hoc manner as is common today. This serves in contrast somewhat to “CPI targeting,” which is common among central banks today, in which there is an official CPI target, but this is not quite binding and serves as only a sort of primary guideline for central bank action that is undertaken in an ad-hoc fashion. The difference between the two comes when CPI is acceptable but “real” GDP is low. The NGDP format would be more aggressively “easy.” Or, when CPI is acceptable and “real” GDP is high. NGDP targeting would be more aggressively “tight.”
However, all this “ad-hoc” policy by central banks, whatever their claimed policy targets might be, has itself been unpopular. Are we really going to just make it up as we go along? Just pull rabbits out of hats — interest on reserves, QE, negative interest rates, “modern monetary theory,” “cashless society” — whenever it seems good for a giggle? Shouldn’t this process have some sort of democratic accountability? This has led to interest in various “rules-based” systems, which could be applied to CPI targeting, NGDP targeting, Taylor rules or a variety of other proposals. Basically, you could program a computer to do it, using certain rigorous, pre-defined protocols, and then you wouldn’t have “ad-hoc” policy made by central bank committees making stuff up as they go along, or, as traders might say when nobody else is listening, “pulling $hi+ out of their @$$.” This raises a few issues: one of them is whether the computer system can actually achieve the goal. Would NGDP, in this automatic system, actually be stable, or hit its target? Another is: once we decide to “let the computer do it,” then the system is on autopilot without human intervention, and we have to live with whatever happens, in terms of interest rates, foreign exchange rates, economic performance and so forth. In practice, I think it would not be very long before the human element re-introduced itself. This could take several forms:
- The policy target is changed. The CPI target, the NGDP target, etc., is reviewed (by central bankers, by Congressional order, by Presidential order) and then it is changed. The NGDP target used to be 3% and now it is 5%. Now we have the question: how is this decision made? Who makes it? How often is it made? What is the process by which it is made? Is the outcome going to be an improvement, or will it make things worse? Wall Street traders want to know.
- The statistics that go into the process are gamed. It won’t take very long (about fifteen minutes) before someone figures out that you can introduce a human element in the “automatic” system by jiggering the figures that it uses as inputs: CPI, NGDP etc. When the economy is weak and an election is coming up, maybe we will change how the CPI (or NGDP) is calculated. Today, I don’t think that NGDP is jiggered too much. “Real” GDP is definitely jiggered today, by gaming the “GDP deflator” (i.e., inflation component). If you lowball the GDP deflator, then “real GDP” looks better. The CPI has been so thoroughly abused that it sometimes seems like little more than propaganda.
- Overt human override. Paul Volcker tried the “Monetarist Experiment” in 1979-1982, which was a loose sort of “rules based” system. It didn’t work very well, and in 1982 he threw it out and did something else. This of course raises the question: Will the central bank override the “automatic” system? Will Congress demand an override? The president? When? What are the chances? What will replace it? Wall Street traders want to know.
These might seem somewhat theoretical now, because we are not yet in this situation. But, when the dollar drops 20% on the foreign exchange market, and interest rates jump two percentage points across the board, it will be a big deal. When Volcker began the “Monetarist experiment” in 1979, the extreme volatility of forex, gold, interest rates, commodity prices and economic performance that followed was not expected. So, when people “don’t expect” such things today, that is the norm for these things. When all that stuff actually began to actually happen, the solution was: get rid of the damn thing.
Or, let’s say the economy is all fired up for some serious growth — after President Trump passes his brilliant Flat Tax plan with a 17% uniform rate for businesses and individuals — but the whole thing is messed up by an NGDP target that squashes the economy down to 3.65% nominal growth (the 3.65% figure was recently suggested by an NGDP advocate). The monetary restriction this requires might lead to a soaring dollar on forex markets, accompanied by complaints from exporters and a wave of defaults by dollar-denominated borrowers worldwide — just as the rise in the dollar in 1982 blew up all of Latin America, which defaulted en masse and then had a “lost decade” of hyperinflation when their currencies broke down. (This dollar rise in 1982 had some relation to the first Reagan tax cut, so these scenarios are not so fanciful as they may at first appear. Another dollar rise in 1997-98 was related to another tax cut at that time, which reduced capital gains taxes.) For comparison, during the Bretton Woods gold standard era of the 1960s, U.S. nominal GDP was about 7.0% annualized, and Japan’s was 16.9%. These growth rates are certainly possible. Or, I should say that they are possible with a gold standard system, since they are obviously impossible in an NGDP targeting system with a 3.65% target.
(The 7.0% figure for the 1960s may have had a little inflationary boost, since the actual market dollar/gold ratio during much of the decade was more in the $38-$40/oz. range, indicative of dollar weakness against gold compared to the $35/oz. parity. Nevertheless, an NGDP figure of this level, or even higher, could certainly have been possible if there had been no tax increases by Johnson and then Nixon.)
When the currency’s value is stable, then “real” GDP and nominal GDP are essentially the same. The official CPI in Japan was 5.5% per year in the 1960s. Neither the government nor the Bank of Japan cared, and it wasn’t a problem. This business of separating “real” GDP and “inflation” basically came about as a result of the currency depreciation and floating currencies that emerged after 1971.
Let’s say the economy is not doing so well, because President Elizabeth Warren introduced 70% top income tax rates, wealth taxes, is harassing big business in general, and is taxing capital gains and dividends as regular income, with a top rate of 70%, which rises to 80% or more when some State taxes are added. Of course these are not indexed to inflation, so “real” capital gains after tax can be negative, which, as you might guess, is not so good for capitalism. Since we have an NGDP target, it seems like disaster is avoided because NGDP is the same 3.65% it would have been if there were no tax changes, or other threats to big business. But, it would take a lot of monetary distortion to accomplish this outcome. In the end, it would probably amount to a big decline in dollar value. How big? That is an interesting question. So, we would have some big Warren tax hikes, which still haven’t been fixed, and on top of that, a big devaluation. I think that, in this case Two Wrongs would not Make a Right. (I actually wrote a book about that, The Magic Formula.)
Let’s try to imagine what something like this might look like at the international level. The dollar’s value would be going up and down, and nobody would be able to do anything about it since the system is basically on autopilot. Interest rates might do funny things. (Note that NGDP targeting proposals typically have no interest rate targets, so they would be left to the free market.) Today, there are about 40 countries that are part of the “dollar bloc,” and which explicitly link the value of their currencies to the dollar. What are they going to think about the dollar’s value, and dollar interest rates, going here and there according to domestic U.S. conditions? The rise in the dollar’s value in 1982 and 1998 blew up dollar pegs worldwide. The rise in 1985 would have blown up dollar pegs worldwide, if they hadn’t already been blown up in 1982. In any case, it required a Plaza Accord to beat the dollar down. The fall in the dollar’s value in 1971 led countries around the world to abandon their currencies’ link to the dollar. Several additional countries have currencies that are “stabilized” against the dollar, using something like a crawling peg. China is one of them. Also, there are a lot of entities worldwide with dollar-denominated debt or other obligations, for example from countries that do not have a dollar peg and instead have some crapulous local floating currency, and as a result can only borrow in dollars because nobody would lend in local currencies. What are all these countries, and entities using dollars worldwide going to do? Will they find some other solution, and dump the dollar for international use?
Today, central banks generally try to maintain somewhat stable exchange rates, although they rarely like to talk about this. The present “ad hoc” central bank methods allow central banks to keep things from potentially getting too far out of line. With some sort of automatic “rules-based” system, all this goes out the window. You get what you get. What does it look like if several countries worldwide start using NGDP targeting? Or do we expect them to accommodate the U.S.’s follies, and try to keep exchange rates from varying too much (basically following the dollar loosely on this travels here and there). I bet that wouldn’t be too popular among foreigners.
All of these automatic “rules-based” systems, whether a CPI target, an NGDP target or some other thing, tend to be intensely domestic-focused. The question of how it plays out internationally is hardly addressed.
That is enough to think about for now. I will say more about it, even before getting into what the NGDP advocates themselves have to say.
