In a tweet, former Obama economic advisor Austan Goolsbee asked:
q for NGDP target fans: if fed target=5% when actual=0, implicitly means target 5% infl for 1 yr. Wouldn’t mkt doubt 5% infl just temporary?
The answer (from a fellow UC alum) is no. A good example occurred in 2007-08:
July 2007 to July 2008: CPI rises 5.5%
2007:Q3 to 2008:Q3: Real GDP falls by 0.3%
That data is pretty close to the supply shock Austan has in mind. And during the third quarter of 2008 inflation expectations (TIPS spreads) fluctuated above and below 2%. Investors know that inflation spikes due to supply shocks are transitory. Under NGDP targeting, inflation doesn’t have momentum; NGDP growth does.
There is a separate issue of what happens if trend RGDP growth slows. I favor stabilizing NGDP per capita. If it slows due to lower productivity, then slightly higher inflation is actually desirable, and vice versa. During the productivity boom of the late 1990s, the Fed should have aimed for below trend inflation, to allow room for higher inflation in the next recession.
PS. I presume Goolsbee means “real” when he says “actual.” NGDP is “actual” GDP. RGDP is a number pulled out of thin air by government bureaucrats. It doesn’t correspond to anything in the real world, unless you think a government bureaucrat’s estimate of how much more you enjoy watching I Love Lucy on a 4k 80 inch OLED TV compared to a black and white TV is constitutes “interesting philosophical speculation.”
HT: MFFA
READER COMMENTS
R Richard Schweitzer
Jun 3 2015 at 12:59pm
The more one reads about such “Targets” to be shot at by “experts,” the more it seems appropriate to suggest that all should be first competent in Zen target shooting.
Glenn
Jun 3 2015 at 1:29pm
Scott says:
PS. I presume Goolsbee means “real” when he says “actual.” NGDP is “actual” GDP. RGDP is a number pulled out of thin air by government bureaucrats. It doesn’t correspond to anything in the real world, unless you think a government bureaucrat’s estimate of how much more you enjoy watching I Love Lucy on a 4k 80 inch OLED TV compared to a black and white TV is constitutes “interesting philosophical speculation.”
Wow-o! I was taught exactly the opposite. So if I double prices (in cash) and the cash supply simultaneously, “actual” GDP has gone up by two-fold? And your bureaucratic construction that claims there has been no change in “actual” GDP is “actually” so much alchemy and not to be trusted?
Triumph of ideology over reality, indeed.
Certainly, RGDP is an imperfect measure. But it is far superior to NGDP, which is TRULY meaningless (as opposed to RGDP, which is de jure meaningful but de facto imprecise). And in any case, empirical evidence strongly suggests inflation has a unit root, which I think was the point.
Jose Romeu Robazzi
Jun 4 2015 at 9:09am
Prof. Sumner
I keep talking to people about NGDP targeting, and most like the idea, but have the same two concerns:
1. It is hard to explain to people (as opposed to inflation targeting, which is supposed to be “intuitive”)
2. It will require massive levels of intervention (as opposed two your claim that ajustments will be minor)
So I guess that these two concerns have to be stressed as much as you can …
Scott Sumner
Jun 4 2015 at 9:20am
Jose, Thanks for the question:
1. I’d guess 99% of American have absolutely no idea what inflation targeting is. For instance, they don’t understand the distinction between supply and demand-side inflation, and they don’t understand that when the Fed tries to increase the inflation rate it is trying to increase the real incomes of Americans. Heck, they don’t even understand why the Fed would ever try to increase the inflation rate.
I think NGDPLT is easier to explain; just keep the incomes of Americans growing at a steady 4% to 5% per year. What could be simpler?
2. The second point is exactly backwards. Without NGDPLT we have to do massive interventions like QE. With NGDPLT, we would not have to do QE.
ThomasH
Jun 4 2015 at 4:04pm
Perhaps you have (you probably have) covered this before, but I do not see much difference between NGDPL trend rate targeting and PL trend rate targeting. It seem that most of the advantage of NGDPL trend targeting is that it’s a level and progressive departures imply progressively greater efforts to return it to trend, but that is also the case with PL trend targeting. The main problem with Fed policy since 2007 has been that it’s an inflation rate ceiling policy not an (average) inflation rate policy (= PL trend policy). In the presence of really sticky prices and really large shocks could not a higher PL trend target be superior to a lower NGDPL trend target.
Second, why would Fed purchase or sale of NGDP futures contracts be less “interventonary” than purchase of LT assets as a means of achieving its NGDPL target? Would the same argument apply to a PL futures market?
W. Peden
Jun 5 2015 at 5:09am
“I think NGDPLT is easier to explain; just keep the incomes of Americans growing at a steady 4% to 5% per year. What could be simpler?”
I’d add the qualification “average” to “incomes”, otherwise it might sound like you’re proposing some sort of 1970’s-style wage policy.
By the way Scott, I find this graph interesting-
https://research.stlouisfed.org/fred2/graph/?g=1dui
The past few years in the US have had exceptional stability in NGDP growth rates, even in comparison to the Great Moderation. Given the Fed hasn’t even been targeting NGDP growth in this period, it proves that an exceptional degree of NGDP growth stability is possible, and presumably it is achievable under more suitable arrangements e.g. NGDPLT with a futures market.
Mike Rulle
Jun 5 2015 at 1:40pm
Your writing has become more direct, broader in scope, more humorous and more persuasive. I like your comment on “4k 80 inch OLED TV” which made me laugh. I assume you mean with NGDP you pretty much know what you have got in aggregate monetary terms, whereas with RGDP it is too uncertain and subjective to use as target. But RGDP is still an interesting concept.
Comments are closed.