“Doubt”? — how about confusion or impenetrability?

Here is a fun article about two macroeconomic forecasters.  I chuckled at this part:

Showing not even a hint of doubt, Mr. Hatzius said, “The prospect of substantial inflation seems very remote, but the prospect for deflation is far from remote. A double dip is certainly possible but not likely.”


I'm not a macroecomonyist, but that seems neither confusing nor impenetrable to me. It's not taking a firm stand, but given the subject taking a firm stand is not possible. I would say it's a bit heavy on the weasel words for such a short communication, but I'd have a hard time saying even that.

What did you chuck? :)

I don't see any confusion or impenetrability either. The statement seems like a clear assessment of the probabilities for the three possible events (substantial inflation, deflation, and a second dip). Did you want explicit numbers on those probabilities? Did you want conditional probabilities (probability of deflation conditional on no second dip, etc.)? Did you want him to go into the reasons for his assessment? Or are you just thinking that the phrase "without a hint of doubt" contradicts the inherently probabilistic nature of his assessment? (I don't think it's a contradiction: you can high confidence in the distribution of a random variable but still be uncertain about its realization.)

I will note that the article seems a little bit off in its description of the deflationistas. While most of them do support additional fiscal stimulus, few are pinning their hopes on it, and some never supported it in the first place. Right now they (we) are mainly arguing for more aggressive monetary policy.

I like forecasts that are made against type.

For example, a forecast made by a monetarist that is totally Keynesian, predicting AD and unemployment based on a federal stimulus program.

Or, since I work as an antitrust lawyer, person known as classical microeconomist advising on firm marketing strategy based on structure-conduct-performance models, and advising them to create barriers to entry or engage in conduct they would otherwise testify as futile in a competitive market.

Look at some of the forecasting/business modelling, and ask: what overall theory is being used to make this forecast.

Does he think it's going to rain on Tuesday?

He says inflation is unlikely and deflation is not unlikely. I don't understand the problem with that.

What is so funny is what the journalist says : "without a hint of doubt" - and the guy actually says "this might be probable, but this is not impossible." - which well, is actually doubting.

Tyler is getting a lot more snippy lately. I guess his terribad calls about the economy over the last 24 months made him more bitter than I suspected.

Nothing to laugh about to this Bayesian statistician here.

When you don't know equivocate. I guess you don't get well paid for
admitting you haven't got a clue.

The two guys above must think a point estimate is more informative than the whole probability distribution.

Comments for this post are closed