Friday, February 26, 2016

Romer^2 on Friedman on Sanders

Oh dear.  Today's Romer and Romer response to Gerald Friedman’s paper on the economic consequences of Sanders seems to have identified the core problem in GF’s analysis, confusing one-time and ongoing stimulus effects.  According to the R team, F attributed increases in economic growth in perpetuity to single bursts of stimulus, and not just once but repeatedly—in his treatments of demand stimulus, income redistribution and health care.  That plus his belief that the output gap is large enough to accommodate extremely rapid growth over a full decade, explains his headline numbers.  If this is true it’s an embarrassment.

For economists and other academics the message is to always circulate drafts before going public.  For the rest of us, don’t forget that this tussle has nothing to do with the merits of Sanders’ policies.

UPDATE: In my original post I attributed the supercharged growth rates to multiplier issues, primarily in out years.  That appears to be partly correct, in that GF's multiplier depends on his estimate of the output gap, which he thinks would remain substantial through the entire forecast period.  But the larger problem seems to be the multiplicands.  This is the sort of thing that requires actual replication to uncover.  Incidentally, this is turning into a poster case for the value of writing down your model and not jumping immediately into the spreadsheets.

4 comments:

Anonymous said...

> For the rest of us, don’t forget that this tussle has nothing to do with the merits of Sanders’ policies.

+1

ProGrowthLiberal said...

Nice followup on your original review! Romer and Romer did a rebuttal of what I have called junk science. That is the easy part. Now time for the hard work of doing an affirmative analysis as I just posted. Team Republican will have their advocacy analysis. Team Democrat needs a credible analysis.

Thornton Hall said...

Can someone cite the empirical evidence that suggests economists can predict the future?

Sandwichman said...

"Can someone cite the empirical evidence that suggests economists can predict the future?"

There is none.