The international community is scrambling to figure out whether Nick Rowe or Dean Baker is the aggressor.
Nick initially counted on Brad DeLong as an ally, mistaking his Swissesque (how's that - I adjectivized an adjective) neutrality for sympathy. But it was not to be.
Last time this came up I took my usual cowardly "you all have a point" position. I felt that people were talking past each other in what they meant when they referred to "generations" or "the future". Are we talking about an economy at a point in time in the future? Or are we talking about a future cohort? The various possibilities seem to have different implications and different people could be "right" depending on how they frame the question.
Bob Murphy, who insisted that Baker, Landsburg, Krugman, Callahan, and I were crazy back then, had a masterful summary of the arguments here which (if I remember correctly) captures my view of things very well. It really is one of my favorite blog posts of all time. If you want to get back up to speed, make sure you read that.
Krugman and Baker's point back then is that if you think of future GDP, the only real cost was distributional.
DeLong's contribution now, as I read it, is that the other potential cost is crowding out of private investment. That wasn't even considered in the OLG models we were running through. That's probably fine, because it's hard to convince most people that crowding out presents a real cost for thinking about the deficit today (long-term problems are another story, of course).
I don't plan on jumping into this anymore except to re-present what I think the meaningful contours of the discussion are.
I stopped paying attention to this debate last time around because basically everyone seemed to take for granted the Ricardian Equivalence assumption of no growth.
ReplyDeleteTwo thoughts:
Delete1. Growth seems like a practical concern. Clearly with growth the point where you worry about deficits is pushed out a little. But there is still a point where you may worry about it when it outpaces growth. If we all recognize that, we just hold growth constant rather than footnoting the whole discussion with references to the growth rate.
2. You still have to think counterfactually. Even given the previous point that growth is more of a practical concern about when we should start worrying about net costs rather than a reason not to worry about net costs, debt can still impost gross costs.
In discussions of what the deficit should actually be, I absolutely agree we need to talk about growth. In discussions of how meaningful the phrase "we owe it to ourselves" is, I think we can fruitfully abstract away from it.
Well, historically the US has never reached any (sustained) point where debt growth outpaces output growth, so its not just a matter of making a casual simplifying assumption its about making a fundamental assumption about the trajectory of the debt burden. Which, btw, unfairly favors the Nick Rowe/Barro/monetary>fiscal/budget hawk etc narrative.
DeleteAs for your second point, that dovetails with the other handicapping assumption that transfers are simply transfers (taking apples from peter to give apples to paul)and not part of a circular income/output system.
Anyway, what (nondistriburtional) gross costs does debt impose?
Thanks for the props, Daniel. Looking back on that post, I can't believe how much time I spent on it... That took up like a whole weekend I think.
ReplyDeleteI have to say Bob, I've been reading econ blogs for about half a decade and that particular post of yours is my favorite so far. I had to stop reading it at work because I couldn't stop laughing and it really helped clarify some of the points of the debate which I had not gotten up to that point.
Delete"Last time this came up I took my usual cowardly "you all have a point" position. I felt that people were talking past each other in what they meant when they referred to "generations" or "the future". "
ReplyDeleteThat was where I left off too.
http://jpkoning.blogspot.ca/2012/01/debt-generations-savings-and-economic.html
I refuse to wade back into that debate. Tell me if anything new pops up.
JP: nothing new has popped up yet. Your old post still reads very well. It's the Borges Problem. Do we categorise the future by time periods, or by cohorts?
DeleteOooh I didn't know it had a name.
DeleteThe fact that someone else considers this a real facet of the debate makes me more confident in having made the point.
Yep, the Borges problem. Also useful for resolving differences with spouses, as I've since discovered.
ReplyDeleteThe debate on this has been going on nigh two-hundred years, so it is probably insusceptible of resolution.
ReplyDeleteThank you for pointing me to Bob Murphy's summary! He is really good at writing humorous dialogs.