> ...these strong similarities are not what these authors choose to emphasize.
This statement feels misleading, in that the similarities do not appear to be as significant as they are implying. The table they provide very clearly shows a significant difference between their 2010 findings and the HAP paper (-0.1 vs. 2.2 mean, a gap of over 2 percentage points).
> These results are, in fact, of a similar order of magnitude to the detailed country by country results we present in table 1 of the AER paper
Taken literally, this makes no sense. No one has ever claimed the HAP results were of a differing order of magnitude to their own.
Given that this is a rebuttal to a research paper where numerical values are in dispute, I would have thought they would use the term ("order of magnitude") precisely, and not colloquially.
> It is utterly misleading to speak of a 1% growth differential that lasts 10-25 years as small.
Perhaps, but if a 1% growth differential is significant (HAP), then a 2.9% differential is massive (2010 RR), compounding the scale of their error.
After all, their original 2010 work implied a mean 2.9% drop in the growth rate once debt climbed above 90%. That paints a significantly different picture than a decrease of 1 percentage point, resulting net positive growth rate (2.2%) as opposed to negative (-0.1%).
While this is speculative on my part, I feel it's safe to say that had they originally reported these (allegedly) "very similar" numbers ( > 2% vs -0.1 %), their findings would not have gotten the same wide circulation in certain political circles that it did. Given that context, the whole "...but the numbers are kinda close..." argument falls a little flat IMHO.
Could one even be off by an order of magnitude with economic growth numbers? On a base of 2.2, that requires being off by 20%. You could never be off by an order of magnitude because economies don't grow or shrink by 20% pretty much ever.
The original paper concluded that debt overhang above 90% of GDP would result on growth of -0.1%. This was based on three serious flaws – 2 of methodology, one of spreadsheet programming. The corrected figure is 2.2%. To pretend that this is not a big deal is disingenuous considering how widely the 2010 paper was cited. It’s true that the 2012 paper is more accurate and also that higher debt leads to lower growth, but to gloss over the severe flaws of the earlier paper is misleading. I would respect Reinhart and Rogoff a lot more for simply acknowledging the errors and repudiating their previous conclusion; instead they essentially argue their errors ought to have been caught earlier.
This is an unsatisfying response. It doesn't answer the question of why Reinhart and Rogoff chose to exclude certain high-debt years for certain countries or why they chose the particular weighting mechanism that the critics found problematic.
It is a response and that is all that is needed. The target audience for this just needed any rebuttal to paste into their streams to convince themselves that no adjustment of beliefs is required.
While I agree that it doesn't seem to be an adequate response (the whole -0.1 to 2.2 difference looks problematic enough), they claim that their 2012 paper addresses the weighting mechanism.
So we'd need to read that paper too to then argue if it somewhat clarifies their findings.
It's unsatisfying because it's a bunch of smoke and mirrors. They didn't actually address the criticisms. It screams politics more than academics at this point.
Dean Baker's critique of this response is typically incisive:
[Herndon, Ash, and Pollin] found growth was slower in periods with debt levels above 90 percent of GDP than below, but the gap was relatively small and nowhere close to statistically significant. Furthermore, they found a much bigger gap in growth rates around debt-to-GDP ratios of 30 percent. If we think that [Reinhart and Rogoff's] methodology is telling us something important about the world then the take-away should be that we want to keep debt-to-GDP ratios below 30 percent.
Note there is no entry in a debt-to-GDP ratio for assets, just liabilities. So if we believe the R&R story, then we can increase the growth rate through [auctioning off the California coastline].
People are fixated on the lurid detail of the story; do the Google search and see the "bad numbers! Excel spreadsheet! bad numbers!" roll by. The real issue, which Matt Yglesias pointed out last year, is that the purported result is hard to square with reality. Japan has the highest debt ratio in the world. If they disposed of it tomorrow, they would not suddenly be the fastest growing economy.
[+] [-] doktrin|13 years ago|reply
This statement feels misleading, in that the similarities do not appear to be as significant as they are implying. The table they provide very clearly shows a significant difference between their 2010 findings and the HAP paper (-0.1 vs. 2.2 mean, a gap of over 2 percentage points).
> These results are, in fact, of a similar order of magnitude to the detailed country by country results we present in table 1 of the AER paper
Taken literally, this makes no sense. No one has ever claimed the HAP results were of a differing order of magnitude to their own.
Given that this is a rebuttal to a research paper where numerical values are in dispute, I would have thought they would use the term ("order of magnitude") precisely, and not colloquially.
> It is utterly misleading to speak of a 1% growth differential that lasts 10-25 years as small.
Perhaps, but if a 1% growth differential is significant (HAP), then a 2.9% differential is massive (2010 RR), compounding the scale of their error.
After all, their original 2010 work implied a mean 2.9% drop in the growth rate once debt climbed above 90%. That paints a significantly different picture than a decrease of 1 percentage point, resulting net positive growth rate (2.2%) as opposed to negative (-0.1%).
While this is speculative on my part, I feel it's safe to say that had they originally reported these (allegedly) "very similar" numbers ( > 2% vs -0.1 %), their findings would not have gotten the same wide circulation in certain political circles that it did. Given that context, the whole "...but the numbers are kinda close..." argument falls a little flat IMHO.
[+] [-] jamesaguilar|13 years ago|reply
[+] [-] anigbrowl|13 years ago|reply
The original paper concluded that debt overhang above 90% of GDP would result on growth of -0.1%. This was based on three serious flaws – 2 of methodology, one of spreadsheet programming. The corrected figure is 2.2%. To pretend that this is not a big deal is disingenuous considering how widely the 2010 paper was cited. It’s true that the 2012 paper is more accurate and also that higher debt leads to lower growth, but to gloss over the severe flaws of the earlier paper is misleading. I would respect Reinhart and Rogoff a lot more for simply acknowledging the errors and repudiating their previous conclusion; instead they essentially argue their errors ought to have been caught earlier.
[+] [-] gruseom|13 years ago|reply
Or that lower growth leads to higher debt. See Krugman's chart at http://krugman.blogs.nytimes.com/2013/04/16/reinhart-rogoff-....
[+] [-] GabrielF00|13 years ago|reply
[+] [-] jeremyjh|13 years ago|reply
[+] [-] tellarin|13 years ago|reply
So we'd need to read that paper too to then argue if it somewhat clarifies their findings.
[+] [-] ohashi|13 years ago|reply
[+] [-] gruseom|13 years ago|reply
[Herndon, Ash, and Pollin] found growth was slower in periods with debt levels above 90 percent of GDP than below, but the gap was relatively small and nowhere close to statistically significant. Furthermore, they found a much bigger gap in growth rates around debt-to-GDP ratios of 30 percent. If we think that [Reinhart and Rogoff's] methodology is telling us something important about the world then the take-away should be that we want to keep debt-to-GDP ratios below 30 percent.
http://www.cepr.net/index.php/blogs/beat-the-press/quick-tho...
[+] [-] tptacek|13 years ago|reply
Note there is no entry in a debt-to-GDP ratio for assets, just liabilities. So if we believe the R&R story, then we can increase the growth rate through [auctioning off the California coastline].
People are fixated on the lurid detail of the story; do the Google search and see the "bad numbers! Excel spreadsheet! bad numbers!" roll by. The real issue, which Matt Yglesias pointed out last year, is that the purported result is hard to square with reality. Japan has the highest debt ratio in the world. If they disposed of it tomorrow, they would not suddenly be the fastest growing economy.
[+] [-] jacquesm|13 years ago|reply