CEPR - Center for Economic and Policy Research


En Español

Em Português

Other Languages

Home Publications Blogs Beat the Press Failed Rescue Attempt of Reinhart-Rogoff

Failed Rescue Attempt of Reinhart-Rogoff

Wednesday, 01 May 2013 09:59

In a Bloomberg column, Matthew Klein gives a brief discussion of errors in New Zealand GDP numbers that he claims supports Reinhart-Rogoff over their critics at the University of Massachusetts, Thomas Herndon, Michael Ash, and Robert Pollin (HAP). However the numbers he gives do not support his case.

He claims that the UMass trio relied on mistaken numbers from the Maddison Project, which is generally taken as an authoritative source on GDP. However, Klein claims that the data given for New Zealand for the years in question is mistaken. After quoting from the UMass paper he tells readers:

"That sounds pretty damning but it turns out that on the New Zealand question, at least, it's Reinhart and Rogoff who are right, and Herndon, Ash and Pollin are wrong."

Klein refers us to data from Statistics New Zealand for 1946-1950, the five years in question:

"According to Statistics New Zealand, the economy grew by 3.0 percent in 1946, 0.4 percent in 1947, and 3.2 percent in 1948. New Zealand's GDP shrunk by 5 percent in 1949 and then grew by 5 percent in 1950."

Let's see that comes to an average growth rate for these five years of 1.32 percent. This is somewhat less than the 2.58 percent average that HAP got using the Maddison Project data, but Klein's number is far closer to the HAP average than the -7.9 percent growth figure for New Zealand that Reinhart and Rogoff used in their calculations.

It's not clear how this shows Reinhart and Rogoff are right about New Zealand and Herndon, Ash, and Pollin are wrong.



I just checked with Michael Ash and he confirmed that the data in their paper all came from the Reinhart-Rogoff spreadsheet. Insofar as this data is wrong, the fault lies with Reinhart and Rogoff, not Herndon, Ash, and Pollin.

Comments (9)Add Comment
E-Z logic
written by Squeezed turnip, May 01, 2013 11:10
Dean, it's simple. HAP were wrong. Therefore RR were right. See? Klein's mastery of inductive logic is simply stunning.
RR also used Maddison - sort of
written by JimV, May 01, 2013 11:25
There are three excellent posts about the RR fiasco at "Meltdown Economics". The latest one, here


reviews the economic data available for New Zealand from three sources and concludes RR's data came from Maddison, with a couple unexplained discrepancies in the early years (which were not used in RR's "90%" analysis anyway). The blogger shows in earlier posts that no matter which available data source you use, RR's conclusions don't hold water.

(I found the website on a link from Brad DeLong's blog.)
written by Chris, May 01, 2013 11:28
RR were and are so fundamental to the argument about debt that anything that suggests they are wrong is intolerable to the debt fanatics. Hence the repeated attempts to revalidate their wrong conclusions and distract people from their errors. Debt fanaticism has a life of its own; debt is disaster and must be eliminated NOW at any cost.
Reverse Explanation in Economics
written by Hugh Sansom, May 01, 2013 12:41
Scientists and philosophers talk about the direction of causation. What's the direction of causation in economic explanations? As a couple of commenters here have noted, the austerity bombers, deficit hawks (whatever) are fanatically attached to keeping inflation below 2 percent (or even at zero, as Alan Greenspan used to want). That is the causal starting point for them — an article of faith. So any facts must conform to that religious tenet. Somehow or other, Reinhart and Rogoff must be right, whatever the facts. It's very interesting to see where in the spectrum of economic thinking we find more of this dogmatic fanaticism. I think it's clearly on the right, conservative end of the spectrum (which includes much of the so-called center, these days).
Doesn't Replacing (-7.6%) with (+15.6%) Completely Disprove R&R?
written by Mark A. Sadowski, May 01, 2013 2:13
Matthew C. Klein notes the following:

"Historians of New Zealand know that there was a boom in 1951 as a result of the U.S. Army's demand for wool to make uniforms during the Korean War. According to Statistics New Zealand, the real economy expanded by about 16 percent in just that one year. Output contracted in the following years as the demand for wool subsided."

If you think back to Thomas Herndon, Peter Ash, and Robert Pollin's (HAP) original analysis, they pointed out that the unusual weighting meant that the only year that New Zealand was above 90% of GDP in terms of public debt was 1951. And according to Reinhart and Rogoff's dataset real GDP (RGDP) declined by 7.6% in 1951. But according to Statistics New Zealand (SNZ) the correct figure is *positive* 15.6%:


Now go back to HAP and look at Table 2:


Now do the arithmetic. Add together the average growth figures for the seven countries during the time that they had public debt over 90% of GDP but this time replace (-7.6%) with (+15.6%) for New Zealand and what do you find? Instead of an average RGDP growth rate of (-0.1%) for countries with debt over 90% of GDP you find that they had an average RGDP growth rate of (+3.3%).

Now compare that to the results from Reinhart and Rogoff's paper which are listed in Table 3 of HAP (RR 2010b Appendix Table 1). For countries with public debt less than 30% of GDP the average RGDP growth was 4.1% and for countries with public debt between 30% and 60% of GDP, and between 60% and 90% of GDP, the average RGDP growth was 2.8%. Thus the average RGDP growth rate of countries with public debt over 90% is *higher* than than every category except the lowest category.
R&R reminds me of F&P (Fleischmann&Pons) or cold fusion
written by Anna, May 01, 2013 3:53
And I think some are still trying to make the original research reports true after over 20 years. You see, there is a lot of hanging onto things people just want to be true. The new data is always just around the corner.
Live In Building Superintendent
written by dilbert dogbert, May 01, 2013 9:24
If RR and HAP are sensitive to one data point, New Zeeland, then this effort is worthless.
Only the Chief Engineer can graph the curve using one data point.
R-R should do academia a favor and describe their "working paper" process
written by John Wright, May 02, 2013 8:17
I would like R-R to do a post-mortem and describe what was their longer term intent of their "working paper 15639".

This is available at http://www.nber.org/papers/w15639 or as a magazine article in the May 2010 American Economic Review as "Growth in a Time of Debt".

The dictionary definition for "working paper" that seems to apply is "a tentative statement prepared to serve as a basis for discussion or negotiation".

On the front page of the paper the authors thank Olivier Jeanne and Vincent R. Reinhart for helpful comments.

Does this imply Olivier Jeanne and Vincent Reinhart read and vetted the paper?

Did R-R have any research assistants that checked the paper before submission?

The media should interview all the associated professionals that checked the paper before submission to NBER.

And when does a working paper morph into something else?
Is it after peer review?

The R-R concluding remarks are clear, "Our main finding is that across both advanced countries and emerging markets, high debt/GDP levels (90 percent and above) are associated with notably lower growth outcomes."

This is a great opportunity for the media to investigate the actual process that led to an influential, but tentative "working paper" that remained uncorrected for 3 years.
Emeritus Prof of Pol Sci
written by AJ Oliver, May 02, 2013 9:59
It's also absurd to use statistical significance with a small, non-random sample. That whole pile of mierda is rife with fraud.
Keep the pressure on. R&R deserve termination, which leftie Ward Churchill of the U. of Colorado got for doing a WHOLE lot less.

Write comment

(Only one link allowed per comment)

This content has been locked. You can no longer post any comments.


Support this blog, donate
Combined Federal Campaign #79613

About Beat the Press

Dean Baker is co-director of the Center for Economic and Policy Research in Washington, D.C. He is the author of several books, his latest being The End of Loser Liberalism: Making Markets Progressive. Read more about Dean.