google analytics tracking code

Monday, April 29, 2013

Reinhart-Rogoff vs. New Zealand 1951 Topsy-Turvy (historical-statistical nitty-gritty and postmodernist nightmare)

The more one descends into the nitty-gritty of historical reality the more one begins to experience postmodern scientific despair.

[Postedit 27 May 2013: For final summing up of these issues, see post Reinhart-Rogoff vs. New Zealand: Final Round?]

[Post-Editing Note: You need to read the last post on the centrality of 1951 New Zealand to the Reinhart-Rogoff result to understand the motivation for this post.]

I've calculated New Zealand real GDP growth rates for the period 1946-1951 based on three data series available from Statistics New Zealand, using the dating convention

growth rate (t) = 100*(GDP(t) - GDP(t-1))/GDP(t-1).

New Zealand Real GDP Growth Rates in %, Three Methods

Year Maddison Maddison Geary-Khamis $90 Stat NZ $91/92
1947 10.91 9.56 0.42
1948 -9.92 -11.69 3.17
1949 10.79 8.54 -5.00
1950 14.67 12.38 4.97
1951 -7.63 -9.49 15.60
average 3.76 1.86 3.83

None of these series corresponds to the data employed by Reinhart and Rogoff, although the Maddison data comes closest. [Postedit: see next post for partial data reconciliation!] Moreover, the three series are wildly inconsistent, even the two Maddison series, although the Maddison series at least fluctuate in step with one another. The Statistics NZ series seems to lag the other two by one year but still deviates significantly from them (the 1952 growth rate is -5.5%).

If RR had used the Statistics NZ  value of +15,60% for 1951 instead of the -7.6% one they did use, they would have obtained the exact opposite conclusion - that countries with debt ratios over 90% had significantly higher growth rates than lower indebted countries! That is the danger one runs by basing such a far-reaching conclusion on just one observation of the most minor country in the database. In any event, the gyrations in NZ's postwar growth experience had nothing whatsoever to do with its indebtedness but were rather driven by unprecedented industrial disputes and the Korean War wool export boom.

Now some nitty-gritty on the 1951 NZ waterfront lockout/strike.

  1. The waterfront dispute did not entirely close the ports. After locking out the waterside union workers, the government declared a state of emergency and used military troops to reestablish some measure of port activity (as well as in the coal mines).
  2. While it was the longest NZ industrial dispute, it was not the most violent one. That was the Great Strike of 1913.
  3. Support strikes were staged by coal miners, railroad and hydroelectric workers, frozen meat processors and in other sectors amounting to some 22,000 workers during much of the 151 day dispute.
  4. The government ultimately achieved its primary goal in provoking the dispute - the destruction of the waterside union, which had separated from the main labor federation, by branding them communists and blacklisting their leadership and activists.

1 comment: