Keywords: Growth & Innovation; Statistics;

The tabulation below the Chart is the best available longrun series of New Zealand’s volume GDP, volume GDP per capita, with Notes following. The data was originally compiled for The Economic Development of New Zealand.

The units are as follows:
Volume GDP & Volume GDP per capita: 1859/60 = 1000. (This enables the data to be rebased to any chosen year. What is important is percentage difference over time matter.)
Population: ‘estimated population’ thousands.
For further details go to Notes

Volume GDP: 1860-2003

Go to top

Notes
The general principle has been to use the best quality available series.

1. GDP per capita and GDP:
The series are constructed by linking the following series:

1859/60-1916/7
Keith Rankin, “New Zealand’s Gross National Product: 1859-1939″, Review of Income and Wealth, Series 38, Number 1, March 1992. Series from the past are typically of poorer quality than more recent ones. The Rankin series uses the synthetic procedure of money multipliers, rather than direct estimates. It is subject to a number of further weaknesses or problems as follows.
– The Rankin series is for GNP not GDP. The above assumes that there was a constant ratio between the two.
– The Rankin series is for the Non-Maori economy only. The linkage of the non-Maori economy before 1917/18 with the Maori plus non-Maori economy after. assumes that earlier the Maori made the same contribution to GDP as they did in 1917/18.
– The Rankin series assumes that the CPI follows a similar pattern to GDEF, when in fact they dont over short periods. (There was no GDEF for Rankin.)

1917/18-1930/31
Brent Lineham, “New Zealand’s Gross domestic Product 198/38”, New Zealand Economic Papers, Vol 2, No2, 1968. The Lineham series is a nominal series constructed by adding the value added of each production sector, as is the standard practice in the SNA accounting system.
It was deflated by a GDEF reported in Brian Easton, “A GDP Deflator for New Zealand: 1913/14-1976/7”, Massey Economics Papers B9004, December 1990, pp.83-102. The Easton series is a (fixed) weighted combination of the available price series.
Details are also reported in Brian Easton, In Stormy Seas, University of Otago Press, 1997, p.299.

1931/32-1948/49
There is an official nominal National Income or GNP or GDP series going back (sometimes intermittently) to 1931/32, constructed from income statistics. From this a nominal GDP series can be constructed using the Easton GDP deflator (above).
Details are also reported in Brian Easton, In Stormy Seas, University of Otago Press, 1997, p.299.

1949/50-1953/54
The Economic Survey 1956 published a volume GDP series. There is no indication how it was constructed.

1954/55-2002/3
Statistics New Zealand has constructed volume GDP since 1954/55. There are a set series constructed by different methods (and using different prices as bases).
A particular problem is 1977/78 where there is a clear break in the series. At this lap year, the official series suggests a major economic contraction, which corresponds neither to the accounts of the day, nor in any of the official data except for a major fall in the value of inventories. Unfortunately the survey base for manufacturing inventories was changed in that year, and there was no lap. A careful review, reported in Brian Easton The 1977/78 year (unpublished) suggests that after adjusting for the over-reported inventory decline, the contraction was much smaller – of the order of 1.3 percent. The volume GDP data here and in elsewhere in my work is adjusted to this contraction. The change makes no difference to the general topology of the series, although it affects quantities, so the difference does not materially impact upon the overall argument.

Population
The data series corresponds to the official data series except that before 1917/18 it is the series for non-Maori only (that being the series the Rankin uses), but is linked to the all New Zealand series, in effect assuming that the Maori/non-Maori proportions were constant back to 1859/60.
The other complication is that a post-census survey after the 1991 Population Census found there had been an undercount of around 2.5 percent. Before 1991 Statistics New Zealand reported the ‘de facto’ population based on the censuses: after it reports the ‘estimated’ population which allows for the undercount. We have no idea how large the undercount was before 1991, so the entire ‘de facto’ series is increased by 2.5 years. (What else might one do?)
(As an aside, if New Zealand is the only country to allow for an undercount in its population estimates, this may have the effect of lowering its ranking.)

Go to top