Gross Capital Formation and Improved Estimates of Real Gross and Net Capital Stocks to 1990 for the New Zealand Non-Market Production Sector.

Abstract

This paper presents estimates of real capital formation, and real gross and net capital stocks, for the New Zealand public non-market services sector (NZSNA sectors 22 and 23) the major components of which are in the areas of public health, education, welfare, administration and defense, and road and harbour infrastructure. Gross capital formation and capital stock time series for New Zealand are largely incomplete except for those prepared for the period 1950-90 by Bryan Philpott as a part of the Research Project on Economic Planning. This paper builds upon this work by extending the capital formation series of Philpott (1992-b) backwards through time allowing improvements to be made to the retirement estimates that are one of the components of the post-1949 capital stock series of Philpott (1992-a). Gross capital formation series are an important pre-requisite of capital stock estimation by the perpetual inventory method. If empirical information on the pattern of asset retirement within the period of investigation is incomplete, or a capital stock valuation within the desired time period of the capital stock estimation is not available, the capital formation series needs to be taken back to the formation time of the longest-surviving element in the capital stock at the commencement of the stock series. This also defines the time period of relevant price indexes, the second requisite of the perpetual inventory model. A non-trivial part of the road infrastructure capital stock is considered to have a near-eternal life, necessitating the estimation of a gross capital formation series at the commencement of European-based settlement. Because of incomplete empirical information on public capital expenditures and on the course of prices of capital assets prior to c. 1914 estimates of real gross capital formation prior to c 1914 can be expected to be more problematic than those after this time. Stock point estimates are compared with other capital valuations utilising different estimation procedures which indicate that the magnitude of error is likely to be small.