Production functions and the measurement of productivity : an econometric approach - New Zealand as a case study

Abstract

The study of Solow (1957) stimulated debate among economists on the role of technical change in economic growth. The consensus has been that technological change accounts for a significant proportion of gross national product (GNP) growth in industrialised economies.

This study empirically measures the performance of productivity in New Zealand's economy for the 1950-1999 period, using an econometric approach. Over recent years, there have been many empirical papers trying to measure productivity. All these studies adopt different approaches but the findings indicate that the performance of productivity in New Zealand has been very poor. However, none of these studies allows the process of technological progress to be modelled explicitly, as suggested by Griliches (1988), by including human capital as a distinct factor of production.

"Time series" data based on the aggregate economy are used to develop an econometric model that captures the dynamic interaction between GDP, fixed capital, labour employed and human capital. A vector error correction model derived from the vector auto-regression is used to estimate the long run elasticities of a Cobb-Douglas-type production function augmented to include human capital as a distinct factor of production. Estimates suggest that labour employed, fixed capital and human capital are all important determinants of "long run" productivity growth, with labour employed being the dominant source of productivity improvement. The findings from the impulse response and forecast error variance decomposition analyses suggest that the response of fixed capital to GDP shock is remarkably short-lived and it is only labour employed and human capital that contributes to a significant proportion of GDP's forecast error.

The analysis further demonstrates that the inclusion of human capital as a separate factor in the productivity analysis better explains its performance than the two-factor production function. These findings support the argument that human capital needs to be given greater emphasis than fixed capital in determining long run productivity growth in New Zealand. Hence its omission from productivity analysis will lead to unreliable results, which may mislead policy formulation, planning and budgeting decisions. This study finds the performance of productivity in New Zealand over the period considered to be very poor. This was attributed mainly to capital productivity which had a declining trend throughout the period considered.