1  Introduction

Although New Zealand per capita GDP growth has increased since the early 1990s, New Zealand’s per capita GDP is still below the OECD average. In 2001 New Zealand per capita GDP was about 84% of total OECD per capita GDP (The Treasury 2004). From a growth accounting perspective, improvements in per capita GDP can arise from increases in labour productivity and from labour utilisation, and therefore from increases in labour participation. The purpose of this paper is to evaluate the potential for further increases in labour participation to contribute to raising New Zealand per capita GDP and closing the gap between New Zealand and OECD per capita GDP.

The paper contributes to two ongoing policy debates in New Zealand that involve claims about the effect of increased labour force participation on GDP. The first debate concerns family policy. The claim is that measures such as free childcare that raise employment rates of young women could generate substantial increases in GDP. The second debate concerns broader economic strategy. The claim, made by commentators such as the OECD (2003a: 17), is that general increases in labour force participation could make a substantial contribution towards closing the income gap between New Zealand and wealthier OECD countries.[1]

Labour force participation measures the proportion of the population over 15 years of age who are either working (employed) or without paid work and actively seeking work (unemployed). Increased participation could affect GDP by increasing employment.

This paper investigates the plausibility of these claims by calculating the effect on GDP of hypothetical increases in participation. Increases in GDP can be expressed as the product of GDP per extra employee and the increase in employment. The paper derives values for “increase in employment” and “GDP per extra employee”, and then calculates the associated “increase in GDP”. We investigate what GDP would have been in 2001 if, instead of the actual participation rates, we had experienced a different, higher, set of rates. There are many ways to carry out such an exercise, involving varying tradeoffs between realism, tractability, and transparency.

We have chosen methods that sacrifice some realism in return for a relatively high degree of tractability and transparency. We set up hypothetical scenarios, based on New Zealand’s historical participation trends and on the experiences of comparable OECD countries. In the first scenario, which we call the “young women” scenario; participation rates for women aged 25-34 equal an “adjusted” average for the five countries, from a sample of 24 OECD countries, with the highest rates. The adjustment, which is described in more detail in Appendix 4, attempts to correct for the fact that most OECD countries, including New Zealand, include women on paid maternity leave in their labour force statistics. This scenario addresses the claim that policies to raise participation rates of young women have substantial growth pay-offs. In the second scenario, which we call the “overall” scenario, participation rates for the working-age population as a whole equal the average across the five countries in our sample with the highest rates. This scenario addresses the claim that higher participation overall could raise New Zealand’s income relative to other OECD countries.

Estimates of GDP per extra employee should allow for the possibility that the new employees work different hours, and have different productivity, from existing employees. The estimates should ideally incorporate factors such as learning on the job, the effect of taxes used to fund employment policies, and changes in the ratio between capital and labour. They should also allow for measurement artefacts: typically, some of the measured increase in GDP that occurs when employment rates increase simply reflects activities moving out of unmeasured home production (parents looking after children at home) into measured market production (childcare workers looking after children at crèches) (Waring 1988).

We do some of the things on this list. We allow for different hours and productivity, and we calculate the extra capital required to maintain constant capital to labour ratios. Our analysis of tax, learning on the job, and measurement artefacts is, however, confined to a few comments in Section 5.

The next section of the paper compares trends in New Zealand employment rates with trends across the OECD. Section 3 explains how we generate our scenarios and our assumptions about hours and productivity. Section 4 presents the results for the effects on GDP. The final section summarises the conclusions and discusses policy implications.