The fall (and rise) of labour share in New Zealand

ABSTRACT

The share of national income going to labour in New Zealand fell substantially between the 1970s and the end of the century. Approximately half of this decline was then recovered in the following decade. In this paper, we argue that the decline from the mid-1980s onwards is due to public sector reforms. Corporatisation re-orientated the public trading enterprises away from a broad range of social and trading objectives towards generating profits, while increased fiscal discipline in non-market government departments reduced payroll costs. Consistent with this hypothesis, we show that most of the decline in aggregate labour share from the mid-1980s onwards can be attributed to a significant fall in the labour share of the public sector. To more formally analyse the effects of the reforms, we build a simple model of structural transition. The model yields several predictions that are consistent with observed trends in sectoral labour share. First, there is a large and permanent decline in public sector labour share after the reforms. Second, there is a smaller, short-run decline in private sector labour share that is reversed over the long run. The model can, therefore, explain not only the decline in aggregate labour share from the mid-1980s onwards; it can also explain the partial recovery in labour share beginning in 2002.

KEYWORDS:

• Labour share
• inequality
• economic reform
• state-owned enterprises

JEL Classification:

• D33
• L3
• N17
• O15

Acknowledgement

We thank Debasis Bandyopadhyay, Ian Duncan, Tim Hazledine and Steve Poletti for their comments. The views expressed herein are those of the authors and do not necessarily reflect the views of the Bureau of Economic Analysis or the Department of Commerce.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Note that gross operating surplus in the non-market sector is set equal to imputed depreciation, so that LS can only decline if compensation of employees falls, and assuming the reforms did not affect depreciation.

2. While the model does not include a government non-market sector, any decline in such employment would strengthen this result.

3. Specifically, to make it false would require that most rent sharing was done through work rules and private labor demand is very inelastic. The first element is required since fair wages do not affect government labor demand and private LS is moved by an influx former government workers. If labor demand is inelastic, wages must fall a great deal for the private sector to hire them. It is not obvious whether the first is true or not, particularly if there are former non-market government employees, but most estimates put labor elasticities somewhat below one making it unlikely for that condition to hold.

4. Note that the shift-share results are qualitatively unaffected if we instead use 1995 as the beginning date of the transition period, or if we use 1983 as the beginning date of the pre-reform period. See Tables A.3 and A.4.

5. The private market sector is of course an imperfect control for the public market sector, and thus the methodology is far from being a genuine natural experiment.

6. These data also indicate that earlier industrial relations legislation over the reform period, such as the Labour Relations Act 1987, had a minimal effect on union density between 1985 and early 1991.

Additional information

Funding

This work was supported by the University of Auckland Faculty Research Development Fund [grant number 3709796].