Income Distribution in New Zealand

By George Barker, Institute of Policy Studies, 1996

Des O'Dea
Consultant Economist
PO Box 19,253
Hamilton

Introduction

1996 is the United Nations International Year for the Eradication of Poverty. It is a reflection on economists that probably most of us in the profession are quite unaware of that fact.

Of course many economists are interested in "equity" and income distribution issues, of which poverty issues are but one part. Much good work has been done on these, and a substantial literature has developed in New Zealand over the past 15 years. Major works were Easton (1983), NZ Planning Council (1988 and 1990), and Statistics New Zealand (1990) – the latter two influenced the work of Suzanne Snively and her research team at DSW in the late 1980s on measuring the redistributive impacts of the government's budget, and developing appropriate concepts and estimation techniques to do so. And there have been other significant contributions – for example from Treasury, addressing income adequacy issues (Brashares 1993); at the Social Policy Agency; and at Massey University (e.g. Chatterjee 1995).

For a period in the early 1990s, however, there was something of a lull in work on distributional issues. Not entirely, of course. Stephens, Waldegrave and Frater began their ground-breaking work on new ways of measuring poverty (1995). Microsimulation models were developed by Business and Economic Research Ltd for analysing wealth accumulation and distribution (Strombergen 1995). But on the whole, and with the 1991 demise of the Planning council, there seemed less interest, at least in official circles, in research on distributional issues. Perhaps it would be more accurate to say that there was less official willingness to fund such work.

Now, there seems to be some revival of interest in the field. The New Zealand economy has emerged from a decade's reforms, indubitably in sounder shape and with its long-term growth prospects much improved. One natural tendency as we draw breath, figuratively speaking, is to look back and try to assess how the sacrifices of the reform years were distributed. A recent contribution by Easton (1996) does this. A second is to draw lessons from the experience for the future. Are there worrying inequities, now or in prospect? How do we decide what is inequitable? And, most important, what are the underlying factors which generate a more, or less, equitable distribution of resources and, when we have identified them, what if any are the policy implications?

George Barker's book reviewed here starts to address these questions. It should be added that it might better be called a booklet, the main text amounting to only about 40 pages. Such brevity in this technically complex area is commendable.
 

The Research Programe

The book is a first report on work barker and others commenced at Treasury, and which he is now continuing as a research programme under the auspices of the Institute of Policy Studies at Victoria University of Wellington. Some will remember an earlier presentation of parts of the work at the Association of Economists conference at Lincoln last year.

The aim of the book is to lay some groundwork for a forthcoming series of publications on the research programme, and also to present some preliminary results. The aim of the series is to "make more widely accessible a new view that has emerged in the academic and professional community on income distribution over the past 10 to 20 years", leading to better informed debate "which draws especially on a more dynamic analysis or perspective" (p. 1).

The most interesting chapter is that (Chapter 2) in which "The New View on Income Distribution" is discussed. I had better say immediately that I am not really convinced that there is much here that is very new. My comments may, however, be unfair to the author, in that I have not read many of the references he cites in his discussion. He has clearly done a substantial amount of reading in the more philosophical literature on "fairness" and "equality", from Rawls to Friedman, and also in "human capital" theory.

The "new view" consists of "new data" and "new theory", the latter with both normative and positive components. The new data consist of longitudinal or panel data-sets, tracking individuals' incomes over a number of time-periods. Such data are now available in a number of countries. The point Barker makes here is that using such data (or simulated data in some instances) one finds that income inequality is less for income cumulated over a number of years, or over a lifetime, that it is for cross-sectional data a year at a time, as used in the New Zealand analyses referred to above. This is because, typically, a large proportion of individuals in the lowest income quantile in a given year will next year be in a higher quantile (and likewise for other inter-quantile movements). That is, there is a considerable degree of income mobility in the population.

I think this is indisputable, but also something that has been recognised for many years. I very much agree with the author's viewpoint that, as and when such data become available, they should be used to carry out "dynamic" analyses of income distribution in addition to the "static" analyses now carried out on an annual cross-sectional data, and that we will learn a great deal from dynamic analysis. The author goes on to imply that static snapshots showing an increase in inequality are not a proper basis for policy recommendations. This is because changes in income mobility from year to year may have increased "static" inequality, but not "dynamic" inequality, if I have read him correctly.

Here I part company with the author. It would certainly be possible to construct models in which income mobility increases in ways which give the effects he postulates. To say that it is possible, however, is not to say that it actually happens. I find it difficult to believe that a significant increase in, say, either the level of unemployment or the rate of marital breakdown will not lead to a consequent significant increase in income inequality, on either cross-sectional or longitudinal measures. In most real life cases I believe that "cross-sectional" and "longitudinal" measures of inequality will move in the same direction. The burden of proof is on the author to construct realistic models, or exhibit real data, which show the converse. It must be said that this would be a fascinating research project.

I turn now to the "new theory". The discussion here is fairly general, with wide-ranging lists of the outcomes to be taken into account when assessing whether a given distribution is better or worse than another, and also of the influences which help explain how income distributions emerge. A noticeable feature is that the author is trying to integrate into his theoretical framework factors explaining the "size of the cake" as well as "how the cake is divided. That is efficiency and growth as well as equity. This is ambitious, perhaps too much so. But one can see that in a dynamic analysis framework, the attempt at formally integrating efficiency and equity objectives does make more sense than in the more traditional kinds of analysis.

Quite a lot of emphasis is put throughout on human capital determinants. Its importance is undeniable, but the dismissal of property income as playing empirically only a small role in the overall outcome seems too dismissive, at least without citing evidence to the contrary. The impact of property income from sources not covered by the standard household surveys, such as capital gains and imputed home-ownership rental income, can be substantial and is often overlooked (see NZPC 1990).
 

Descriptive Material for the 1980s and 1990s

From this chapter on theory, the author then turns to descriptive material. Mostly this is from the standard statistical sources, or is derived from the work of others on income distribution. At this stage the author's research does not appear to have encompassed any new statistical measurement of income distribution.

The material discusses first the available "dynamic" data published on income mobility – there is not much for New Zealand – and lifecycle earnings profiles. Next are sections on changes in the inequality of personal and household income distributions in recent decades, and where and why income fell (for most people) over the 1987-1991 and post-1991 periods. There are charts with the customary breakdowns by gender, age, ethnicity, and educational experience.

It must be said that this descriptive material is disappointingly thin. It is not without interest, but contains little novel information. The level of discussion suffers by comparison with that in Easton (1996).

Specimen points that concerned me were:

Lack of knowledge of some of the data peculiarities in the various series. For example, that census income definitions changed in 1981 (p. 11), as discussed in Easton (1996), also the use of OECD measures of New Zealand's unemployment rate (p. 24), where census rates would have given a better historical perspective. One wonders if there are not similar faults in some of the other time-series cited from a number of sources. This is a research area where the advice of a statistician with expert knowledge of the data sources is essential.

Sometimes uncertainty as to whether the population being graphed was, for example, the whole adult population, or just those in employment.

The lack of any discussion in the sections on household disposable income of the potential impact of changes in tax rates and benefit levels, particularly the benefit cuts of 1991.

In a discussion of the lower earnings of Maori, a comment is mad that part of the reason could be that Maori tend to be younger. Fair enough, but one would then expect that an attempt would be made to standardise by age to get an undistorted comparison. Perhaps this is the sort of more detailed analysis to be expected in subsequent work.

A lack of attention to the contribution that gender shifts could be making to changes in income inequality – for example more females in the labour force with average income growing relative to average male income. Again Easton has some interesting comments on this point.

The restriction of the analyses to "market income" and "disposable income". The important development resulting from Snively's work in the late 1980s, and included in the other major reports of the time, was the concept of "final income", allocating all government expenditures and receipts to households. The estimation was complex, but the results important in that benefits to households from government expenditures other than social welfare – for example on education and health – were specifically taken into account. The effect was to significantly reduce overall inequality. I would hope that Barker's research programme is able to incorporate these broader income concepts in its subsequent reports.

A number of the criticisms above will appear to be rather tedious points of detail. Unfortunately, in income distribution analysis this sort of attention to detail is necessary.

A definite plus to the report, I should add, is a data section giving all the data used for the charts. There are a few things more frustrating than trying to access data which have been charted, but not tabulated. The old Planning Council was a particular sinner in this regard.

Conclusion

The project Barker is undertaking is an important one, and one wishes him well in his endeavours. I would have wished to see more indication of the contents of the reports he is intending to produce. However, the results are likely to be exciting, especially if he can develop the "dynamic" analysis he aspires to.

Judging from the problems listed above, however, I suspect that his research programme may be currently under-resourced. At a minimum, future reports from him and his team would benefit from ongoing comment and review, more than seems to have been the case for this first publication, by those who have detailed knowledge of the data sources, of the fish-hooks hidden in the data, and of how these were overcome in earlier work. Given this support, I look forward to a series of interesting reports from his research programme.

References

Brashares, E. (1993) "Assessing Income Adequacy in New Zealand" New Zealand Economic Papers 27 (2) 185-207.

Chatterjee, S. and L. Tobisson (1995) "Growing Apart: Aspects of Personal Income Distribution in New Zealand in the 1980s" paper presented to the New Zealand Economists Conference, Lincoln, August.

Easton, Brian (1993) "Income Distribution in New Zealand" New Zealand Institute of Economic Research, Research paper 28, Wellington.

Easton, Brian (1996) "Income Distribution" Chapter 4 in Brian Silverstone, Alan Bollard and Ralph Lattimore (eds.) A Study of Economic Reform: The Case of New Zealand, Elsevier Science B.V.

New Zealand Planning Council (NZPC) (1988) For Richer or Poorer: Income and Wealth in New Zealand, New Zealand Planning Council, Wellington.

New Zealand Planning Council (NZPC) (1990) Who Gets What? The Distribution of Income and Wealth in New Zealand, New Zealand Planning Council, Income Distribution Group, Wellington.

Statistics New Zealand (1990) The Fiscal Impact on Income Distribution 1987/88, Government Printer, Wellington.

Stephens, R., C. Waldegrave and P. Frater (1995) "Measuring Poverty in New Zealand" Social Policy Journal of New Zealand, Issue Five, December.

Stroombergen A., D. Rose and J. Miller (1995) Wealth Accumulation and Distribution: Analysis with a Dynamic Microsimulation Model, Business and Economic Research Limited, Wellington, August.