Pension payments and receipts by New Zealand birth cohorts, 1916–1986

Abstract

This paper analyses how demographic change has affected the costs and benefits that New Zealand's pay-as-you-go retirement income scheme has provided or will provide to cohorts born between 1916 and 1986. The analysis is based on historic data and Statistics New Zealand projections of future population trends. The results show that previous rapid population growth means cohorts born prior to 1980 can expect to pay half as much as they can expect to get in retirement benefits, but the costs of the scheme are getting larger and larger for subsequent cohorts.

Keywords:

• Retirement incomes
• intergenerational transfers
• government pension schemes

JEL codes:

• E24
• H55

Acknowledgements

I would like to thank Matthew Bell, Arthur Grimes, Tui Head, Malcolm Menzies, and Peter Neilson for helpful discussions and comments on this paper. The paper was written while the author was consulting for the Financial Services Council (New Zealand), but they did not fund this research.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. New Zealand has had a universal government retirement income scheme since 1940. The current scheme, New Zealand Superannuation, provides a retirement income (or pension) to all people over age 65 meeting a residency requirement. The pension does not vary with age, income, or previous contributions, but it is taxed. A brief history of the schemes is provided in Section 2.1.

2. Standard descriptions of this literature include Diamond (1997), Feldstein and Liebman (2002), de La Croix and Michel (2002), and Lindbeck and Persson (2003). For a contrary position, see Barr (2001, 2002).

3. Sinn (2000) provides an elegant exposition of the argument.

4. Coleman (2014) estimates how the opportunity cost of a pay-as-you-go and a save-as-you-go funded retirement income scheme paying benefits similar to those paid by New Zealand Superannuation vary with the return to capital, the growth rate of the size of each cohort, the growth rate of incomes and retirement income benefits, and the average length of time each cohort receives a benefit. The opportunity cost of a pay-as-you schemes varies from 17% of the cost of the save-as-you-go scheme when the real rate of return is 3%, the population growth rate is 1% per year, and the growth rate of incomes is 1.5%, to 257% if the real return to capital is 5%, the population growth rate is 0%, and the growth rate of incomes is 1.2%.

5. See, for instance, Diamond (1997), Feldstein and Samwick (1997), Feldstein (2005), or, in a New Zealand context, Littlewood (2010).

6. For additional details of the history until 1990, see the New Zealand Official Yearbook 1965, section 6, the New Zealand Official Yearbook 1990, chapter 7, and the Periodic Report Group (1997).

7. Between 0.5% and 2% of people old enough to qualify for Government Superannuation earned a sufficiently large amount that their entitlement was fully abated. See the Appendix for more details.

8. In 1975, the Labour government made membership of a superannuation scheme compulsory, and established the New Zealand Superannuation Fund to receive contributions from people not contributing to a private scheme. The initial contributions were modest, and the scheme was abolished and replaced by National Superannuation in 1977.

9. In theory it would be possible to calculate the average costs and benefits of New Zealand Superannuation for New Zealand born or foreign born people who spent any period of their lives as New Zealand residents. These calculations would be extremely difficult to make given that the incomes of New Zealand born and foreign born people are not distinguished in the census data. The calculations made in the paper calculate the average of these costs and benefits for New Zealand born and foreign born members of each cohort, weighted by the number of years each group spends in New Zealand. Consequently they can be interpreted as the costs and benefits for each cohort of New Zealand residents, irrespective of birth place.

10. The numbers are interpolated using data from the 1936 and 1951 censuses.

11. The numbers in Table 1 are not adjusted for the surcharge tax 1984–1998. During this time the following numbers of each cohort are estimated to have earned so much that their retirement income was fully abated: 1916–2900; 1921–4800; 1926–10900; 1931–6100. These reduce the receipts by between 0.5% and 1.7% for these cohorts. See the appendix for more details.

12. The dependency ratio prior to 1951 is assumed to be 26%.

13. Hence Gross Domestic Product is 16% higher than Net Domestic Product.

14. The maximum income band in the census was reduced between 1981 and 1991 despite a 160% increase in the price level. This means the 1991 census significantly under-measures the income of high income people.

15. The ratio of female to male incomes 1981 was 35%, compared with 29% in 1976, 55% in 1986, and 59% in 1996.