Intergenerational
sustainability of
Superannuation

impacts: the
New Zealand

Commissioned report prepared for the Commission for Financial
Capability’s 2019 Review of Retirement Income policies:
Term of Reference number 6 (TOR 6):
“An assessment of the impact of current retirement income
policies on current and future generations, with due
consideration given to the fiscal sustainability of current New
Zealand superannuation settings.”
Susan St John1 and Claire Dale2

1

Susan St John, Hon Associate Professor Economics, Director Retirement Policy and Research Centre
s.stjohn@auckland.ac.nz,
2
Dr M Claire Dale, Research Fellow, Retirement Policy and Research Centre, m.dale@auckland.ac.nz.

1

 Contents
1.

Introduction ................................................................................... 4

2.

New Zealand Superannuation – an overview ........................................ 6
Parameters of New Zealand Superannuation ............................................. 8

3.

Pre-retirement and post retirement cohorts ....................................... 12
Labour force participation of older people ............................................... 12
Low income, pre-retirement group ........................................................ 14
Declining rates of home ownership ........................................................ 16
Changing nature of work ..................................................................... 17
How well are those over 65+doing? ....................................................... 18
Summary .......................................................................................... 22

4.

Fiscal sustainability and intergenerational equity: current arrangements 23
Demography...................................................................................... 23
Fiscal pressures ................................................................................. 24
Role of prefunding .............................................................................. 26

5.

Age of eligibility issues ................................................................... 29
International context .......................................................................... 29
Case for raising the age in New Zealand ................................................. 31
New Zealand’s experience with raising the age ........................................ 34
How much is saved by raising the age? .................................................. 35

6.

Reduction in rates of NZS ............................................................... 37
Indexation ........................................................................................ 38
Rationalisation of different rates of NZS ................................................. 41

7.

Means testing and alternatives ........................................................ 43
The New Zealand Superannuation Grant- a blue skies proposition .............. 46
PIE tax regime ................................................................................... 50
Costing ............................................................................................. 50
Summary .......................................................................................... 52

8.

Notes on other measures to improve fiscal sustainability ..................... 54
Increasing residency requirements for NZS............................................. 54
Compulsory saving and tax concessions ................................................. 54
Growing GDP, raising taxes, prefunding ................................................. 55

9.

Conclusion ................................................................................... 56

References ........................................................................................... 56
Appendix NZSG costings
2

 Acknowledgments
The authors acknowledge assistance from the CFFC who commissioned this report from
the Retirement Policy and Research Centre. It is the work of the authors alone and does
not necessarily reflect the views of the CFFC.
The authors also thank Matthew Bell, Treasury NZ for assistance with the costings and
his review of this paper. This does not imply Treasury’s endorsement of all or any part of
this paper. They also thank Bryan Perry, Ministry of Social Development, for his
assistance with analysis of Household Economic Survey but again this does not imply any
endorsement by MSD. Any errors remain the responsibility of the authors alone. This
report is provided to the CFFC to inform the 2019 Review with due care but within the
limitation of data sources and is not to be used for other purposes without permission of
authors.

3

 1. Introduction
In addition to the enlarged burden of pension costs, a further consequence of
increasing longevity (and medical and technological advances) is that total
healthcare expenditure in OECD countries has climbed faster than GDP, at an
average annual rate of 2% over the past 50 years, and some estimates suggest
that by 2040 total expenditure could grow by another 50-100% as the number of
people living with Alzheimer's disease and other dementia is expected to almost
double to 65 million by 2030, with 72% of cases in the developing world. (OECD,
2017a; World Economic Forum, 2013).
For many years there has been growing concern in OECD countries that an ageing
population means rising pension and healthcare costs will put increasing pressure on
governments’ budgets. As the numbers of those aged over 85 years grow relative to the
older age group (65 years+) the expenditure on the older ‘old’ will take a growing share.
Some of that cost will be met privately but there will be fiscal implications and
intergenerational tensions around whose needs should be prioritised.
An additional concern in OECD countries is that ageing inequality is increasing and with it,
growing poverty issues especially for new retirees and women:
Growing disparities in labour market conditions will likely result in higher pensioner
poverty in the future. Moreover, many countries have already lowered pension
promises as a response to population ageing. Financing of pensions and long-term
caring is likely to become more difficult. Therefore, the sustained and broadly
shared improvement in the living standards of elderly people seen over past
decades may not continue in the future. (OECD, 2017b, p. 20).
New Zealand has a younger demographic profile than many other OECD countries but
faces similar tensions. The retirement income framework comprises a universal state
pension, New Zealand Superannuation (NZS) at age 65, and an auto-enrolment private
savings scheme, KiwiSaver. The amount spent on subsidising KiwiSaver attracts little
attention as it is relatively small, however the cost to the Crown accounts of NZS is often
viewed as ‘unsustainable’, or ‘unaffordable’ especially as the number over 65 swells
relative to those of working age. The growing numbers of older persons and the ageing of
the older population itself also portends a rapid growth in health and welfare expenditure
both in absolute terms and relative to other state spending. After years of belief that elder
poverty was a thing of the past in New Zealand, there is new evidence that for many lowincome people, the state pension is no longer enough to live on.
Some might view increasing fiscal costs as an inevitable consequence of demographic
change, but long-term projections suggest that something must give as the debt
projections under current settings appear fiscally explosive (see Table 9, discussed in
section 4). Some people are willing to let the inevitable adjustments just emerge over
time, while others suggest the answer lies in proactive changes that might require some

4

 kind of means testing or a rise in the age of access to NZS. Some put their faith in the
growth of GDP to mitigate the ‘burden’, or in growing the sovereign fund, the New Zealand
Superannuation Fund (NZSF).
This report first outlines the international context and the current parameters of NZS.
Section 3 assesses how well pre-retirement and post retirement cohorts are doing and
Section 4 examines issues of fiscal sustainability as the demographic picture changes over
the next 40 years.

If state resources are limited for whatever reason,3 there will be

competition for the marginal dollar of state expenditure. Some spending on older wealthier
New Zealanders may be viewed as “crowding out” other more desirable spending on
younger less affluent people. If so, the question becomes how should the current and
future costs of retirement income policies best be contained, or moderated, without further
harming the retirement of those who are already struggling? The remainder of this report
addresses these issues.
If it is decided that the ‘affordability’ of New Zealand Superannuation (NZS) should be
improved, then there are three major policy responses:
•
•

Raising the qualifying age for NZS,
Lowering the level of the NZS, including changing the indexation of NZS

•

Increasing the degree to which NZS is targeted.

These are examined for their potential in sections 5-7. In section 8, other changes to save
costs, such as increasing residency requirements for NZS; making KiwiSaver compulsory:
growing GDP, raising taxes, prefunding NZS are also briefly discussed. The potential for
cost saving around long term care is discussed in St John & Dale (2019).
The Government’s Terms of Reference for the 2019 Review reflect the Government’s
interest that policies work to enhance outcomes for disadvantaged groups and begin by
requiring:
An assessment of the effectiveness of current retirement policies for financially
vulnerable and low-income groups, and recommendations for any policies that
could improve their retirement outcomes. (Hon Kris Faafoi, 2019)
This explicit requirement echoes the international concern that societies are increasingly
ageing unequally.
“Ageing unequally” refers to inequality that develops throughout the life course and
materialises in old age. It is often the result of specific episodes during people’s
lives that tend to cumulate their detrimental effects on health and income at old
age. Ageing unequally is not a new phenomenon, but while the current generation
of older people is experiencing higher incomes and lower poverty risks than
previous ones in most countries, the younger generations are likely to face again
higher inequality in old age. They are expected to live longer, but have been

3

These may be self-imposed, with restrictive fiscal rules for example, or the perceived difficulties of raising taxes
equitably.

5

 experiencing more unstable labour market conditions, and widening inequalities in
the distribution of earnings and household income. This contributes to widening
inequality in old age, while socio-economic disparities in health status remain large.
(OECD, 2017b, p. 15).
Mindful of this lens, this report examines the main policy tools for their potential to
contribute to fiscal sustainability, and to improve perceptions of intergenerational equity
and retirement outcomes for low income retirees.

2. New Zealand Superannuation – an overview
Figure 1 shows the spending on public pensions as a proportion of GDP for OECD countries
and suggests that NZ spends a modest amount compared to many demographically older
European countries. The New Zealand Treasury projections discussed in section 4 show
that the gross cost of NZS is expected to reach only 7.9% of GDP by 2060, well below the
percentage already spent by a large number of countries today. Comparative figures in
Figure 1 also overstate expenditure on the public pension in New Zealand because the
state pension is paid gross but is fully taxable as income for all recipients.
Other countries design and manage public pensions very differently to New Zealand.
Among developed countries, New Zealand has taken a unique approach that puts universal
non-contributory, flat rate NZS at the centre. In pensions jargon, NZS is a tier one (or
pillar one) scheme. Other countries place less reliance on the basic tier one pension, and
more on tier two, compulsory, contributory state pensions. These social insurance schemes
are usually financed by contributions from both employers and employees.
Compulsory social security contributions are, like taxes in New
Zealand’s case, used to pay for the existing retirees’ pensions on a
Pay-As-You-Go (PAYGO) basis. Because these social insurance
schemes are mandatory, managed by the state, and contributions
resemble taxes, expenditure from them is classified as “public”.

PAYGO schemes do
not have a fund that
accumulates
contributions like a
private savings fund.
Current contributions
are used to make
current payments.

Another hidden form of state spending on retirement incomes comes from tax relief and
subsidies to private retirement income provision. These tax-subsidised savings schemes
may be also mandated by the government, but as they are managed by the private sector,
they are not classified as public schemes. In Australia for example, the second tier
compulsory savings scheme called the Superannuation Guarantee (SG) requires a 9.5%
contribution from the employer, set to rise to 10% in 2021. The SG is ‘tax-effective’
because it has lower (concessional) tax rates for fund contributions and earnings, and taxfree withdrawals at the prescribed age of access.
Many countries encourage private saving for retirement, whether mandatory or voluntary,
to a much greater extent than does New Zealand.

6

 The cost of these fiscal incentives is measured in many OECD countries using the
concept of “tax expenditures”, developed in the 1960s. This attempts to quantify
the value of the preferential tax treatment relative to a benchmark tax treatment.
The idea is that this is the amount the government would have to provide as a
subsidy (a direct expenditure) to achieve the same effect. Data on tax expenditures
for retirement savings are available for 21 OECD countries. Over two-thirds of these
figures are 0.2% of GDP or less. And in only four countries –Australia, Canada,
Germany and the United Kingdom – are reported tax expenditures worth 1% of
GDP or more. (OECD, 2017a, p. 144)
Ireland, Australia and Canada look like they spend less on pensions than New Zealand
(Figure 1), but their tax expenditures on private retirement saving are 0.9%, 1.9%, and
2% of GDP respectively. In contrast, the tax subsidisation of private provision in New
Zealand is limited to a small tax credit for KiwiSaver costing around 0.3% of GDP.4
Figure 1. Public spending on pensions as % of GDP, 2017, or latest available
(OECD, 2017a)

As noted, the taxation of state pensions also makes a difference, with New Zealand levying
more tax than most other countries. For example, the Age Pension in Australia is not
taxed. In 2018/19, Treasury figures show the gross cost of NZS of 4.9% is only 4.1% of
GDP in net terms, making it similar to Australia in Figure 1.5

4

With the home start subsidy, KiwiSaver related subsidies for 2019 are forecast to be $957 million. As different
countries measure tax expenditures differently, the OECD (2017) cautions against simply comparing the
percentages.
5
Treasury estimates based on the New Zealand Superannuation Contribution Rate Model, BEFU 2019 New
Zealand
Treasury.
https://treasury.govt.nz/publications/information-release/new-zealand-superannuationfund-contribution-rate-model-2019.

7

 Given different countries have very different approaches it is hard to draw conclusions
from the data in Figure 1. Typically, in OECD countries, social insurance programmes pay
earnings-related pensions that reward a person’s contributions history, while the basic
state or tier one pension is limited and often means-tested. In the US for example, some
protections for low paid people are built into the formula for social security pensions, but
those who fall through the cracks are assisted with means-tested social assistance,
supplementary or welfare payments.
The coverage under public pensions is also very different. In Australia there is a strict
means test on the age pension, around 20% of retirees do not get any (see section 7). In
Ireland, many women fail to access a full state pension, because it is based on having a
full-contributions record in the paid workforce. Some require means-tested top-ups to
survive, others are dependent on their spouses (St John, 2018a). New Zealand is at the
other extreme: paying the same gross pension (adjusted for marital status and living
arrangements) to everyone who qualifies on residency grounds alone.6
The different approaches taken also result in differing outcomes for those on different
multiples of the average wage when working. Pensions based on contributions inevitably
deliver higher pensions to higher income people. New Zealand pays a flat gross rate but
delivers less net state pension to higher income people because it is taxable under the
(slightly) progressive tax structure. Consequently, NZS delivers a high replacement rate7
for low income people, but, compared to countries with mandatory savings schemes and
contributory social insurance programmes, a much lower replacement rate for high income
retirees (OECD, 2017a, pp. 101-103).

Parameters of New Zealand Superannuation
Table 1 shows the three different 2019 NZS rates and the amounts retained after tax by
individuals who are taxed at the lowest tax rate, and those taxed at the top rate. Tables
2 and 3 provide detailed statistics and the five-year rate of change.
NZS is both price and wage-linked. The net married couple rate after tax at the primary
income rates adjusted annually by changes in the Consumer Price Index (CPI) providing
it is within a band of 66% and 72.5% of the net average wage. Since the floor of 66% was
reached in 2000 (McTaggart, 2005) it has in effect been indexed to wages.

6

With the exception of those who have a state pension from another country (St John and Dale 2016).
There are various ways to calculate replacement rates, all are for fulltime workers at the average wage. Table
4.2 (OECD 2017a) includes mandatory pensions, public and private; Table 4.3 also includes voluntary pension
schemes.
7

8

 Eligibility requires that 10 years residency is met with 5 after the age of 50. When a
superannuitant has lived in another country and has a pension that is classified as an
overseas state pension, deductions will apply (see Dale & St John, 2016).
If admitted to residential care and the pensioner has few assets and income, the pension
helps pay the costs with a small allowance retained for personal spending. The details of
these arrangements are set out in St John & Dale (2019).
The net sharing and living alone rates are set as:
•
•

60% of the net married couple rate for single people sharing accommodation
65% of the net married couple rate for single people who are living alone.

Table 1: New Zealand Superannuation (NZS) and Jobseeker Support rates in NZ$
at 1 April 20198
Category

% Net average
wage

Annual rate
Gross

Annual Net
Primary Tax

Annual Net
33% Tax

NZS Single, living alone

43%

$24,722

$21,380

$16,564

NZS Single, sharing

40%

$22,731

$19,735

$15,223

33%

$18,742

$16,446

$12,557

23%

$12,723

$11,387

19%

$10,602

$9,488

29%

$16,063

$14,232

24%

$13,251

$11,860

NZS Married person or
partner in civil union or de
facto relationship (each)
Jobseeker Single, 25+
years
Jobseeker Married, civil
union or de facto couple
(without children, each)
Supported living
payment single 18+
Supported living
payment (married couple
each)

Table 1 also shows how net welfare benefits as a proportion of the net average wage
compare to NZS. As discussed in section 3, over time, a large gap has opened between
net benefits of various kinds and net NZS because welfare benefits are not wage indexed.
Table 1 shows that the net NZS varies depending on the tax rate paid. New Zealand has
a fairly flat tax structure with the lowest income single living alone paying a 13.5% average
tax rate. High income superannuitants pay 33%. The highest income single living alone
superannuitant receives 77.5% of the net pension paid to the lowest income single living
alone superannuitant.
Thus, tax is mildly progressive and provides a modest degree of claw-back even though
the pension is universal. For 13 years (1985-1998) an additional surcharge to ‘other
income’ applied in effect removing the advantage of getting NZS from high income people

8

see Work and Income website: http://www.workandincome.govt.nz/.

9

 (St John, 2018c). Since the abolition of the surcharge over 20 years ago, the only claw
back is through the progressivity of the tax scale. From 2000 the top tax rate was raised
to 39% but by 1 October 2010 it had been again reduced to 33%.
Table 2. NZS and veterans pensions as at Mar 2019 (source MSD, 2019)9

Tables 2 and 3 show that there are around 15,000 under-age recipients, generally spouses
of superannuitants included in a special Non-Qualifying Partner means-tested rate of NZS.
From 2020, this category will not be available to new retirees (Cabinet Social Wellbeing
Committee, 2019; The Treasury, 2019).10 It is not further discussed in this paper.
In Table 3, the Veteran’s pension has been added to the NZS figures as for purposes of
this report they are equivalent. Table 3 shows the increase in the total numbers over the
five-year period to 2019.

9

See https://www.msd.govt.nz/about-msd-and-our-work/publications-resources/statistics/benefit/archive2019.html.
10
Saving a cumulative $139.9 million by 2023, and $103.5 million per annum thereafter.

10

 Table 3. Five-year increases NZS plus Vet pension (MSD 2019)
Recipient characteristic

5 yr %
change

Mar-14

Mar-19

Male

299,600

362,889

21%

Female

355,397

418,870

18%

Gender

Age Group
Under 60 years

3,676

3,822

4%

60-64 years

10,281

11,123

8%

65-69 years

212,469

235,901

11%

70-74 years

160,938

208502

30%

75-79 years

111,153

146,135

31%

80-84 years

80,802

90,981

13%

85-89 years

50,150

54,681

9%

90 years and over

25,528

30,614

20%

395,253

484,663

23%

Māori

32,360

44,239

37%

Pacific peoples

14,960

19,854

33%

All other ethnicities

86,216

123,405

43%

126,208

109,598

-13%

33,221

41,945

26%

129,287

129,720

0%

3,775

6,579

74%

654,997

781,759

19%

Ethnic Group
NZ European

Unspecified
Receipt of additional support
Accommodation Supplement
Disability Allowance
Temporary Additional Support/Special Benefit
Total NZS+ Vet pensions

Of interest is the rapidly changing ethnic mix. NZ European have increased marginally to
be 62% of all recipients by 2019, likely due to improved longevity in this group. However,
NZ European numbers have grown more slowly (23%) than for Maori and Pasifika and all
other ethnicities (overall 40.4%). The non-specified ethnicity group has fallen by 13%.
Table 4 shows the numbers receiving different rates of NZS. Approximately 62% are
married, 13% are single sharing and 25% live alone. Gross superannuation for 2018/19
was $14.562 billion, of which approximately $9 billion was paid to married persons, $2
billion to single sharing and $3.5 billion to superannuitants living alone.
Table 4. Characteristics of NZS and Vet pensions as at March 2019
(Ministry of Social Development, 2019) Rate type
Individuals in a married, civil union, or de facto couple
where each qualifies for NZS/VP
Single sharing accommodation
Single living alone
Single
Non-qualified partner
Total

11

NZS + Vet pension
Female
Male
208,608 261,564

total
470,172

Share %
60.14

68,336
130,434
13
11,479
418,870

103,614
194,712
30
13,231
781,759

13.25
24.91
0.00
1.69
100

35,278
64,278
17
1,752
362,889

 3. Pre-retirement and post retirement cohorts
Labour force participation of older people
Older people of working age in New Zealand have a high labour force participation rate
(LFPR) as shown in Table 5, and to date a low unemployment rate. Even by the age 6064 years, nearly 72% are still in the labour force (meaning employed for 1 or more hours
per week). As at March 2019, the LFPR of people aged 65+ is 23.7% with 170,100 still in
paid employment.
Table 5. Labour market statistics for older workers: Quarterly Employment
Survey March 2019
Not in
labour
force

Labour force
Employed Unemployed

Total

WorkingLabour force
age
Employment Unemployment
participation
population
rate
rate
rate
(2 )(3 )

(000)
267.0

9.0

275.9

254.3

6.7

261.1

197.9

5.0

202.9

170.1

2.1

172.2

(%)
50–54 years
36.0
311.9
55–59 years
53.3
314.3
60–64 years
72.6
275.6
65 years and over
553.7
725.9

88.5

85.6

3.3

83.1

80.9

2.6

73.6

71.8

2.5

23.7

23.4

1.2

Of those still in the labour market at 65+, most are under 75 years of age. Table 6,
based on census 2013 data11 shows that one third of women and nearly one half of men
aged 65-69 are in the labour force. By ages 70-74 this drops to 15.5% of women and
27% of men, with relatively very few over age 75 still in the labour force.
Table 6. Labour force participation (LFP) aged 65+ by age group using 2013
census data (Littlewood, 2014, Table 3)
LFP
Age
group

Males
LFPR %

Males
number

Females
LFPR %

Females
number

65-69

48.10

45,972

33.20

33,330

70-74

27.00

19,467

15.50

12,090

75-79

14.70

7,293

6.70

3,846

80-84

8.40

3,000

3.80

1,707

85-89

5.80

1,512

2.70

1,281

90+

n.a.

n.a.

n.a.

n.a.

Participation in paid work by older persons in New Zealand is high. The LFPR for those
aged over 65-69 is second only to Iceland in the OECD (OECD, 2017a, p. 125). Compared

11

2018 census data is not yet available.

12

 to Australia, our nearest neighbour, Table 7 shows that the LFPR of New Zealanders aged
55 and over, have been significantly higher.
Table 7. Older workers Labour Force Participation Rate: (ILO statistics base)12

Age
band
total

male

female

LFPR

LFPR

Percentage points
difference

Australia

NZ

NZ v Australia

2018

2030

2018

2030

2018

2030

55-64

66.7

71.4

80.8

84.5

14.1

13.1

65+

13.2

15.7

24.4

27.4

11.2

11.7

55-64

73

76.4

87.1

89.1

14.1

12.7

65+

17.2

19.8

31.5

34.7

14.3

14.9

55-64

60.6

66.6

75

80.3

14.4

13.7

65+

9.7

12

18.2

21

8.5

9

The LFPR of all those aged 65+ (24.4%) is nearly twice that of Australia (13.2%). For
females the LFPR in the pre-retirement age 55-64 is 14.4 percentage points higher in New
Zealand. By age 65+ New Zealand women are nearly twice as likely to be still in the
labour force. In part this reflects lower disincentives to be in paid work in New Zealand
once pension-qualifying age is reached.13
The International Labour Organisation (ILO) projections in Table 7 show that in both
countries the LFPR is expected to increase at all ages 55+ by 2030. Such projections are
to some extent speculative: the ‘future of work’ is likely to be very different than work
today and gains in LFPRs are likely to slow as the proportion of younger retirees falls
relative to all aged 65+.
Nevertheless, older citizens are clearly making significant contributions to the paid
workforce. The latest projections are summarised by the Ministry of Social Development
(2019):
The proportion of the population aged 65 and over in the labour force is expected
to rise from the current rate of 23.7 percent14 to reach up to 26 percent by 2036.15
The labour force aged 65 years and over increased from 23,000 in March 1991 to
57,900 in March 2006 and to 146,000 in 2016. It was 172,200 in March 201916 and
is projected to reach 311,000 by 2038. This means that around 8 percent of our
total labour force will be aged 65 and over by 2023 (up from 7 percent in 2019),
and 10 percent by 2038.17

12
13
14
15
16
17

See International Labour Organisation (ILO) https://ilostat.ilo.org/.
The LFPR for the 65+ group in Australia will increase as the age for the age pension increases.
Household Labour Force Survey: March 2019 quarter. Statistics New Zealand, Wellington, May 2019.
National Labour Force Projections: 2017 (base) – 2068, Statistics New Zealand, Wellington, December 2017.
Household Labour Force Survey, Statistics New Zealand, Wellington, extracted from Infoshare 31 July 2019.
National Labour Force Projections: 2017 (base) – 2068, Statistics New Zealand, Wellington, December 2017.

13

 Even so, many do not work full-time. Of the 100,200 aged 65+ earning wages and salaries
in 2018, 46,200 are female and 54,000 are male. Only 50% percent of women work fulltime, compared to 62% of men.

Approximately, 60,000 of those over 65 have self-

employment income.18
A Commission for Financial Capability survey (2016, p. 5) with 2,989 respondents revealed
the main reason 54% of the 65+ group in paid work remain in the workforce is financial
necessity, while a minority (36%) work mainly for satisfaction and value. It might be
surmised that this latter group are largely white-collar professional people, although some
lower paid persons may also derive social contact and/or mental health benefits from work.
The impact of retirement on physical health is an important focus of ageing. Research
findings using longitudinal data from the New Zealand Health, Work and Retirement Study
(Szabó, Allen, Stephens, & Alpass, 2019), examining physical functioning 8 years pre- and
post-retirement indicate that retirement can be beneficial for those with poor health and
limited resources. Unsurprisingly, for the wealthy and healthy, retirement does not
necessarily advantage health. Importantly, the researchers confirmed that NZS may partly
address inequalities experienced by older persons in poor health and socio-economic
circumstances prior to retirement, meaning their circumstances improved post-retirement.
These findings suggest there are advantages in maintaining the current policy regime
around provision of NZS.
There is also a downside to the increase in numbers of older workers participating in the
workforce. While the oldest workers (75+) had the lowest number of accident
compensation claims in 2017 consistent with the demographic structure of the working
population, they had the highest claim rate of all age groups at 190 claims per 1,000 FTEs.
This group also had the highest incidence rate for more serious claims, at 66 claims per
1,000 FTEs, compared with 14 claims per 1,000 FTEs for the total population.19

Low income, pre-retirement group
In New Zealand, older working age people (45-64 years) who are living on their own have
the second highest rate of income poverty after sole parents: doubled from 10% in 1988
to 23% in the early 2000s, and currently at 29% (McKenzie, 2019). Evidence suggests
while the depth of poverty is largest among beneficiaries and their children, (Perry, 2019
forthcoming, p. 64),20 there is significant poverty and hardship among the older lowincome working age population.

Statistics New Zealand info data, accessed July 2019.
See https://www.stats.govt.nz/information-releases/injury-statistics-work-related-claims-2017.
20
Sole-parent households with dependent children have the highest low-income rates of all household types of
around 45%, compared with a population rate of 15%. Using the 60% AHC REL measure, the rates are around
60% compared with 21% respectively. (Perry, 2019 forthcoming, p. 142)
18
19

14

 Of those aged 50-64 years, while 714,400 are employed in paid work (NZ Stats 2019),
around 120,600 receive a welfare benefit with more than half of these on the benefit for
over 5 years, and with two thirds of this group on a benefit for reasons of health disability
(McKenzie, 2019). A further number had an ACC payment including 14,440 on earnings
related compensation for an average of 117 days21, and 14,130 people aged 50 to 64, not
receiving a main social security benefit (ie low income workers) were receiving one or
more weekly supplementary benefits (McKenzie, 2019).
These figures provide a snap-shot at a point in time, but the probability of a person needing
benefit, ACC or supplementary assistance between the ages of 50 and 65 is much higher.
In later working age, periods of unemployment, redundancy, ill health or caregiving duties
may impact severely on the ability to prepare financially for retirement.
Welfare benefits are widely perceived as inadequate a outlined in the Welfare Expert
Advisory Group report Whakamana Tangata, (2019). The net single 25+ Jobseeker rate is
barely over 50% of NZS for a single recipient taxed at the primary rate living alone (Table
1). Even if NZS is taxed at the top tax rate, perhaps because the superannuitant works
full-time, the net NZS is still 45% more than the net Jobseeker Support benefit rate paid
to an unemployed adult. The inadequacy of benefits and other assistance22 means that
many people on benefits accumulate private debt, and/or have debts to public agencies.
For example, as at June 2018, the Ministry of Social Development (MSD) was owed $557.8
million as recoverable hardship assistance, and a further $768.7 million by clients who had
received money from MSD to which they were not entitled (Welfare Expert Advisory Group,
2019, p. 85). Of such debts, at the end of December 2018, over 56,000 current
beneficiaries aged 50 to 64 years owed the MSD a total of $182 million (McKenzie, 2019).
Using the DPB, sole parent benefit (now called sole parent support) to illustrate the general
case for benefits, Figure 2 shows the gap between welfare benefits and NZS has grown
despite the occasional one-off boosts to welfare. NZS has been linked to average wages,
which have risen markedly, while benefits are linked to CPI prices only.

While the

indexation for benefits will change to wage indexation from April 2020 and prevent the
gap from widening further, it will not close that gap.

21

In 2016, 216,000 people aged 50 and over had one or more ACC claims for a fall-related injury accepted.
Data are not presented on the type of injury resulting from the fall, which may range from a simple bruise
through to a head injury. See https://www.hqsc.govt.nz/our-programmes/health-qualityevaluation/projects/atlas-of-healthcare-variation/falls/. Earnings related compensation data is from 2019 OIA
22
As detailed in the WEAG report, pp. 96 – 98.

15

 Figure 2. Comparisons between average wages, NZ Super and benefit levels23

Those on benefits also face a stringent income test whose parameters have not been
changed since the 1980s. Earned mounts over $80-100 per week are treated punitively.
In contrast there is no disincentive for those aged 65+ to earn over and above NZS.

Declining rates of home ownership
As well as providing security of tenure, home ownership avoids escalation in rent costs.
Table 8 shows that all ages, except those over 80, experienced declining rates of home
ownership between 2001 and 2013 on census figures (2018 data not available).
Table 8. Home ownership rates (Census data, Mckenzie, 2019)
Age Group

2001

2006

2013

25-29

26.4

22.9

18.4

30-34

47.7

43.6

36.0

35-39

61.0

56.6

49.6

40-44

69.1

64.8

57.9

45-49

74.3

70.5

63.8

50-54

77.9

74.8

68.4

55-59

79.3

78.0

74.9

60-64

79.3

79.1

74.9

65-69

80.2

79.3

77.3

70-74

80.6

79.5

77.5

75-79

78.7

78.4

75.9

80-84

72.6

74.5

73.1

85+

55.4

59.0

60.4

23

Adapted from Perry (2014) Ministry of Social Development, page 82, Figure C.8A.

16

 While there are measurement issues that reduce the usefulness of data on home
ownership,24 the decline in home ownership in the pre-retirement age groups 50-64
appears significant and shows the problem is not just younger people delaying home
purchases. This lower owner occupancy can be expected to feed into lower rates of home
ownership for those over 65 in the years to come. When available, Census 2018 data will
likely show a further decline in home ownership in those approaching retirement.

Changing nature of work
In addition to housing problems, the rapidly changing nature of work may make
employment at older ages more difficult with more calls on welfare support. The Welfare
Expert Advisory Group report25 (2019, p. 130) warns of the coming disruptions:
The welfare system needs to respond to employment changes brought on by the
growing use of robotics in manufacturing and digital technologies (the 'fourth
Industrial Revolution'), the gig-economy, and New Zealand’s transition to a lowcarbon economy. The magnitude of these changes cannot be predicted with
confidence, but major job losses and considerable structural change in the labour
market are virtually certain.
Research by Hyslop and Townsend (2017, p. 26) confirms that people displaced from work
generally re-attach to the labour market, but they have poorer long-term outcomes than
those who have not lost their jobs. This highlights the importance of the systems in place
to re-attach people to the labour market and ensure that their skills remain relevant to
employers: “job-displacement has substantial long-term effects on workers labour market
outcomes and income support”.
In a changing labour market of more casualised work, higher redundancy, youth
unemployment, the welfare system and retraining and support provisions appear seriously
lacking. In this environment many older workers are likely to struggle to obtain sufficient
income from work and have more periods between paid employment which will affect their
ability to maintain assets let alone save for their retirement. They may be less likely to get
retrained than younger people. It is very likely that hardship amongst older working age
people will rise in the future:
Support for displaced workers is particularly weak. Compared with OECD best
practice, New Zealand has an inadequate system of dealing with job loss,
redundancies and labour market shocks (OECD, 2017). Redundancy pay is not
required by law, the stand-down provisions between work and benefit entitlement
see many workers and families plunged into poverty. (Welfare Expert Advisory
Group, 2019, p. 46)

24

See, Chamberlain and Littlewood (2019).
Also see http://weag.govt.nz/assets/documents/WEAG-report/backgrounddocuments/29c1cc3696/Employment-and-Labour-market-010419.pdf.
25

17

 How well are those over 65+doing?
As a basic income, provided on residency grounds, not on contributions or work periods,
along with high rates of mortgage-free home ownership, NZS appears to have been
outstandingly successful to date in preventing poverty among most of those over
65. Compared to the rest of the population, their fixed line income-based after housing
cost poverty rates have been very low (Figure 3a). Similarly, hardship rates based on nonincome measures are also low, 3% compared with 13% for children and 8% overall (Perry,
2019 forthcoming).
Figure 3a. Poverty rates after housing costs for older people compared to general
population. Source: Perry (2019, forthcoming)

International comparisons of income poverty rates can be misleading. On the 60% before
housing costs (BHC) measure NZ appears to have high rates of poverty (around 33%),
yet on the BHC 50% measure the NZ rates are very low (4-5%).26 This difference arises
because of the high number of older New Zealanders whose income is NZS and little or no
more, and the value of the NZS being not far above the 50% BHC threshold in the last 15
years. Material hardship measures provide more realistic international comparisons. Using
the EU’s official hardship measure, the rate for 65+ New Zealanders was 3% in 2017,
which ranks NZ in the top five EU nations along with Norway, Sweden and Denmark (Perry,
2018, p. 48).
Nevertheless, in the last five years relative living standards have slipped. While NZS
(married rate) has been kept at 66% of net wages, with the rise of more two-earner
families, NZS has fallen relative to median household incomes (Perry, 2019 forthcoming,
figure 1.3). Figure 3b shows the steady increase in poverty as measured by the 50% AHC

26

These figures are based on the EU’s approach, using the modified OECD equivalence scale. See Perry (2018,
p. 173 for more detail).

18

 contemporary poverty line. This corroborates the suspicion that many more retirees are
affected by rising rents as home ownership falls.
Figure 3b Poverty rates using a 50% contemporary line measure. Source: Perry
(2019 forthcoming)

The benign poverty picture that was shown in Figure 3a may change as less well-prepared
cohorts enter retirement. As shown in Table 8, the rates of home ownership among late
working age people is dropping (see). Fewer people will reach 65 with a mortgage-free
home and many face high weekly rental costs as the New Zealand property market soars.
Figure 4 charts significant changes in tenure for those over 65, showing a drop in those
with no mortgage and a rise in numbers with a mortgage.
Figure 4. Tenure for individuals aged 65+. Source: Perry, (2019 forthcoming)

19

 Towards the end of the second decade of the 21st century an increasing number of
superannuitants require additional support in the form of the Accommodation Supplement
(AS).27 The numbers of those over 65 accessing the AS has increased from under 4% in
the early 2000s to 5.4% in 2019 or 42,000 people (see Table 3). An additional 14,241
people aged 65 or over in public housing receive the Income-Related Rent Subsidy that
means they pay no more than 25% of their income in rent (McKenzie, 2019).
Other supports available to NZS recipients are: Disability Allowance , Temporary Additional
Support (see Table 2 and 3 above) and hardship grants which are primarily for food.
Evidence from the NGO sector also points to a growing poverty problem in the older age
groups despite the extra MSD support. Budgeting services and foodbanks report increased
demand for food parcels for this age group. Darryl Evans, CEO of a major South Auckland
budgeting centre reports that from a client base of 1,300 over age 65, 29% need
assistance with food costs.28 Others may be accessing low- or no-interest loans from
microfinance providers,29 or borrowing from less benign providers of credit.
It is fair to say that the people who actually ask for food parcels are the tip of the ‘need’
iceberg. The Auckland City Mission data30 (Figures 5) shows a worrying trend of increased
monthly demand for food parcels by the group aged over 60.31
Figure 5. Number of food parcels requests persons aged 60+ (Auckland City
Mission)

NUMBER OF REQUESTS

600
500
400
300
200
100

Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May

0

2015

2016

2017

2018

2019

DATE
Figure 6 analyses this food need indicator by age bracket. While the absolute demand is
unadjusted for the numbers in the respective age categories, 92% of the food parcels

27

The Accommodation Supplement is an asset and income-tested top-up assistance payment for rent or home
ownership costs such as mortgages.
28
Personal communication.
29
In 2018, of 130 applicants to Nga Tangata Microfinance for debt relief loans, 10 were aged 65+.
30
Personal communication Helen Robinson, GM Auckland City Mission.
31
Data covers hubs in Papakura, Mangere & Manurewa as well as central Auckland (ACM June 2019).

20

 requested by those over 60 are for those between 60 and 75 years of age. The picture is
suggestive of a growing hardship picture among people in that age group.32
While Christmas is the peak time for food insecurity, Figure 6 shows that all months for
those aged 60-75 have seen a significant increase. The figures for 2019 are continuing
this trend. Figure 7 may indicate that women are more likely to be in food poverty than
men aged over 60. It may also indicate that women are more likely to seek outside help.
Figure 6. Number of food parcel requests from persons 60+ by age. (Auckland
City Mission)

NUMBER OF REQUESTS

250
200
60-65
65-70

150

70-75
100

75-80
80-85

50

85-90
90-95

Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May

0

2015

2016

2017

2018

95-100

2019

DATE
Figure 7. Number of food parcel requests from persons 60+ by gender
(Auckland City Mission)

NUMBER OF REQUESTS

350
300
250
200
Female

150

Male

100

Unknown

50
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May

0

2015

2016

2017

2018

2019

DATE

32

In the 2018 ANZ Financial Wellbeing Survey of around 828,000 people in New Zealand, only 23% had no real
financial worries (Prendergast, Blackmore, Kempson, Russell, & Kutin, 2018, p. 5).

21

 Further research is needed to see if the increased demand for food from those aged over
60 are corroborated by City Missions in other parts of New Zealand. While the Wellington
City mission has not observed higher demand for food parcels for those over 60, they
report a steady increase in other support required for this age group (60+) over the last
two financial years 2017- 2019. They note there are considerably more males than females
over 65 years that are provided with both food and other support.
While the decline of homeownership, and the rise in rental costs is a major explanation for
food insecurity, some new retirees may have been independently affected by the
experience of the Global Financial Crisis (GFC), the casualisation of work, divorce, ill
health, or some combination of these. Causation however is likely to be both ways. For
example, poorer health can lead to poorer housing as well as poorer housing leading to
poorer health. After three consecutive, annual, nationally representative New Zealand
Health Surveys that compared owner‐occupiers, private renters and public renters,
Pledger, McDonald, Dunn, Cumming, & Saville-Smith (2019) concluded that rental tenure
is associated with poorer health.33
Rising numbers of older renters, who are often living alone, on low incomes and in
poorer health than owner‐occupiers, will have important implications for future
health and housing policy. Much of this policy is currently premised on high rates
of home ownership and changing patterns of tenure will need to be considered and
adapted to. There are also implications for services in terms of the future demand
for care, in helping deliver policies that support people to remain healthy and in
their homes for longer, and in ensuring that the diverse health, care and support
needs of older people living in their homes are met. (Pledger, et al., 2019, p. 188)

Summary
Already there are signs of financial stress among some of the 65+ group. As measures to
improve fiscal sustainability of retirement policies are discussed it is also important to
understand the particular problems that will be faced by the cohort aged 50-65 as it
approaches retirement.34 The poorer life experience of many in cohorts entering retirement
in the next 15 years may see more new retirees at 65 less well prepared financially, less
healthy, and with less security of housing tenure.
As well, the changing nature of work, digitalising and automation will have a profound
impact on the ongoing employment of older workers who may find themselves without
skills needed in the new 21st Century economy and face frequent redundancy. Older people
may be in competition for the limited resources available for retraining and education with
similarly affected young workers.

33
34

In their report, pooled data were analysed for 15,626 older adults (aged 55+) from these three studies.
It also reflects the requirement of Retirement Income Policy Review Terms of Reference No. 1.

22

 While New Zealand has one of the highest LFPR for cohorts aged over 65 in the OECD, it
must be acknowledged that this is primarily part-time work. The majority of those aged
65+ are not working at all (77.4%) and of those working it is predominantly from financial
necessity although they have a basic income from NZS. The LFPR for women has increased
faster than for men older age categories and increasingly women may seek to work past
the age of 65 to make up for time spent out of the workforce in a caregiving role.
Many retirees, even if in part time paid work, make valuable contributions in unpaid work,
especially providing the caregiving required for the young when parents work, and for
older parents who have become frail. Volunteer work of all kinds, meals on wheels,
supporting the Arts, mentoring the next cohorts in business and the professions, search
and rescue, ambassadors for tourism, environmental projects, budgeting services, to
name a few are the lifeblood of a community and can be facilitated by the provision of a
basic income as the NZS provides (see section 7).
This section suggests that as younger cohorts enter the 65+ years traditionally associated
with retirement, a rise in elder poverty in New Zealand can be expected, especially when
measured in after housing costs terms. Many will feel forced to continue paid work, but
this has an opportunity cost for them and for society as further discussed in section 7.

4. Fiscal sustainability and intergenerational equity: current
arrangements
Demography
Figure 8 is a stylistic picture of New Zealand’s demographic change. There were around 5
people of working age to each one aged over 65 in 2010. By 2060 this falls to just over 2
of working age to each one over 65.35 Another way to put this: the 65+ dependency ratio
(the number of people aged 65+ per 100 people aged 15–64 years) was about 20 per 100
in 2010 and will rise to 50 per 100 in 2060.
Figure 8. Structural Ageing in New Zealand 2010 to 2060
Projected ratios of different age groups
15-64 years
65+ years
2010
2020
2030
2060

35

The actual figures are 2.8 at working age to each person aged 65+ by 2032, 2.5 in 2050 and 2.1 in 2068.
National Population Projections: 2014 (base)–2068.
http://archive.stats.govt.nz/browse_for_stats/population/estimates_and_projections/NationalPopulationProjecti
ons_HOTP2014.aspx.

23

 Fiscal pressures
The 2019 Budget records NZS gross expenses of $13.7 billion in the most recent completed
fiscal year of 2017/18. It forecasts these expenses to increase from this level by $1.8
billion to $15.5 billion in 2019/20 and by $4.8 billion to $18.5 billion by 2022/23, the end
of the forecast period (Figure 9). In this final forecast year NZS represents around 53% of
core Crown social assistance spending and 17% of core Crown expenditure.
Figure 9. Growth of NZS recipients and expenses (The Treasury, 2019)

The growth in expenditure of 34.8% shown in Figure 9 reflects both an expected increase
in recipient numbers from almost 741,300 in 2017/18 to approximately 872,900 by 2023
(17.8% increase) and continued indexation of NZS to wages (The Treasury, 2019, p. 29).
Based on Budget updates of the New Zealand Superannuation Fund (NZSF) Contribution
Rate Model on the Treasury website, Figure 10 shows the projected growth of NZS
(Budgets 2000, 2009, 2019) under pressures from an ageing population (Bell, 2019b).
Figure
10. New
Zealand
(NZS)expenditure
expenditure
(gross and as a
New
ZealandSuperannuation
Superannuation (NZS)
(gross)
Percentage (%)
a percentage
of Nominal
percentage ofas
Nominal
GDP (Bell,
2019,GDP
Fig. in
2)history and projections
12

Projection

History

10

8

6

4

2

0
1997

2002

2007

2012

Budget 2000 projection

2017

2022

2027

Budget 2009 projection

24

2032

2037

2042

Current projection

2047

2052

2057

Year ended June

 Over time the projections in Figure 10 appear to show that there is less of a problem than
had been thought.

Growth of GDP has been higher than expected, and one factor

contributing to this has been the greater LFPR of older people. The most recent projection
shows a steady upward trend but lower growth path in the proportion of the GDP
represented by NZS expenditure. These are gross expenditure figures so the net
projections are even less stark. However, an ageing population and improving technology
will mean health expenditure will also grow faster than GDP. Health expenditure is more
difficult to forecast than NZS but, based on different assumptions, Treasury provides a
variety of projections ranging from around 8.5% to nearly 12% of GDP by 2060.
Table 9 from the Treasury’s 2016 Statement on the Long-Term Fiscal Position projects
expenditure under current policy settings using historical spending patterns: the
combination of NZS and healthcare expenditure rises from 11% in 2015, to a projected
17.6%, while welfare and education remain relatively static.
Table 9. Projections for ‘historical spending patterns’ scenario as % of GDP
(The Treasury, 2016, Table 6.1)
2015

2030

2045

2060

Healthcare
New Zealand Superannuation
(NZS)

6.20

6.80

8.30

9.70

4.80

6.30

7.20

7.90

Education

5.30

5.40

5.50

5.70

Law and order

1.50

1.40

1.40

1.40

Welfare (excluding NZS)

4.20

4.50

4.70

4.70

Other expenses

6.30

6.70

6.70

6.70

Debt-financing costs

1.60

2.20

5.30

11.00

Expenses

30.00

33.30

39.10

47.10

Tax revenue

27.60

28.60

28.60

28.60

2.30

2.40

2.40

2.50

Revenue

29.90

31.00

31.00

31.10

Operating balance

-0.10

-2.30

-8.10

-16.00

Primary expenses

28.40

31.10

33.80

36.10

Primary balance

0.50

-1.20

-4.00

-6.30

Capital expenditure

0.70

0.90

1.00

1.00

Net debt

25.10

32.50

94.00

205.80

Net debt incl NZSF

12.90

11.50

68.90

174.10

Net worth

13.80

16.10

-41.30

-146.30

Other revenue

Note: All variables are on a core Crown basis; New Zealand Superannuation expenses
are on a gross basis; bracketed numbers represent negative values; primary expenses
are expenses excluding debt-financing costs and the primary balance is the difference
between revenue (excluding interest revenue and dividends) and primary expenses;
these projections represent a "what if" scenario.

25

 If current policies are unchanged, including keeping tax revenue at 28.6% of GDP the
projections show (Table 9) that net debt grows exponentially as a % of GDP. These
projections will be updated when the next Long-Term Fiscal Statement is published in
2020. While individual figures in the table will undoubtedly change, the general path of
the key fiscal indicator, net debt to GDP, is very likely to follow a similar upward trajectory
to unsustainable levels.
It is not suggested that the outcomes by 2060 shown in Table 9 will occur. One of the
strong assumptions is that taxes as a percentage of GDP remains fixed. Ceteris paribus,
over time as relatively more is spent on ageing cohorts, there will be relatively less for the
working age cohorts. Many of working age and their families have much higher rates of
deprivation and income poverty so that there will be inevitable pressure for changes long
before 2060. Either changes can be planned for and the fiscal pressures managed or there
will be ad hoc, knee-jerk, ill-thought-out changes in response to the perceived crisis.

Role of prefunding
One of the problems of expressing the demographic pressures in simplistic income and
expenditure terms is that no account is taken of what is happening to the overall Crown
balance sheet. The ‘what if’ scenario in Table 8 includes what happens to Crown assets
and liabilities and net worth as the projected operating surplus becomes a deficit.
Under the PAYGO system, NZS is funded by current
taxpayers, as are subsidies for old age care. At the same
time, current taxpayers are contributing to the NZSF. This
fund has been accumulating assets since 2003 although
regular contributions were suspended in 2009 and not

NZSF: in essence, in the early
years, taxpayers provide it
with funds that are invested to
grow so that, in future years,
money can be taken from the
NZSF to help future taxpayers
cover the cost of higher public
pension expenses.

restarted until 2018. In 2019, the size of the fund is nearly
$42 billion and the cumulative contributions of $9.5 billion over the years 2018/19 to
2022/23 is expected to increase the size of the NZSF to nearly $64 billion by the end of
this period (Table 10).
Table 10. The NZ Superannuation Fund (The Treasury, 2019)
Year ended 30 June
Contributions
Fund size

2019 $m
1,000
41,811

2020 $m
1,460
46,093

2021 $m
2,120
51,345

2022 $m
2,420
57,276

2023 $m
2,553
63,769

It is difficult to know the counterfactual: if NZSF did not exist, the contributions to it may
have gone towards lowering taxes or may have been spent on other things, some of which
might or might not have increased productive spending to benefit the real growth of the
economy. The Fund has opportunity costs. From a real resource point of view, if more is
allocated to the ageing population, then there will be a smaller slice of current output of

26

 goods and services for the working age population, regardless of how or where the funding
comes from.
Moreover, NZSF does not prevent future tax increases: the future cost of NZS will continue
to rise as may other spending categories. Accounting for government spending is not
affected by where the spending is financed from, so that the Fund does not help the
government keep within its fiscal caps (government expenditure on NZS is not changed
when there are capital contributions made from the NZSF), nor does it avert pressure from
competing demands.
In the meantime, the recent resumption of allocations from the budget surplus to the NZSF
ask the working age population to not only fund the existing superannuitants, but also pay
for some of their own, perhaps less generous, future age pensions. These contributions
may preclude other useful government spending or redistribution.
While some argue that funding improves savings and investment rates, evidence that
funded or Save As You Go (SAYGO) schemes improve the size of the rates GDP pie itself
is scarce (Barr, 2001; Barr & Diamond, 2009). The NZSF is best thought of as performing
a tax smoothing function.
Another common misunderstanding is that the NZSF will build up and then be drawn down
and disappear, but the Fund is not set up that way. The Treasury explains the way the
Fund works in detail and the complexity of the calculations involved (Bell, 2019b). While
each year, a calculation is made such that over a forty-year horizon, the tax paid to meet
the total costs of net NZS (the contribution rate) could be kept constant, assuming at the
end of the 40 years the Fund is exhausted, each year, a new 40-year time horizon is used
to set a new contribution rate (tax smoothed rate). Figure 11 illustrates the latest
projection of the tax smoothing role.
Figure 11. Tax smoothing role of the NZ Super Fund (Bell, 2019b, Fig. 3)
% GDP

NZSF Contribution rate versus Net NZS expenditure

9
8
7
6
5
4
3
2
1

0
23/24

28/29

33/34

38/39

43/44

48/49

Aggregate Net NZS expense

53/54

58/59

63/64

68/69

NZSF Contribution rate

27

73/74

78/79

83/84

88/89

93/94

98/99

Year ended 30 June

 Each year the difference between the cost of NZS and the tax smoothed contributions rate
is the capital contribution made to the Fund, or the capital withdrawal taken from the
Fund. Under the current projections, as shown in Figure 11, the capital contributions
virtually stop in the mid-2030s and significant contributions from the Fund do not start
until the mid-2050s.
By the end of the century the annual contribution from the Fund from that time is expected
to be about 1% of GDP. Looking to the end of the current 40-year horizon, by 2059/60
the projected annual cost of NZS is $105 billion with the Fund meeting $7 billion or just
6.7% of the net costs of NZS.36 Figure 12 shows that the Fund does not reduce much on
budget 2019 projections, even by the end of the century, as s when it will be worth over
30% of GDP.
Figure 12. NZSF closing balance, as a percentage of GDP, as projected at Budget
2012, Budget 2016 and Budget 2019 (Bell, 2019b, Fig. 13)
% of Nominal GDP
40

35

30

25

20

15

10

5

0
1999/00

2009/10

2019/20

Budget 2012

2029/30

2039/40

Budget 2016

2049/50

2059/60

Budget 2019

2069/70

2079/80

2089/90

2099/00

Year ended 30 June

Summary
While the NZSF may be helpful eventually, there are two considerations. First, there is an
opportunity cost of the funds contributed (expected to be $9.5 billion over the forecast
period 2019-2023): there are alternative uses for this money. The second consideration
is that the public may be misled into thinking that the fund makes their pensions secure.
In fact, it provides no guarantee of any of the parameters such as level of NZS or its
indexation or continued universality, and it only very partially funds NZS. It does not alter
any of the real resource costs of an ageing population. In the meantime, the decision to
exclude the NZSF assets from the net debt fiscal rule that the government has adopted

36

See https://treasury.govt.nz/publications/information-release/new-zealand-superannuation-fundcontribution-rate-model-2019.

28

 may unduly constrain the type of real investment today that might better prepare society
for the inevitable pressures of an ageing population.
The discussion in this section shows that the outcome under current settings is not
sustainable. An increase in taxation is one way to proceed although to date an increase in
personal income tax rates, GST and capital gains tax have been ruled out. Many other
sectors such as welfare, justice, education and health are in need of serious injections of
money and are unlikely to be sources of cost saving. This leaves reducing future NZS
payments through the use of one or more of the three main levers: the age of eligibility,
the level of payments, and some form of means-testing.
While raising the eligibility age is often discussed as if it were the only option, a carefully
considered mix of the three main levers: raising the age, lowering the rate, and incometesting, might most effectively maintain the best features of NZS. The first two levers are
briefly discussed in sections 5 and 6, followed by a more detailed proposal for use of the
third lever in section 7. This third lever has been seldom discussed seriously in New
Zealand since the late 1990s when the surcharge was abolished.

5. Age of eligibility issues
International context
It is not self-evident that New Zealand is much out
of

step

with

other countries.

While

several

countries are proposing raising the age of eligibility
for state pensions, the average expected age on
current policies across the OECD will be just 66 by
2060 (OECD, 2017a).
Figure 13 shows that under current legislation a
few countries are aiming to have their retirement
age considerably higher by 2060. But it is a two-

What are other countries doing?
In the Netherlands, Denmark, Belgium,
France, Germany and Spain the age
pension qualifying age is already 67. The
UK age for the British State Pension will
rise from 2018 to reach age 68 by 2046,
while in Ireland the age rises to 68 by
2028. The age will be 67 in Australia by
2023, and in the US by 2027.
A few outliers, eg France, Greece,
Luxembourg, Slovenia and Turkey,
expect the age to be below 65 in 2060.
Overall, the average expected age on
current policies across the OECD will be
just 66 by 2060 (OECD, 2017a).

edged sword. Under contributory schemes, pensions become more valuable if accessed
later. While this may be good for the individual, it may not save costs in the long run.
Moreover, the problems of poverty are re-emerging among younger less affluent retirees
who are having to wait longer, often without the benefit of extra years of paid work.
Of those countries whose age is expected to increase, Denmark stands out: the age of 74
is by 2060 reflects the indexing of the qualifying age to expected improvements in
longevity.
Finland, Italy, the Netherlands, Portugal and the Slovak Republic are also linking the age
of eligibility to life expectancy (OECD 2017a). Nevertheless, there are controversies over
how higher ages are working out for manual workers.

29

 Japan has one of the oldest demographic profiles and while they are talking about raising
the age,37 they have not yet done so. It is possible for people to choose to postpone
retirement in Japan as it is in many countries, but this does not save any money:
… the incentives to continue working after the retirement age are large and go
beyond the increases that would be justified to compensate for the short retirement
period (OECD, 2017a, p. 12).
Figure 13. At what age will young people retire? (OECD, 2017)38

Three countries are actually reducing their age in response to concerns about austerity
policies and elder poverty. In Canada for example, in 2012 the Conservative government

37

See https://www.reuters.com/article/us-japan-economy-retirement/retiring-late-as-pensions-underwhelmmore-japanese-opt-to-prolong-employment-idUSKCN1RM0G.
38
Note that the age of entitlement to a state pension and the actual age of retirement may differ. In New Zealand
there is no age of retirement.

30

 decided to raise the age for the basic pension (Old Age Security) from 65 to 67 by 2023.
Under Justin Trudeau the Liberals won the 2015 Federal election with reversing the age
rise as a key policy. The argument was that it was better to encourage Canadians to stay
in the workforce rather than force them to stay and increase poverty of those who cannot:
Reversing the previous government’s arbitrary decision to [raise] the OAS age of
eligibility was a commitment we made with the middle class and the most
vulnerable Canadians in mind, and it was absolutely the right thing to do. (Duclos,
Canada’s Minister of Families, Children and Social Development).39

Case for raising the age in New Zealand
Lifting the age of eligibility for NZS has been the most discussed lever to help sustainability
and is widely seen as necessary for fiscal sustainability. The New Zealand Treasury
(2013),40 amongst other options, modelled the possibility of raising the eligibility age for
NZS for the Long-term Fiscal Projections in light of an ageing population and ever
improving longevity.
This option may appear logical as the average length of time spent in retirement lengthens
with improvements to longevity. Figure 14 shows these improvements over time with
some possible levelling off since 2013-15 and a compression of the difference between life
expectancy for men and women.
Figure 14. Average life expectancy at 65 (StatsNZ, 2019)

Retirement Commissioners have emphasised the need to raise the age in the last three
three-yearly reviews of retirement incomes policy (Retirement Commission, 2010, 2013;
Retirement Commissioner, 2016). The 2013 Retirement Income Review suggested that
the age of eligibility should reflect improving life expectancy and the lengthening of

39

Minister of Families, Children and Social Development Mandate Letter from the Office of the Prime Minister:
https://pm.gc.ca/en/mandate-letters/minister-families-children-and-social-development-mandate-letter.
40
There was no inference that Treasury supported this policy, but it was modelled as an option.

31

 average length of time spent in retirement. One way to do this, was to “keep constant the
proportion of life in which people receive publicly funded retirement income” (Retirement
Commission, 2013, p. 36).
Figure 14 however disguises the spread of mortality experience. As Figure 15 illuminates,
while the most common age of death is 90 there is a wide spread of actual experiences.
Figure 15. Estimated number of deaths at each age (from 65 to 100) for 100,000
female New Zealanders who reach their 65th birthday in 2015 (O'Connell et al.,
2015, Fig. 6)

By the time today’s 25 year olds reach 65, it is expected men will live another 25 years
and women another 27 years on average (Commission for Financial Literacy and
Retirement Income, 2013 p.36).41
Much of the gain in life expectancy at birth has
been attributed to the gains made at older ages
with improving medical technology likely to
further increase life expectancy at older ages over
time. A woman aged 90 today can expect to live
on average another 4.6 years, a 60% increase
since 1950-52.42 Thus it is possible for life
expectancy to increase because there is an
extension of life at older ages rather than

A recent comparative study of Denmark and
Sweden has shown improvements in health
in cohorts of female centenarians in
Denmark but no such improvements in
Sweden. This difference may be explained
by reduced spending on public services,
including healthcare for the elderly, in
Sweden in the early 1990s, due to a series
of economic crises. While no evidence of an
increase in Sweden was found in recent
years in Denmark the very oldest were
observed to die at higher and higher ages,
and the age at which only 6% of
centenarians survive rose consistently
among those born between 1870 and 1904
(Medford, Vaupel, & Christensen, 2019).

everyone living longer. This is a plausible picture of the future when the inequality in
access to health outcomes and lifestyles help the longer lived and healthy to live yet

41
42

Discussed further in St John and Dale (2019)
See http://archive.stats.govt.nz/browse_for_stats/health/life_expectancy/NZLifeTables_HOTP12-14.aspx.

32

 longer, while others experience declining life expectancy (O'Connell, et al., 2015).
Evidence provided for the CFFC 2019 review from the Health, Work and Retirement
surveys suggests that around 15% of adults in the longitudinal study were on a trajectory
of financial hardship as they entered retirement, with this group also experiencing ill health
and disadvantage (Allen, 2019)
Figure 16 illustrates the different life expectancies for Maori and non-Maori. Not only is the
life expectancy worse for both male and female Maori (there is a similar result for Pasifika
peoples), but Maori spend fewer years without disability and spend a higher proportion of
their total years in retirement with a disability.43
Figure 16. Maori & non- Maori life expectancy at age 65 (Ministry of Health, 2013)
While life expectancy for
Maori is increasing, the
incidence of ill health,
rates of poverty and
benefit receipt are also
higher for this population.

Figure 17 shows the projected increase in disability for the Maori population compared to
the non-Maori.
Figure 17. Estimated increase in population disability (Kerse, 2019, Fig.10)

43

See https://www.health.govt.nz/system/files/documents/publications/healthy-ageingstrategy_june_2017.pdf.

33

 New Zealand’s experience with raising the age
While the state pension age was increased rapidly by 5 years, with little warning, over the
period 1992-2002, the consensus today is that raising the age needs to be well-signalled.
The recommendation from the 2016 Retirement Review44 was to increase the age of NZS
eligibility to 67 over eight years between 2027 and 2034. The National Government found
even that an insufficient lead in-time and sought a mandate at the 2017 election for a 20year notice period with a rise to 67 phased in over four years between 2037/38 and
2040/41.45
Giving an extended period of notice would ensure families and individuals will have
time to adapt to these changes. (English, 2017)
The Labour-led coalition elected in 2017 was opposed to the raising of the age and Prime
Minister Jacinda Ardern promised to resign rather than raise the age.
It wasn't irresponsible to leave the age of eligibility at 65, despite New Zealand's
growing population, she said. It's irresponsible not to save for it. … Ardern said she
was moved by people in hard labouring jobs who said they could not continue
working beyond 65. (Walters, 2017)
While both major political parties have shown a reluctance for a rise in the age, New
Zealand‘s

closest

neighbour,

Australia, began

to

increase the age for the Age Pension in 2017 and will
achieve 67 after only six years in 2023.
Rather than aligning the age with Australia, New
Zealand’s reciprocity agreement with Australia has
been revised so that:

Australia’s Age Pension age has
been slowly increasing from 65 to 67
years. From 1 July 2019, the age for
the Age Pension is 66. It increases by
6 months every 2 years until Age
Pension age reaches 67 on 1 July
2023. These increases only affect
people born on or after 1 July 1952.
(Centrelink, 2019)

… in order to claim the Australian pension or
New Zealand superannuation under the agreement, an individual must have
reached the higher of the pension eligibility age in both countries, regardless of the
country in which the claim is made.46
Both the current policy of raising the pension qualifying age in Australia to 67 and the
threat to raise it further have been controversial. In July 2019, the age at which a person
can access the pension increased to 66 from 65.5 years. The Sydney Morning Herald
reports:
Older Australians struggling to make ends meet or looking to boost their quality of
life are flooding the national jobs market in record numbers but many are finding
their skills and experience unwanted by prospective employers.
National Seniors chief advocate Ian Henschke said many older Australians faced
prejudice as they tried to get a job which would worsen as more people sought
work in their 60s.He said people made redundant in their late 50s or early 60s often

44

See https://www.cffc.org.nz/reviews-and-reports/retirement-income-policy-review/.
The National opposition has confirmed in 2019 this is still their policy.
46
See https://www.msd.govt.nz/about-msd-and-our-work/publications-resources/regulatory-impactstatements/revised-social-security-agreement-with-australia.html.
45

34

 faced years on Newstart as they tried to get back into work before accessing the
more generous age pension. (Wright, 2019)
Some of the debate about raising the age yet further is outlined in Box 1.
Box 1. Recent controversy in Australia around raising pension age
In 2014, Treasurer Joe Hockey’s budget was the first to propose increasing the Australian Age
Pension qualifying age to 70 years, starting in 2053. That proposal to increase the pension age
was later scrapped.
Then in June 2017, Australians born in or after 1966 heard Social Services Minister Christian
Porter confirm the Turnbull government would increase the pension qualifying age to 70 by 2035,
18 years earlier than the original proposal.
Subject to legislation, we are proposing to progressively increase the age pension age …
from 67 to 70 years, over 10 years from 2025 to 2035. By 2054/55, the number of
Australians aged over 65 will more than double to 8.9 million, representing about onefifth of the total population and placing enormous strain on the age pension system. The
cost to taxpayers of the age pension would rise from $39.5 billion in 2013/14 to $72.3
billion in 2023/24. As the measure is rolled out, the government expects it to raise $3.6
billion between 2025 and 2029 and more in later years. (Porter, quoted in Myer, 2017)
The NewDaily (5 June 2017) quoted some strong objections to the policy from the Combined
Pensioners and Superannuants Association:
It’s unreasonable. They don’t say what people who lose their jobs before they’re 70 and
can’t get another one will do. The move could threaten the superannuation of older
people with small balances who become unemployed…. If they put their super) in pension
mode it could threaten their Newstart rights under the assets test, but people don’t
understand that.
Sandra Buckley, CEO of Women in Superannuation, highlighted consequences for women:
It is predominately women who earn less than $37,000 and who have low super
balances. Women are consistently being paid less and earning less than men over their
lifetime so once again, they will be penalised in later life for the gender pay and super
gaps that are proving to be extremely difficult to overcome.
David Atkin, CEO of building industry super fund Cbus, said:
Shifting the access to the age pension to 70 is a “blunt policy tool” that fails to consider
the needs of manual worker. Our members simply can’t keep doing these physically
demanding jobs until they’re 70.
Dr Cassandra Goldie, CEO of the Australian Council of Social Services, said:
The move will mean unemployed older workers would be pushed onto Newstart benefits
which are far lower than the pension until they reach 70. We will consequently see a rise
in poverty among the over 65s.

How much is saved by raising the age?
As described in Section 4 and in reference to Table 9, current policy settings in New
Zealand projected out to 2060 are unsustainable: deficit costs feed an unconstrained
growth of net debt and negative net worth.
Will raising the age address this problem?
Treasury have modelled fiscal projections based on a set of policy changes that stabilise
net debt by 2060;
•
•
•
•

long-run tax to GDP is lifted by 1%,
NZS eligibility age is raised to 67 by 2027/28, with 1 year increments every 15
years to reflect improved longevity,
health growth is restricted to GDP +1% &
other spending reduced so net debt is held below 20% of GDP.

35

 This scenario was produced as an illustrative example, of one out of many potential sets
of options, to address future fiscal pressures. It does not represent government policy or
Treasury advice, and should not be construed as being either of those.
Figure 18 illustrates that this combination stabilises net debt compared to the
unconstrained scenario and separates out the component contributions to that
stabilisation.
Core Crown expenses as %s of Nominal GDP
Percentage
(%)Core Crown expenses as %s of Nominal GDP in history and different
Figure
18.
in history and different scenario projections
12
scenario
projections (Bell, 2019a).
11
10

9
8
7
6
5

4
3
2
1

0
NZS

Non-NZS welfare
History 1997/98

Health

History 2017/18

Education

Other operational
expenses

Unconstrained 2059/60

Interest costs

Stabilised 2059/60

The raising of the age for NZS appears to reduce the % of GDP spent on NZS by around
0.8 percentage points by 2060. This is about 10% saving of the gross costs of NZS. A
saving of 10% of NZS eventually is significant but this saving has been achieved by raising
the age to 69 by the end of the projections period and there will be many offsetting costs.
Given the profile of pre-retired people discussed in section 3, a significant proportion will
not be able to support themselves for these extra years and may further run down their
assets and increase their debts to be even worse off at age 67 by 2027/8, and at 69 by
2060.
A critical disadvantage of relying on raising the age of eligibility to improve NZS
sustainability is that many people with physically demanding jobs are disabled or sick by
age 65 and unable to work further. Others lack the required skills or education to meet
the changing market requirements. The cost saving from raising the age would have to
take account of the costs of supporting people who could not work past 65 and would
require another form of state assistance. Retraining also is another cost and could not
provide the whole answer for those who struggle to work past 65. The use of conventional
welfare benefits with stringent income tests are particularly harsh for those over 50 who

36

 spend sustained times unemployed, care-giving or incapacitated.

They are likely to

exhaust their private retirement resources before reaching the new higher age of eligibility.
Table 1 showed the net Supported Living Payment (SLP) for a married person is only
$11,860 compared to $$16,446 for NZS. There is an over $9000 annual difference for a
couple. A couple on the Jobseeker benefit, has a joint income of only $18,976 per annum,
a difference of nearly $14,000 per annum to a couple on NZS, and they can earn only $80
between them before harsh abatement occurs to make any extra not worthwhile. The
welfare benefits are well below usual poverty lines and supplementary means-tested
assistance will normally be needed. This illustrates that the raising of the age, under
current welfare policies would have a severe impact on those who cannot support
themselves.
Another consideration is that the voluntary work of the retired population has an economic
and social value, projected to be worth around $47 billion (2016 dollars) by 2061 (Office
for Seniors, 2018, p. 14). Grandparents also make a large contribution to the care of their
grandchildren and other significant community activities. An entitlement to a basic income
at age 65 offers the option for many to retire from paid work to do valuable voluntary
work. It is also a very effective way to support older workers who wish to do paid work,
but to work fewer hours and have more flexibility than full time work. It enables the
combination of the benefits of some work with the benefits of income to support other
worthwhile activities.
In summary: some raising of the age may prove inevitable to reflect improved average
longevity, greater participation in the workforce and to align with other countries such as
Australia. However, if the only way to do this politically is to give a long lead-in time,
there will be little or no potential for immediate savings from using this lever. Moreover,
there may be substantial disadvantages both for low income individuals and for society
who must cope with the costs of unemployment and increased poverty and loss of older
persons’ socially valuable unpaid contributions. While it is a potential source of some
saving, there will be many offsetting costs to factor in.

6. Reduction in rates of NZS
If raising the age cannot deliver needed savings for the foreseeable future, then a second
possibility is to reduce the payment level of NZS. Politically this would be difficult, while
welfare benefits were cut nearly 30 years ago, that was under a first past the post
government.

A gentler approach would be to achieve cuts over time by changing the

indexation basis for NZS, another is to reform the marital status and living arrangements
basis of determining rates of NZS.

37

 Indexation
Projections show that fiscal savings from indexing the annual payment of NZS to inflation
(the CPI) rather than wages would lead to significant long-term savings (The New Zealand
Treasury 2009, pp. 57-58). The real spending power of NZS would be protected but the
rate of NZS would fall relative to average wages.
After a period of time, it is likely that the same issues of poverty found among the
beneficiary community would emerge. There are two other objections: the CPI is not a
good measure of the cost of living for low income people as it seriously underweights basic
items such as food, rent and interest. A second problem is that it conflicts with an intrinsic
value underpinning NZS: older people should have enough income to feel they can
participate and belong to society. Having a wages link for NZS is a crucial protection of
this value.
The lowest quintile of disposable income has very different patterns of spending when
compared to the middle income quintile. Other price indexes can be developed that better
reflect the expenditure patterns of lower quintiles. Figure 19 below shows these differences
Figure 19. Diversity in spending patterns between middle and lowest disposable
income quintiles47

Household living-costs price indexes (HLPIs) do provide measures of the inflation
experienced by specific household groups, such as beneficiaries or Māori. The
resulting index better reflects the actual inflation experienced by the group. For
example, low-income households typically spend proportionally more than average
on household energy and grocery food. They spend proportionally less on new cars,
international airfares, and restaurant meals.48

47
48

See https://www.stats.govt.nz/methods/measuring-child-poverty-fixed-line-measure.
See https://www.stats.govt.nz/methods/measuring-child-poverty-fixed-line-measure.

38

 The problem is, even if a HLPI price index is used that is better suited to set the level of
NZS for low income superannuitants, it still only protects an absolute living standard. As
average living standards rise, lower quintiles of retirees will fall further behind others.
Figure 20 shows the dominance of NZS as a source of a long-term income support for low
decile superannuitants. Any reduction in the level of NZS may have a dramatic effect on
social inclusion of older people over time.
Figure 20. Percentage of income from government sources (Perry 2019
forthcoming)

The great majority of older New Zealanders (aged 66+) are very dependent on NZS
and other government transfers for their income. 40% have less than $100 pw
from other sources, 40% of singles have no other income, the next 20% have on
average around 70% of their income from NZS and other government transfers…
[Couples] generally have higher per capita non-government income than do those
in single person EFUs. (Perry, 2019 forthcoming)
In summary, the level of NZS needs to be high enough to prevent hardship and while
the current level does that for most, particularly for those who are home-owners, some
pensioners clearly still struggle. Figure 20 shows how dependent at least 40-50% of
superannuitants are on NZS. Single supernnauitants on average get nearly 60% of their
income from NZS.
Various studies show that a healthy retirement that allows participation in society requires
additional income stream of around $10,000-$12,000 as discussed in St John and Dale
(2019). This suggests the current level of NZS as a long-term support for people with no
or little other income is far from generous.
There are also lessons to be learned from the failure to index welfare benefits to wages
that has helped create serious levels of poverty (Welfare Expert Advisory Group, 2019).

39

 Women, already at risk (see Figure 21), are the most affected in the welfare system and
it is likely they will also be most affected by price indexation of NZS.
Figure 21. Women are more at risk of old age poverty: relative income poverty
among 66+ year olds, 2014 or latest available (OECD, 2017b, p. 39)

A Westpac NZ survey of more than 1,000 KiwiSaver members found that women are
contributing less into KiwiSaver than men, have lower average balances and are less likely
to have other investments that can pay for their retirement. Nearly a third of women in
the survey have less than $5,000 in their KiwiSaver accounts (vs 19% of men) while only
4% of women had more than $50,000 (vs 13% of men). The national gender pay gap
remains at 9.3%,49 and this research shows the flow-on effect on retirement savings.
“Simply put, it appears men have more income available to set aside for retirement and
this is impacting KiwiSaver balances" said Simon Power, Westpac's general manager of
consumer banking and wealth (Personal Finance, 2018).
Figure 22. Median value ($) of KiwiSaver schemes, by age group and sex, year
ended June 201850

49

As reported in NZ Herald, 22 August 2019, ‘More work to do’ on pay gap: PM, p. A7.
Note error bars show variability in the estimates https://www.stats.govt.nz/news/kiwisaver-nest-eggs-grow5000-in-three-years.
50

40

 As shown in Figure 22, and as has been the case in the Australian compulsory scheme,
median balances in the KiwiSaver savings scheme for women in the older working age
(40-54 years) and those approaching retirement (55-64 years) are significantly below
those of men. The biggest gap is for those still contributing after 65 years.
Therefore, reducing either the level of NZS or the relativity to wages over time can be
expected to undermine the desirable achievement of low hardship rates for the 65+ group,
adversely impact on women, and further increase the expenditure on supplementary
assistance.

Rationalisation of different rates of NZS
Another approach to decreasing the level of NZS is to rationalise the three different rates
for NZS. As shown in Table 1, there is a net married rate, a single sharing rate at 60% of
the married rate, and a single living alone rate at 65% of the married rate. As previous
Retirement Commission Reviews and Periodic Report Groups Reviews (eg, 1997, 2007,
2010) have noted, the difference between single sharing and married rates are hard to
justify.
Rationales from the MSD (2016, pp. 10 - 11) are:
Couples living together in a married, civil union or de facto relationship are paid
less than double the single rate because it is considered that they can take
advantage of certain economies of scale that individuals in shared accommodation
cannot. Thus, the rate paid to a married person is less than that paid to a single
person. For example, a married couple:
•

could be able to enjoy lower accommodation costs than two single people

•

could be able to have their personal household effects on one insurance
policy whereas two single people who are sharing accommodation would be
more likely to have separate insurance costs totalling a higher amount

•

could share vehicle expenses, while two single people may be more likely to
have their own individual transport and vehicle costs

•

could generally share meals, while two single people
accommodation may not have merged their lives to that extent.

sharing

None of these rationales are particularly convincing or substantive. It is fair to say the
different rates are historical and they are unsuited to a modern world of flexible living
arrangements and relationships (St John, MacLennan, Anderson, & Fountain, 2014). There
are many more factors at play than just culture, but

Figure 23 shows that living

arrangements are more varied for the Maori population with at least 25% living with
‘others’ and far fewer with a spouse.

41

 Figure 23. Living arrangement by sex and ethnic group51 (Kerse, 2019)

Current policy also criminalises those who commit ‘relationship fraud’ by failing to declare
relationships in the nature of marriage. Healey & Curtin (2019) found from an Official
Information Act request that only 122 superannuitants were prosecuted in 2017/18. While
the MSD do not as aggressively pursue superannuitants as they do other welfare
beneficiaries for this so-called crime, aligning the rates will eliminate the anxiety many
may feel around their relationship or living alone status.
In summary: There is a case therefore to pay the same flat rate to everyone, set
somewhere between the married person and single sharing rate. As shown in Table 4,
around 25% of superannuitants live alone and possibly the majority of these would need
accommodation assistance. The elimination of the living alone rate would require an
additional needs-based payment where high housing costs are demonstrated. The meanstested AS is currently already accessed by 42,000 of those over 65 (Table 3). This payment
could be further adapted for the over 65 group to assist with high housing costs,
independently of whether superannuitants are sharing, married or living alone.
Using the data in Table 4 and the rates of NZS in 2019 (Table 1), paying all NZS recipients
the married rate reduces the gross cost by around $1.3 billion (8.5%). If all were paid the
single sharing rate, the gross cost increases by around $1.3 billion and by $2.6 billion if
all were paid the living alone rate.
Whether or not there is a separate rate for living alone, the alignment of the married and
single rates appears justified.52 To save costs without direct cuts, the single sharing rate
could be frozen until the married rate catches up with normal annual adjustments.

51

LiLACS NZ first wave of data collection.
The Retirement Commissioner’s Review (2010, p. 13) endorsed the alignment of the single sharing and married
person rates In the interests of simplicity it suggested that the living alone rate remained unchanged.
52

42

 Alternatively the single rate could be CPI indexed until the married rate indexed to wages
catches up.
In an April 2019 paper to the Cabinet Social Wellbeing Committee, the Office of the
Minister for Social Development signalled an intent to change some aspects of NZS that
would move the scheme towards more individual entitlement by ‘modernising and
simplifying.53 So far this affects only the non-qualifying partner (NQP) and spousal
deduction, but it may also indicate a willingness to move in the direction of aligning rates.

7. Means testing and alternatives
This leaves the ‘third rail’54 of superannuation policy:
some form of ‘claw-back’ from those who do not ‘need’ it.
This has been a politically unattractive option because of
New Zealand’s history, (see St John, 2015; St John,
2018b).
There are a number of ways to save costs by reducing
access to NZS by the well-off. Probably few people would
wish to contemplate a welfare-type income test as in the
benefit system. The last time that was attempted (1991
budget) there was outrage and anger among the powerful
superannuitants’ lobby and the legislation that would have
changed NZS into a welfare benefit was reversed before it

Means tests take other income
and
assets
into account
in
determining the amount of benefit a
person is entitled to. A simpler
version is an income test alone.
Welfare benefits in NZ are subject
to a stringent income test that aims
to target payments to only those
who ‘need’ them.
A gentle test that affects only the
top end may be described as an
affluence test.
A progressive income tax and a
taxable
benefit
automatically
ensures some income testing or
clawback.
A basic non-taxable income and
other income taxed at progressive
rates is another way to operate an
affluence test.

was implemented (St John, 1999; St John & Ashton, 1993).
The Australian means test described in Box 2 takes account of both income and assets,
using joint income and assets for couples. While 80% of those eligible by age get at least
some age pension55, this model is most unlikely to be acceptable for New Zealand.
However, an income-based test, would not preclude counting as much imputed income
from assets as feasible over time.

53

See https://www.msd.govt.nz/about-msd-and-our-work/publications-resources/information-releases/nzs-vpmodernisation-and-simplification.html.
54
Touch it and you die. The phrase ‘third rail’ is a metaphor in politics to denote an idea or topic that is so
‘charged’ and ‘untouchable’ that any politician or public official who dares to broach the subject would invariably
suffer politically. The third rail in a railway is the exposed electrical conductor that carries high voltage power.
Stepping on the high-voltage third rail usually results in electrocution. The use of the term in politics serves to
emphasise the ‘shock’ that results from raising the controversial idea, and the ‘political death’ (or political suicide)
that the unaware or provocative politician would encounter as a result. (Wikipedia).
55
See https://www.austaxpolicy.com/better-targeting-australias-age-pension/.

43

 Box 2. Age Pension means test in Australia56
Rates of Age Pension March 2019 (including energy and pension supplements)
Single: A$926 per fortnight (A$24,081 per year)
Couple (each): A$698.10 per fortnight (A$18,051 per year)
Couple (combined): A$1,396 (A$36,301 per year)
Couples separated due to illness each receive the Single rate
Income test
To qualify for a full Age Pension a single person must have income below A$174 per fortnight
(A$4,524 per year). Above that the age pension is reduced by 50% for each extra dollar until
there is no entitlement by $2,026 per fortnight ($52,686 per year).
As in New Zealand, the Australian couple rate of the full Age Pension is lower than for two single
people. A couple’s combined income must be below A$308 per fortnight (A$8,008 per year), for
a full Age Pension. A part Age Pension is payable up to a joint income of $3,100 per fortnight
(A$80,610 per year).
From1 July 2019, there is a work bonus: earnings up to $300 per person per fortnight from
working is not included in the Age Pension income test.
Assets test
To qualify for a full Age Pension a single person’s assets must below A$263,250 if a home owner,
or A$473,750 if not. A part Age Pension is paid when assets are below A$572,000 for a
homeowner, or A$782,500 if not. For a couple combined assets must be below $394,500 if they
own their home, or $605,000 if not. A couple may be eligible for a part Age Pension if assets are
worth up to $860,000 if they own their own home, or $1,070,500 if not.

From 1985 to 1998 New Zealand operated a surcharge on superannuitants’ other income
(Preston, 2001). This too was unpopular and complex for people to understand.
Nevertheless, it did deliver useful savings.
While the surcharge was complicated and contentious, it performed a useful costsaving function without imposing hardship. Some better-off retirees did not bother
claiming the state pension, and most of those still in high-paid work received little
after-tax benefit from it.
The fiscal cost of abolishing the surcharge in 1998 was estimated to be $400m or
10% of the net cost of NZS. This indicates that the surcharge created a 10% fiscal
saving on the net cost of NZS. (St John, 2015, p. 8)
In New Zealand, the challenge is to find a way to apply an income (or “affluence’ test) that
could be seen as fair simple and acceptable, with enough useful savings to take the
pressure off relying solely on raising the qualifying age or reducing the rate of NZS as the
principal levers.
The justification for considering this lever seriously is intergenerational equity. Wealthy
recipients of NZS may still be in well-paid work and/or have other large private incomes
and assets, and sometimes annuities (see St John and Dale 2019). This group is likely to
have accumulated their wealth with tax-free capital gains and may have gained
substantially from the 2010 income tax cuts and lower Portfolio Investment Entity (PIE)
rates of tax.57 Under the PIE regime, the top rate of tax is 28% compared to the top rate

56

See https://www.superguide.com.au/accessing-superannuation/age-pension-rates.
“The number of super-rich earners on the New Zealand taxman's radar has skyrocketed in the past five
years with 350 people now worth more than $50 million — some of whom are in a fight over more than $85
57

44

 of 33%, providing a five-percentage point advantage. Increasingly the younger working
age population who are struggling in the property market and may have large student
debts are questioning this largess.
Figure 2458 for the case of a married superannuitant illustrates this generosity by showing
the addition to net income provided by NZS at all income levels. Where there is no other
income the NZS payment is a net $16,446 (Table 1). When earned income exceeds
$70,000, NZS is taxed at the highest marginal rate so that the effective net NZS payment
is reduced to $12,577. No matter how high the earned income is once 33% is reached the
additional income is always $12,577 as shown in Figure1. For those on higher rates of
NZS, the net gain at higher incomes is more: $15,230 for single sharing and $16,564 for
living alone.
Figure 24. Disposable income with NZS, married rate NZS 1 April, 2019
Disposable income

The 1993 Accord59 endorsed the principle that the net amount of NZS should reduce as
total income increases, ether by a surcharge or a progressive tax regime that had
equivalent effect. The 1997 Periodic Report Group on Retirement Incomes noted:
We strongly support the sentiment that there are higher priorities for government
resources. Therefore we regret the impending abolition of the surcharge…
The abolition of the surcharge will provide a breathing space in which we can inform
and educate the community about the future shape of public provision and explain
why some kind of targeting mechanism will be needed in future. (Periodic Report
Group, 1997, p. 47)

million in disputed tax. The group - labelled High Wealth Individuals (HWI) by Inland Revenue - has soared 75
per cent since 2013, while the country's poorest residents continue to struggle.” (Kirkness, 2019)
58
Figures in this section are updated from St John (2018).
59
Between the three major political parties: Labour, National and the Alliance in 1993.

45

 As suggested above, other levers may impact unfairly on those least able to manage. In
contrast, only those with significant ‘other’ income will be affected if a clawback from the
top end can be devised in a simple and fair way. Finding a way for the top line to meet the
bottom line in Figure 24, by reducing the generosity of net NZS at the top end is worth
exploring.

The New Zealand Superannuation Grant- a blue skies proposition60
This section draws on St John (2018b) and suggests various options for a tax-based clawback to improve sustainability with modelling to show approximately how much revenue
could be saved.
A basic income approach might offer this possibility and align with the Government’s future
of work programme, led by the Productivity Commission (2019), where the 21st century
workplace no long provides certainty of employment or sufficient hours of work for many
workers.
In a basic income approach, each person has a universal grant that is not part of taxable
income. When additional income is earned, it is taxed under a progressive tax regime so
that the tax system does the work of providing a claw back of the universal grant for high
income people. Extra income is not unduly discouraged. The attraction for using this
approach for NZS is that it retains simplicity and universality while reigning in the
expenditure at the top end and providing some useful additional revenue to balance
intergenerational concerns and to reduce the inequality within retirement.
It also offers a compromise between aggressive means testing as applied in the welfare
system (or the affluence test as applied in Australia), and a fully universal taxable pension
approach such as for the current NZS. It offers people flexibility in their employment
choices in early retirement.
Taking a ‘basic income’ approach may be simple to implement and operate but it requires
a new way of thinking. The basic income, named here the ‘New Zealand Superannuation
Grant’ (NZSG), would be paid to all superannuitants as a weekly non-taxable grant. Then,
for any other gross income, a separate tax scale would apply for each additional dollar. 61
For illustrative purposes it is proposed that the NZSG is the same for everyone (married;
single sharing; single living alone) and any extra supplement for high housing costs would

60

The authors gratefully acknowledge the modelling of these results provided with the help of Matthew Bell , NZ
Treasury but this in no way implies any endorsement of these policies.
61
Paying the pension as a non-taxable grant and a progressive tax on other income makes the pension analogous
to universal payments such as the old Family Benefit. It fits the labour government’s ideas of progressive
universalism, introduced with Best Start, Winter Energy Payment, tertiary study fees.

46

 be part of the welfare system.62 While the NZSG could be set at any level, Figure 24 shows
it as equal to the current (after-primary tax) rate of NZS, i.e. $16,564 for a married person.
A break-even point exists (Figure 25) where the NZSG, plus extra income from work or
investment, net of the new tax rate, is equal to the disposable income of an ordinary
taxpayer paying the usual rates of income tax. This point is effectively where the gain from
the NZSG has been effectively clawed back (i.e. offset by the additional tax).
The scenario depicted in Figure 25 with a flat tax at 39% on other income shows the
breakeven point occurs when the NZSG recipient’s ‘other’ income is $123,000.
Figure 25. Scenario 1. Flat tax of 39% on other income: married rate NZS 1
April, 2019
Disposable income

This proposal is technically different to the surcharge of 1985-1998 because the NZSG
payment is not part of taxable income. The surcharge was exceedingly complex, applying
until the net advantage from NZS was equal to the surcharge paid and could mean different
end points (when NZS had been fully clawed back) for different taxpayers. Few could follow
the calculations. The surcharge was also perceived as an additional, discriminating tax that
could result in marginal rates of tax exceeding 50% (see St John (1991) for further
discussion of how the surcharge worked).
Given that for 80% of NZS recipients, NZS provides the majority of their income, a tiered
structure may be useful to give some relief to those with limited extra income. Figure 26
illustrates a tiered scenario; with rates of 17.5% for the first $15,000 of other income, and
39% on each dollar above that. At $ 177,000, the break-even point in this case is much

62

As shown in Table 3 around 180,000 superannuitants get the single, living alone rate. Of these, many would
continue to require a supplementary payment to reflect higher costs. A suitably modified accommodation
supplement may be required.

47

 higher than in Figure 25. Figure 27 offers a third intermediate tax scenario, and shows a
cut out point of $101,000
Figure 26. Scenario 2. Two-tiered rate of 17.5% (for first $15,000 earned) and
39% above $15,000: married rate NZS 1 April, 2019
Disposable income

Figure 27 20% tax first $20,000 earned and 45% beyond that; married rate NZS
1 April, 2019
Disposable income

The break-even point is very sensitive to the tax rate chosen, or in the case of a tiered tax
schedule, to the highest rate of the schedule.
In the example scenario depicted in Figure 26 with a flat tax at 39% on other income, if
the recipient of NZSG receives more than $123,000 then it would be rational for them to

48

 either: forego the NZSG and be treated as an ordinary taxpayer, or to apply for a refund
of any tax overpaid on income above the cut-out at the end of the year.
Whether other income is from paid work or from investments, and whether it reduces or
disappears, the right to the basic income floor of the NZSG remains. Thus, the NZSG is
the prototype of a basic income that provides automatic income security as of right.
There are annual losses in disposable income relative to current settings as shown in Table
11 for the three tax scenarios depicted in Figures 25, 26, and 27 at bands of extra income
earned. Any losses for people with small amounts of additional income these may be
compensated in other ways,63 or minimised as in the two-tiered tax approach of tax
scenario two.
Table 11. Losses of non-NZS disposable (rounded) income for superannuitants
relative to current system 2019/20 NZS married rate. 64
Non-NZS taxable income ($ per year)
Tax Scenarios
One

Two

Three

$5,000

$1,071

-$4

$121

$10,000

$2,146

-$4

$246

$15,000

$3,221

-$4

$371

$20,000

$4,296

$1,071

$496

$25,000

$5,371

$2,146

$1,871

$50,000

$8,153

$4,928

$6,153

$75,000

$9,691

$6,466

$9,191

$122,800

Around Cut-out
point for One

$12,559

$9,334

N/A

$176,500

Around Cut-out
point for Two

N/A

$12,566

N/A

$103,100

Around Cut-out
point for Three

$11,377

$8,152

$12,563

Once in place, the NZSG would be far less complicated than other forms of clawback such
as the surcharge, a welfare-type means-test directly on NZS, or even a negative income
tax approach. As with any targeting regime, an increase in the degree of targeting will
result in some avoidance activity. New Zealand’s past experience shows that opportunities
and incentives for tax avoidance were features of the surcharge. It must be noted here
however that the NZSG proposal is not nearly as harsh as the welfare means-test that
applies to rest home care subsidies (see St John and Dale, 2019). It provides a gentle

63

The Winter Energy Payment could be added in to the NZG for example.

49

 clawback using the principle of progressive taxation which, it can be argued, is the natural
counterpart of universal provision. The NZSG is consistent with the current arrangements
that do not require any retirement test and therefore there is no disincentive to work.
Another concern may be that the NZSG would need to be carefully packaged so as not to
adversely influence the decision to save. This of course would be much more of a problem
if a full means-test was proposed including an asset-test rather than the proposed incometest operated through the tax system.

PIE tax regime
The integrity of the NZSG approach would require that the top PIE rate be aligned so that
for the tax scenarios given, the top NZSG PIE rate would be 39% or 45% for the third
scenario. Alternatively, gross PIE income could be imputed as ‘income’ to be taxed at 39%
less the tax already paid by the PIE on the member’s behalf (similar to the imputation
regime).65 The same argument applies to income earned through trusts, companies and
overseas vehicles.66 PIE income is included in the other income used to abate tax credits
in the other major redistributive programme in New Zealand –Working for Families.67
The holding of net financial wealth is highly skewed, favouring older age groups. It is
therefore likely that there is significant PIE income among the top 20% of superannuitants.
The costing done in the next section does not capture undertaxed PIE income as it is not
included in the Household Expenditure Survey data base.
Treatment of current annuities and defined benefit pensions raise other complex but not
insoluble problems. In the past, such annuities were apportioned 50% as income for
surcharge purposes. There may also be opportunities. If for example there was desire to
encourage annuitisation, an annuity to a limited value could be added to the NZSG grant
instead of apportioned 50% as income as a means of making it attractive to middle income
people in the absence of compulsory annuitisation (St John, 2016; St John & Dale, 2019).

Costing68
The fiscal saving possible by using the NZSG approach depends on the decision about the
alignment of rates and critically on the tax rates chosen. If the degree of targeting was
similar to the surcharge as it operated at the end of the 1990s, savings of the order of
10% or more could be expected (Periodic Report Group, 1997).

65

There is a case for using a consistent definition of taxable income for everyone, not just superannuitants. Thus
extensions to the definition of taxable income, such as apply in Working for Families, might capture other possible
avoidance activity.
66
The issues around the need for an overall reform of these vehicles so that they are taxed at the individual’s
appropriate marginal tax rate are explored in Chamberlain & Littlewood (2010, 2019).
67
See https://www.ird.govt.nz/situations/i-am-a-pie-investor-with-a-student-loan-or-working-for-families.
68
The authors gratefully acknowledge the Modelling of these results provided with the help of Matthew Bell , NZ
Treasury but this in no way implies any endorsement of these policies.

50

 Table 12 shows the current projected costs and average numbers of superannuitants. A
surcharge equivalent for the 2019/20 year, would produce saving of $1.54 billion in gross
terms. Moreover, in contrast to raising the age such savings could be reaped much sooner.
Table 12. New Zealand Superannuation (NZS) aggregate expenditure and
average recipient numbers over June-end (fiscal) year - history, forecast &
projection (The Treasury, 2019)
Fiscal Year (Year ended 30 June) →

2017/18

2018/19

2019/20

2020/21

2021/22

2022/23

Nominal Gross Domestic Product (GDP) ($billion)

288.812

299.713

316.957

334.009

350.210

366.568

Gross New Zealand Superannuation expenditure ($billion)

13.699

14.562

15.488

16.384

17.409

18.468

Gross NZS expenditure as percentage of nominal GDP

4.7%

4.9%

4.9%

4.9%

5.0%

5.0%

11.625

12.333

13.095

13.824

14.660

15.528

Net NZS expenditure as percentage of nominal GDP

Net (of tax) NZS expenditure

4.0%

4.1%

4.1%

4.1%

4.2%

4.2%

Average number of NZS recipients in Fiscal Year (thousands)

741

767

794

821

846

873

Annual percentage growth of NZS recipients

3.4%

3.5%

3.5%

3.3%

3.1%

3.1%

The most recent Household Expenditure Survey was used by the New Zealand Treasury to
model the cost saving from implementing a possible NZSG for the March-end year
2017/18. This costing was produced to assist in the production of this paper. It does not
represent government policy or Treasury advice. The Treasury model assumes:
•

All eligible people elect the option that delivers the higher disposable income, even
if only by $1 per annum. In other words, the only people who turn down the NZSG
are those whose non-NZS income exceeds the ‘break even’ point, where they would
end up with the same disposable income under either option.

•

There are no behavioural responses, in particular, no change to labour supply or
average hours worked by eligible superannuitants.

There are a total of 12 scenarios, which are 4 NZS net rate options costed by three different
alternative tax regimes. The 12 scenarios are summarised in the Appendix.
The 4 NZS net rate options are:
1. Anyone who receives NZS gets the net married person rate
2. Any married person who receives NZS gets the net married person rate (including
the non-qualified spouses) and any single person who receives NZS gets the net
single sharing rate
3. Anyone who receives NZS gets the net rate they are currently entitled to
4. Anyone who receives NZS gets the net single sharing rate
The 3 alternative tax regimes are:
a. 39% flat tax rate on all non-NZS taxable income
b. 17.5% on the first $15,000 of non-NZS taxable income and then 39% on non-NZS
taxable income above $15,000 per year
c. 20% on the first $20,000 of non-NZS taxable income and then 45% on non-NZS
taxable income above $20,000 per year
The true cost to the government of providing the public pension is the aggregate net
(after-tax) NZS expense. Relative to its value in each year, costed under the NZS and
personal tax regime existing in that year, modelling showed that savings in net NZS was
possible for all net rate options and alternative tax regimes except net rate option number

51

 4 where everyone has the single sharing rate. The results are summarised for the first
three net rate options 1, 2, 3 above respectively:
•
•
•

Tax regime a (39% tax on all other income) produced overall savings of 23%,
16%, 14%.
Tax regime b (17.5% of first $15,000, 39% on balance) produced overall savings
of 16%, 10%, 7%.
Tax regime c (20% on the first $20,000, 45% on balance) produced overall
savings 17%, 11%, 9%.

These figures assumed an immediate adjustment of all rates to the prescribed NZS net
rate option. In practice the alignment of the rates would be phased in over time and the
savings would increase more gradually. The costings also take no account of the additional
supplements required by many of those living alone with high housing costs. Over time,
as the baby boomers swell the numbers over age 65, savings will likely increase. This will
be reinforced if the tax thresholds for the NZSG tax schedule are unadjusted for inflation.
The scenario of aligning the single living alone and single sharing rate to the married rate
with a flat tax schedule of 39% achieves the most saving (23%). Appendix 1 shows that
over one third or 8.5 percentage points of this saving is due to the alignment of the rates.
The more gentle tax schedule of the 17.5%/39% combination saves 16% of NZS with just
over half of this due to alignment of the rates, and the more aggressive tax schedule of
20%/45% saves 17%.
Even if the net rates are not changed as for net rate option 3, the costings show that 914% savings are possible as modelled under the three tax regimes.
For the scenario of aligning the single to the married rate and a flat tax schedule of 39%,
around 3.7%, or 27,500 of age-eligible superannuitants are unlikely to apply as they would
not gain from the NZSG. If the more gentle combination of 17.5%/39% is applied only
1.5% drop out, but for more aggressive tax schedule of 20%/45% around 4.6% drop out.
It is likely these figures are very much understated as many would find it not worth the
bother to ask for the NZSG especially if they are in well-paid work. It is also likely that
others would be deterred because more income, such as PIE income would be captured.
Thus the savings set out in Appendix 1 for the 12 scenarios are all likely to be
underestimates of the true potential of the NZSG approach. However if the living alone
rate is aligned to the married rate there will be the need for separate assistance with
accommodation costs for many low income retirees.

Summary
The proposed NZSG option that simplifies the treatment of relationship status by paying a
single rate of NZSG for all – a tax-free grant equal to the net amount now paid to a married

52

 person is most effective at saving costs although additional payments for those with high
accommodation costs would be required.
This preliminary analysis suggests that the combined approach of using a separate tax
schedule for other income and freezing the single rates so that over time there is alignment
with the married rate, will give worthwhile savings of at least between 16-23%. If the net
rates are not aligned the savings falls to 7-14%.
Fiscal savings are possible without imposing undue hardship or affecting those with modest
amounts of additional income and can be achieved relatively sooner compared to raising
the eligibility age. Any increase in the eligibility age if longevity continues to improve could
be very gradual with constant monitoring to ensure individuals who, given the arduous
nature of their employment, may expect to retire from work earlier than others, are looked
after.
The NZSG approach does not unduly penalise extra income, depending on the parameters.
Given that for the bottom 60% of NZS recipients, as measured by gross incomes, NZS
payments comprise at least 80% of their total income, and for the bottom 80% of
recipients it comprises at least 55%, the majority of over 65s will face little if any reduction
in disposable income, especially in the 2 tiered tax options. It may therefore be much
easier to introduce than raising the age and hence savings could be reaped earlier.
As with any targeting regime, efforts to maximise returns will lead to some tax planning
activity. However, those who should be paying the top rate of tax of 33% already have an
incentive to reduce their taxable income and some already pay little or no tax. It is
debateable as to whether a marginal 39% or 45% tax rate would substantially change
behaviour but there is the possibility that it could provide the impetus for a full
investigation into, and exposure of, current and potential tax avoidance activities by
wealthy individuals. Under the proposed NZSG, a wealthy person would have to reduce
taxable income to under $15,000 to avoid the 39% rate or $20,000 to avoid the 45% rate
completely.
Where income is in PIE funds the tax regime should ensure that individuals declare their
PIE income and tax paid as they do for dividends from holding shares so as to pay tax at
the appropriate marginal tax rate.
The proposed change would decrease the fiscal cost of NZS through reductions in
payments to high income superannuitants and thus allow more spending or lower taxes
for younger New Zealand taxpayers. It may therefore lead to improved perceptions of
inter-and intra-generational equity.

53

 If it is agreed that the cost of net NZS should be reduced by increasing the degree of
targeting, using the tax system and the proposed NZSG has a number of potential
advantages compared to other targeting regimes:
•
•
•

•

Relatively simple: Simplicity in administration when compared to other income
tests and the old surcharge.
Universality is maintained. The grant is paid irrespective of other income as a basic
income grant if eligible people elect to take it
Flexibility: The choice of tax rates for other income allows flexibility and clarity in
reaching a desired breakeven point and required fiscal savings. It also provides
choice and clarity for very high-income superannuitants who are not denied access
to the basic income floor of NZSG if their situation changes.
The NZSG could be usefully extended as a basic income to other groups such as
those in their 60s on the supported living payment if required.

8. Notes on other measures to improve fiscal sustainability
Increasing residency requirements for NZS
Work under the National-led Government who sought to raise the age of eligibility for NZS
to 67 years, starting from 2037, and in that same year increase residency requirements
(with grandparenting) to 20 years, showed that by 2041, $4 billion or 0.6% of GDP could
be saved (Office of the Minister of Finance, 2016). It was estimated that 120,000 fewer
people would be eligible for NZS in the year 2041. Of this number, only 6,800 were those
who needed to wait longer to meet the new 20 year residency requirements. It was
estimated that there would be an offsetting cost from raising residency of an additional
3,100 more Jobseeker and Supported Living Payments. Overall net savings in 2041 were
expected to be only $195 million.
Raising residency requirements also means that revenue would be foregone from the
overseas pensions direct deductions policy. Dale & St John (2016) argue that any increase
in residency should be in the context of reforming section 70 of the Social Security Act.
They proposed a residency increase of 25 years and abolition of the deduction of an
overseas state pension from a person’s NZS while retaining and updating certain reciprocal
agreements such as with the UK and Australia. The cost of reform of section 70, long
overdue, might simply offset the savings from increasing residency.
Even without this offset, increasing residency is not a serious contender for improving
fiscal sustainability and especially not in the short run.

In terms of intergenerational

perceptions of fairness however it may be a popular policy.

Compulsory saving and tax concessions
The Tax Working Group (2019) proposed some increases in tax incentives for KiwiSaver
in the context of the introduction of a CGT. That exercise was a reminder that such
incentives are difficult to design without them being complex, expensive and regressive,
and disadvantageous to women, while probably impacting little on overall savings

54

 (Retirement Policy and Research Centre, 2018).69 Arguments for and against tax
concessions are many and further explore in Chamberlain and Littlewood (2019) and
Inland Revenue & The Treasury (2019). However even if tax incentives increase KiwiSaver
saving there is no fiscal saving unless KiwiSaver reduces entitlement to NZS implying some
kind of means test.
Cullen (2012) outlined how an annuity purchased from KiwiSaver might be used to reduce
entitlement to NZS. Over time NZS could be made to disappear, shifting provision for
retirement from the state to the uncertainties of private provision. It is unlikely that such
an option would be acceptable in New Zealand.

Growing GDP, raising taxes, prefunding
While in theory, a growing GDP should enlarge the pie to be shared and make the provision
of universal NZS easier, the past three decades have shown that growth alone does not
solve mal-distribution of resources and increasing inequality and inequity. Rising GDP has
reduced the projection of cost of NZS as a % of GDP, however GDP per capita is a better
measure of success than GDP alone. With rising immigration, the growth of GDP per capita
has been low.
The fruits of per capita growth have not been shared well, nor does the future suggest
that growth will be of the things that enhance well-being. Much GDP will be devoted to
repairing the effects of poor regulation eg leaky buildings, climate change, environmental
damage and pollution, underfunding of education, a run-down health system, poor dental
health, as well as costs associated with obesity, dementia and diseases such as cancer
and diabetes.
Raising taxes may help pay for universal NZS, but if PAYE income tax and GST are
increased the impact will be severe on lower and middle income people. A wealth tax of
some kind has potential. Although it is a fairer way to go, a wealth tax will not solve the
issue of competing demands: universal NZS may be seen as a low priority in the face of
the demonstrable needs of many of working age and their children.
Increasing prefunding via the NZSF does not hold out much promise either under existing
arrangements. Increased contributions to the NZSF do not make NZS cheaper but do
preclude other more beneficial uses of the surpluses generated from taxation.

69

See https://cdn.auckland.ac.nz/assets/business/about/our-research/research-institutes-andcentres/RPRC/Submissions/Response%20to%20the%20Tax%20working%20group%20interim%20report%20s
avingfinal.pdf.

55

 9. Conclusion
This paper has outlined a number of possible policy tools that could be used to reduce the
cost to Government of retirement income support, and to enhance fiscal sustainability and
intergenerational perceptions of equity.
A range of evidence suggests that care is needed in designing policies that will reduce the
costs of NZS.

Projections under current policy settings imply however that action is

needed. Early action is always better than abrupt, dislocating action later.
If raising taxes is ruled out, and raising the age of eligibility is likely to be particularly
problematic in a rapidly changing jobs market, then there are only two other levers capable
of raising significant revenue: lowering the rate of NZS, and income-testing. If it is
desirable to save at least 10% of the net cost of NZS with little delay and without creating
poverty, then a basic income approach with some alignment of married, single and living
alone rates as suggested in this paper is worthy of consideration.
Reform also requires a holistic approach. Are reforms for long term care and private
annuities suggested elsewhere compatible with reforms to the structure of NZS? Are these
reforms compatible with a well-being approach and with overall principles that are fit for
the changed family structure and work environment of the 21 st century?
Policy development and implementation requires a long-term perspective not a three-year
election cycle which encourages a short-term bias where “vote-maximising politicians have
strong incentives to discount the future” (Boston, 2017, p. 167). A revisiting of the
possibility of a political accord backed by an ongoing retirement policy taskforce is required
to progress the reforms suggested in this paper.

References
Allen, J. (2019). The wellbeing and vulnerability of older New Zealand adults in retirement:
a background paper prepared for the Commission for Financial Capability’s 2019
Review of Retirement Income Policy. Auckland: Health and Ageing Research Team,
Massey University.
Barr, N. (2001). The welfare state as piggy bank. Information, risk uncertainty and the
role of the state. Oxford: Oxford University Press.
Barr, N., & Diamond, P. (2009). Reforming pensions: Principles, analytical errors and
policy corrections. International Social Security Review, 62(2), 5 - 29.
Bell, M. (2019a). Long-term Projections & Scenarios under an Ageing Population Paper
presented at the Retirement Policy Research Centre Retirement Income Policy
Summit , 26th April, Auckland University.
Bell, M. (2019b). The New Zealand Superannuation Fund – clarifying how it works and
explaining its outcomes. Paper presented at the 60th Annual Conference of the New
Zealand Association of Economists, 3 - 5 July 2019, Victoria University, Wellington.
https://storage.googleapis.com/wzukusers/user-

56

 30969499/documents/57c278a3401d49aca9d6b6603342d494/02_7_19_NZAE_C
onference_Programme_DRAFT.pdf
Boston, J. (2017). Governing for the Future: Designing Democratic Institutions for a Better
Tomorrow (Vol. 25). Bingley, UK: Emerald Group Publishing Ltd.
Cabinet Social Wellbeing Committee. (2019). Minute of Decision SWC-19-MIN-0050
Revised. Wellington: Retrieved from https://www.msd.govt.nz/documents/aboutmsd-and-our-work/publications-resources/information-releases/nzs-vpmodernisation-and-simplification/cabinet-minute-nzs-vp-modernisation-andsimplication-swc-19-min-0050.pdf.
Centrelink. (2019). Age-Pension: who can get it. Retrieved from Australia Government
Department
of
Human
Services
website:
https://www.humanservices.gov.au/individuals/services/centrelink/agepension/who-can-get-it
Chamberlain, M., & Littlewood, M. (2010). Towards a more rational tax treatment of
collective investment vehicles and their investors. Auckland: Retirement Policy and
Research Centre, Department of Economics, University of Auckland.
Chamberlain, M., & Littlewood, M. (2019). Informing the 2019 Review of retirement
income policies – 133 questions that New Zealand needs answered. Auckland.
Commission for Financial Capability. (2016). Ageing workforce, June, Final results digital
survey Review of Retirement Income Policies (pp. 29). Auckland: Commission for
Financial Capability.
Commission for Financial Literacy and Retirement Income. (2013). Focussing on the Future
Review of Retirement Income Policies. Wellington.
Cullen, M. (2012). The Political Economy of Long-Term Fiscal Planning from a Social
Democratic Perspective Affording Our Future conference. Retrieved from doi:
http://www.victoria.ac.nz/sacl/about/cpf/publications/pdfs/2.3-Cullen-paper.pdf.
Dale, M. C., & St John, S. (2016). New Zealand Superannuation policy and overseas state
pensions 2016 Review of Retirement Incomes Policy. Auckland: Retirement Policy
and Research Centre, prepared for the Commission for Financial Capability.
English, B. (2017). New Zealand Superannuation Cabinet paper.
Healey, O., & Curtin, J. (2019). ‘Relationship status’ and the Welfare System in Aotearoa
New Zealand A Report prepared for the Peter McKenzie Project May 2019 Auckland:
Public Policy Institute & Child Poverty Action Group.
Hon Kris Faafoi. (2019). Terms of reference for the 2019 retirement income policy review.
Wellington:
Retrieved
from
https://cffc-assets-prod.s3.ap-southeast2.amazonaws.com/public/Uploads/Retirement-Income-PolicyReview/84d0b92a1d/2019-review-of-retirement-income-policies.pdf.
Hyslop, D., & Townsend, W. (2017). The Longer Term Impacts of Job Displacement on
Labour Market Outcomes. Motu Economic and Public Policy Research, Wellington,
Working paper 17-12.
Inland Revenue & The Treasury. (2019). Tax Policy Report: Joint Report: KiwiSaver and
the Taxation of Retirement Savings. Wellington: Inland Revenue & The Treasury
Retrieved
from
http://taxpolicy.ird.govt.nz/sites/default/files/2019-ir-taxworking-group-15-report-ir2019-081-t2019-297.pdf.
Kerse, N. (2019). How the realities of an ageing population inform the projections. Paper
presented at the RPRC PPI Summit: 2019 Retirement Income Policy Review and
You,
University
of
Auckland.
https://cdn.auckland.ac.nz/assets/business/about/our-research/researchinstitutes-and-

57

 centres/RPRC/OtherPapers/Session%202%20Ageing%20population%20fiscal%20
and%20health%20implications%20.pdf
Kirkness, L. (2019). NZ's super-rich and the tax man: 350 people. Each worth $50m plus.
And
a
fight
over
$85.8m,
NZHerald.
Retrieved
from
https://www.nzherald.co.nz/business/news/article.cfm?c_id=3&objectid=122526
97&utm_source=The+Bulletin&utm_campaign=3be6db732cEMAIL_CAMPAIGN_2018_03_01_COPY_01&utm_medium=email&utm_term=0_5
52336e15a-3be6db732c-533836733
Littlewood, M. (2014). Updating data on older workers PensionBriefing (Vol. 2014-4, pp.
6). University of Auckland: Retirement Policy and Research Centre.
McKenzie, A. (2019). Older New Zealanders: Some Emerging Trends. Paper presented at
the RPRC Summit: The 2019 Retirement Income Policy Review & You, University
of
Auckland.
https://cdn.auckland.ac.nz/assets/business/about/ourresearch/research-institutes-andcentres/RPRC/OtherPapers/Session%204%20Equity%20and%20distribution.pdf
McTaggart, S. (2005). Monitoring the Impact of Social Policy, 1980–2001: Report on
Significant Policy Events Occasional Paper Series, Resource Report 1. Wellington:
Social Policy Evaluation and Research Committee (SPEAR).
Medford, A., Vaupel, J. W., & Christensen, K. (2019). Why has human life expectancy
more than doubled over the past two centuries?, Independent. Retrieved from
https://www.independent.co.uk/news/health/human-life-expectancy-increaseageing-healthcare-sweden-denmark-medicine-a8944151.html
Ministry of Social Development. (2016). Description of New Zealand’s current retirement
income policies. Background paper prepared by the Ministry of Social Development
for the Retirement Commissioner’s 2016 Review of Retirement Income Policy.
Wellington: Ministry of Social Development.
Ministry of Social Development. (2019). Description of New Zealand’s current retirement
income policies. Background paper prepared by the Ministry of Social Development
for the Retirement Commissioner’s 2019 Review of Retirement Income Policy.
Wellington: Ministry of Social Development.
Myer, R. (2017, 5 June). ‘Work till you’re 70’: the Abbott-era measure the Turnbull
government
is
now
backing,
The
NewDaily.
Retrieved
from
https://thenewdaily.com.au/money/retirement/2017/06/05/pension-age-70/
New Zealand Productivity Commission. (2019). Technological change and the future of
work Issues Paper April 2019. Wellington: New Zealand Productivity Commission.
O'Connell, A., Edgar, C., Ormrod, C., Mussett, D., Shirley, J., Benbow, J., . . . Channon,
M. (2015). Income Streaming in Retirement: Options for New Zealand Retirement
Income Interest Group. Wellington: New Zealand Society of Actuaries.
OECD. (2017a). Pensions at a Glance 2017.
OECD. (2017b). Preventing Ageing Unequally. Paris: OECD Publishing.
Office for Seniors. (2018). An Ageing Population Discussion Document He Pukapuka
Matapaki.
Wellington:
New
Zealand
Government
Retrieved
from
http://www.superseniors.msd.govt.nz/documents/ageing-populationconsultation/ageing-population-discussion-document-web.pdf.
Office of the Minister of Finance. (2016). New Zealand Superannuation: Proposal to
Cabinet. Wellington.
Periodic Report Group. (1997). 1997 Retirement Income Report: A review of the current
framework. Interim report. Wellington.

58

 Perry, B. (2018). The material wellbeing of NZ households:: Overview and Key Findings.
Wellington: Ministry of Social Development.
Perry, B. (2019 forthcoming). Household Incomes in New Zealand: trends in indicators of
inequality and hardship 1982 to 2018. Wellington: Ministry of Social Development.
Personal Finance. (2018, 25 February). A third of women have less than $5000 in
KiwiSaver
accounts.
Westpac
Survey,
NZHerald.
Retrieved
from
https://www.nzherald.co.nz/personalfinance/news/article.cfm?c_id=12&objectid=12001571
Pledger, M., McDonald, J., Dunn, P., Cumming, J., & Saville-Smith, K. (2019). The health
of older New Zealanders in relation to housing tenure: analysis of pooled data from
three consecutive, annual New Zealand Health Surveys. Australian and New
Zealand Journal of Public Health, 43(2), 182 - 189.
Prendergast, S., Blackmore, D., Kempson, E., Russell, R., & Kutin, J. (2018). Financial
Wellbeing: a survey of adults in New Zealand, April 2018. Auckland: ANZ Banking
Group Ltd.
Preston, D. (2001). Retirement income in New Zealand: the historical context, from
http://www.retirement.org.nz
Retirement Commission. (2010). 2010 Review of Retirement Income Policy. Wellington:
Retirement Commission.
Retirement Commission. (2013). Focusing on the future: Report to Government 2013
Review of Retirement Incomes Policies. Wellington: Commission for Financial
Literacy and Retirement Incomes.
Retirement Commissioner. (2016). Review retirement income policies. Retrieved from
doi:https://www.cffc.org.nz/assets/Uploads/2016-Review-of-Retirement-IncomePolicies-Tabled-Report-19.12.06.pdf
St John, S. (1991). Reform of the GRI surcharge. Wellington: New Zealand Planning
Council.
St John, S. (1999). Superannuation in the 1990s: where angels fear to tread? In P. Dalziel,
J. Boston & S. St John (Eds.), Redesigning the welfare state in New Zealand.
Auckland: Oxford University Press.
St John, S. (2015). Improving the affordability of New Zealand Superannuation Working
Paper 2015-1. Auckland: Retirement Policy and Research Centre, the University of
Auckland.
St John, S. (2016). KiwiSpend: How to spend like a Kiwi. PensionCommentary, Retirement
Policy and Research Centre 2016-2.
St John, S. (2018a). Feedback on the ‘strawman’ proposal for Irish pension reform from
New Zealand experience: Report for Insurance Ireland and TOR Financial
Consulting: Retirement Policy and Research Centre.
St John, S. (2018b). Improving the affordability of New Zealand Superannuation Working
Paper 2018-1. Auckland: Retirement Policy and Research Centre, the University of
Auckland.
St John, S. (2018c). Improving the affordability of New Zealand Superannuation: New
Zealand Superannuation as a Universal Basic income (Vol. Policy Discussion Paper
2018-1.): Retirement Policy and Research Centre.
St John, S., & Ashton, T. (1993). Private pensions in New Zealand: Can they avert the
crisis? Wellington: Institute of Policy Studies.

59

 St John, S., & Dale, M. C. (2019). Decumulation: the time to act is now 2016 Review of
Retirement Incomes Policy. Auckland: Retirement Policy and Research Centre,
prepared for the Commission for Financial Capability.
St John, S., MacLennan, C., Anderson, H., & Fountain, R. (2014). The complexities of
relationship in the welfare system and the consequences for children. Auckland:
Child Poverty Action Group.
StatsNZ. (2019). Life expectancy at birth, total population, by sex, 1950–52 to 2012–14.
Retrieved 2 August 2019, from StatsNZ https://www.stats.govt.nz/topics/lifeexpectancy?gclid=EAIaIQobChMIq4710tni4wIVw4RwCh366gTiEAAYAiAAEgLtnfD_
BwE
Szabó, Á., Allen, J., Stephens, C., & Alpass, F. (2019). Is retirement associated with
physical health benefits? A longitudinal investigation with older New Zealanders.
Age and Ageing, 48(2), 267-272.
The Treasury. (2013). Affording our future; Statement on New Zealand's long-term fiscal
position. July 2013 Wellington: The New Zealand Treasury.
The Treasury. (2016). He Tirohanga Mokopuna: 2016 Statement on the Long-Term Fiscal
Position Wellington: The Treasury, New Zealand Government.
The Treasury. (2019). Budget Economic and Fiscal Update. (30 May 2019). Retrieved from
https://treasury.govt.nz/sites/default/files/2019-06/befu19.pdf
doi:https://treasury.govt.nz/sites/default/files/2019-06/befu19.pdf
Walters, L. (2017). Jacinda Ardern pledges to quit rather than raise NZ Super age.
Stuff.co.nz,
(5
September).
Retrieved
from
https://www.stuff.co.nz/national/politics/96496437/jacinda-ardern-pledges-toquit-rather-than-raise-nz-super-age
Welfare Expert Advisory Group. (2019). Whakamana Tāngata – Restoring Dignity to Social
Security in New Zealand (pp. 216). Wellington: Welfare Expert Advisory Group.
World Economic Forum. (2013). Healthcare Industry 2013. Sustainable Health Systems:
Visions, Strategies, Critical Uncertainties and Scenarios. A report from the World
Economic Forum in collaboration with McKinsey & Company. Retrieved from
http://www3.weforum.org/docs/WEF_SustainableHealthSystems_Report_2013.pd
f
Wright, S. (2019, 10th August). Over 65s flooding the job market - and finding they're not
so employable, Sydney Morning Herald, 10th August.

60

 Appendix NXSG costings for 2017/2018 March year
Status quo in tax year 2017/18 (1 April 2017 to 31 March 2018)

1.
All
net 2. All singles: net 3. All current net 4. All net single
married rate
single sharing rate rates
sharing rate
(billion)

2017/18 aggregate net of tax cost of New Zealand Superannuation (NZS)

11.846

2017/18 Tax paid by superannuitants on all of their taxable income

5.244

2017/18 Tax paid by superannuitants on non-NZS income

3.555

Scenario 1: Flat tax of 39% on non-NZS taxable income
Aggregate net of tax cost of New Zealand Superannuation

10.454

11.225

11.479

Aggregate tax paid on non-NZS taxable income

4.844

4.882

4.888

5.102

Saving on aggregate net NZS payments

1.392

0.621

0.367

-0.985

Saving on aggregate net NZS as % of status quo aggregate net NZS
Extra tax paid on non-NZS income
Increase in tax revenue as % of status quo tax on non-NZS income
Overall saving relative to status quo
Overall saving as % of status quo aggregate net NZS
Aggregate saving on NZS just from paying net married person rate
Percentage of overall NZS saving due to paying the net rate option
Percentage of superannuitants who no longer receive NZS relative to status quo recipient numbers

12.831

12%

5%

3%

-8%

1.289

1.327

1.333

1.547

36%

37%

37%

44%

2.681

1.948

1.700

0.562

23%

16%

14%

5%

0.988

0.250

0.000

-1.186

37%

12.8%

0%

n/a

3.7%

3.4%

3.3%

1.5%

12.899

Scenario 2: 17.5% tax on non-NZS taxable income up to $15,000 and then 39% beyond that
Aggregate net of tax cost of New Zealand Superannuation

10.695

11.442

11.692

Aggregate tax paid on non-NZS taxable income

4.267

4.279

4.279

4.326

Saving on aggregate net NZS payments

1.151

0.404

0.154

-1.053

Saving on aggregate net NZS as % of status quo aggregate net NZS
Extra tax paid on non-NZS income
Increase in tax revenue as % of status quo tax on non-NZS income

10%

3%

1%

-9%

0.712

0.724

0.724

0.771

20%

20%

20%

22%

1.863

1.128

0.878

-0.282

Overall saving as % of status quo aggregate net NZS

16%

10%

7%

-2%

Percentage of overall NZS saving due to paying the net rate option

53%

22%

0%

n/a

1.5%

1.4%

1.4%

1.0%

12.598

Overall saving relative to status quo

Percentage of superannuitants who no longer receive NZS relative to status quo recipient numbers

Scenario 3: 20% tax on non-NZS taxable income up to $20,000 and then 45% beyond that
Aggregate net of tax cost of New Zealand Superannuation

10.356

11.150

11.402

Aggregate tax paid on non-NZS taxable income

4.123

4.190

4.195

4.275

Saving on aggregate net NZS payments

1.490

0.696

0.444

-0.752

Saving on aggregate net NZS as % of status quo aggregate net NZS
Extra tax paid on non-NZS income
Increase in tax revenue as % of status quo tax on non-NZS income

13%

6%

4%

-6%

0.568

0.635

0.640

0.720

16%

18%

18%

20%

2.058

1.331

1.084

-0.032

Overall saving as % of status quo aggregate net NZS

17%

11%

9%

0%

Percentage of overall NZS saving due to paying the net rate option

48%

19%

0%

n/a

4.6%

4.0%

4.0%

3.3%

Overall saving relative to status quo

Percentage of superannuitants who no longer receive NZS relative to status quo recipient numbers

61