KiwiSaver: A Model Scheme

Alison O’Connell¹
Institute of Policy Studies
Victoria University of Wellington

Abstract

KiwiSaver is the world's first national auto-enrolment savings scheme. So far only one other country - the United Kingdom - has committed to auto-enrolment on a national scale. Both schemes aim to increase the number of people saving for retirement, and they share many design features. However, there are significant differences in the way the schemes are delivered, with implications for the levels of choice, risk and cost for savers and government. The United Kingdom's personal accounts scheme, sizeable though it is, is only part of a complicated private savings landscape, whereas KiwiSaver is fast becoming the predominant vehicle for retirement saving in New Zealand. It offers a working model for countries seeking to create a simple and unified national scheme for lifetime saving.

Introduction

KiwiSaver has transformed savings in New Zealand. It started on 1 July 2007, and at the end of June 2009 had over 1.1 million members (Inland Revenue 2009b). Membership is expected to plateau in 2012 at 1.4 million (Inland Revenue 2009a). 

KiwiSaver contains several innovative features, the main one being auto-enrolment, sometimes called “soft compulsion”. Workers are enrolled automatically into saving and can choose to opt out if they wish, but if they stay in the scheme the employer is compelled to contribute.

Auto-enrolment captures some learning from behavioural economics (see, for example, Madrian and Shea 2001): people have high inertia in relation to saving so are more likely to stay in a savings scheme into which they have been enrolled automatically than make the effort to join it themselves. Although auto-enrolment is used in some employer-based savings schemes in the United States, KiwiSaver is the world’s first auto-enrolment scheme on a national scale. The United Kingdom (UK) has also committed to national auto-enrolment with compulsory employer contributions, and to a new scheme called personal accounts (PAs). The first UK proposals were made a year after KiwiSaver began to be discussed (Pensions Commission 2005), and the start date is planned to be in 2012.

This paper compares these two schemes. It begins with a discussion of the important points of policy context in both countries. It then explains the differences between KiwiSaver and auto-enrolment plans in the UK, in terms of the areas of the markets the schemes cover, and the delivery model used or planned. Next it discusses the policy choices made in each case on what have emerged as the important dimensions: level of investment and provider choice for the saver and levels of risk and cost for both government and saver. The two schemes have taken different positions on these aspects. This paper argues that KiwiSaver encourages more choice and is less risky, although the jury is out on whether it is higher cost. Finally, the paper considers whether KiwiSaver would be a good model for other countries.

Table 1 briefly compares some features of auto-enrolment in the two countries. This paper focuses on the different delivery models, and so there are many points of detail in both KiwiSaver and the UK schemes that are inevitably not covered here. Readers who wish to know more are referred to www.kiwisaver.govt.nz and the relevant UK government publications (Department for Work and Pensions 2006, 2007).

Table 1  Selected Features of Auto-enrolment Retirement Savings in New Zealand and the UK (as at July 2009, current plans for the UK)

*  Employer superannuation contribution tax (formerly known as specified superannuation contribution withholding tax, or SSCWT) up to a maximum of 2% of the employee’s gross pay.
** Depending on fund structure and member's income level.

A word on terminology: KiwiSaver is a defined contribution retirement savings product.² Although money saved can be taken out before age 65 in specific circumstances, and can be taken out at age 65 by someone in work or not, the money saved is intended to be used primarily in retirement. This is the same as for PAs, although UK tax rules make saving for retirement liable to compulsory annuitisation. It is therefore natural to call PAs a pension scheme. This is one of many environmental differences between KiwiSaver and PAs, which should not distract from the overall comparison between the two.

Retirement income policy context: UK and New Zealand

The only two countries in the world – so far – to plan an auto-enrolment savings scheme both did so to extend the number of people saving for retirement. However, the policy developments were aimed at different perceptions of the problem (O’Connell 2006). In both countries active membership of workplace pension or superannuation schemes has been falling. But in 2006, before KiwiSaver, the proportion of working-age New Zealanders making some kind of retirement saving above the public pension³ was around half the 40% achieved in the UK.⁴ The rationale for KiwiSaver in New Zealand was part micro-economic – getting households into the habit of saving was good for the individuals concerned – and part macro-economic − national savings would improve and local capital markets would develop.

The UK already has huge and well-developed capital markets, and any possible macro-economic benefit of a new savings product has not been part of the debate. The UK problem was seen to be that while some people are saving “enough”, large numbers are saving nothing or not enough to make up for the poor public pensions. This was characterised as a problem of access to good savings schemes through the workplace, and so the new personal accounts scheme (PAs) is targeted at those workers whose employers do not already offer a pension scheme.

Following the first proposals by the Pensions Commission in 2005, the Pensions Act 2008 outlined the way in which auto-enrolment will work and formally gave the Personal Accounts Delivery Authority (PADA) executive powers to deliver PAs. So while the outline of the plan has been widely discussed, many details are still being worked on before launch in 2012.

We can already see the extraordinary progress of KiwiSaver. The Inland Revenue’s Evaluation Services report (2008a) on KiwiSaver’s first full year⁵ showed that membership had reached 22% of the eligible population. This size of membership was previously forecast to be reached in 2011. By the end of June 2009 membership had grown further to reach around one-third of the eligible population (Inland Revenue 2009b).

A survey of employers with pre-existing workplace schemes found that most members stayed in those schemes on the introduction of KiwiSaver (Inland Revenue 2008a:34). This means that KiwiSaver members include many people who have not previously saved for their retirement in designated schemes. The total proportion of people saving for their retirement in New Zealand is therefore higher than the one-third of those eligible for KiwiSaver. This is high by international benchmarks. Consider the UK: membership of defined benefit schemes has been reducing fast, until they are now mainly limited to the public sector and very large private companies. The highly incentivised defined contribution market is well developed since major reforms two decades ago, but KiwiSaver has taken less than a year to exceed the proportion of working-age people currently in defined contribution pensions in the UK.⁶

Why has KiwiSaver proved so popular? Around one-third of members surveyed after six months of the scheme starting said that incentives and design features played a part in choosing to join KiwiSaver (Inland Revenue 2008b:21). At that time, the incentives for members were more generous than those shown in Table 1, with an additional fee subsidy of $20 every six months (as well as other differences in minimum level of contribution and employer incentives). KiwiSaver incentives, generous compared to the previous lack of incentives for saving in New Zealand, were probably worth less for each saver than UK pension incentives at the time.⁷ However, given the complexities of the UK’s tax relief system, KiwiSaver’s matching contributions may seem more tangible and easier to understand (and are less regressive).

Incentives do not appear to be the whole story, though. For over 60% of members in the six-month survey, “saving for retirement” was the main reason for joining (Inland Revenue 2008b:21). In other words, the availability of a simple retirement savings product, with lock-in to age 65, has proved attractive in its own right. Auto-enrolment is obviously proving effective at starting people saving, and generic KiwiSaver brand and provider-specific marketing must be helping. KiwiSaver provides a lesson in how attractive a simple and easily accessible private retirement savings product can be. That lesson has eluded policy makers in many countries, including the UK, over many years.

It does not appear to be the case that KiwiSaver is attractive only to a narrow group. KiwiSaver members are of all ages, with young workers aged 18–24 and the over 55s particularly well represented. Although workers opting in have higher incomes than average eligible salary and wage earners, auto-enrolled workers have lower than average incomes. Women make up 51% of members (Inland Revenue 2009a).

KiwiSaver as a "whole market" solution

PAs form one part of the private pension landscape in the UK, but KiwiSaver is a “whole market” solution, as defined by three important features.

First, any permanent resident of New Zealand under 65 can join KiwiSaver at any time, employed or not. As at the end of June 2009, around 61% of KiwiSaver members have opted in. The number of opt-ins seems to have been highest at the start of the scheme, with a smaller peak when employer contributions became compulsory. Around 39% of KiwiSaver members have joined through auto-enrolment. Of those who have been automatically enrolled, 34% opted out (Inland Revenue 2009b). People are continuously being automatically enrolled into KiwiSaver as that occurs on job change. Steady growth in members is expected to continue for some time.

Second, any amount can be contributed at any time. Employees choose a 2%, 4% or 8% of salary regular contribution, but can put more in lump sums whenever they choose. Amounts in existing complying pension schemes can be transferred in.

Third, KiwiSaver is becoming the predominant retirement savings product for New Zealand. Some existing workplace schemes are still open to employees of specific organisations, but the full incentives are only available through KiwiSaver. KiwiSaver is also an important national brand. Because it is a government-sponsored product, available to everyone, the brand has been developed and paid for by government.

This is in marked contrast to the UK situation, where employers will be responsible for putting in place a scheme into which their employees will be enrolled. Many employers already have some kind of workplace pension scheme, but an employer can choose to sign up with the PA scheme. So in the UK there will not be a single brand, a “BritSaver” available to all, but instead a continuing range of variously named pension products with access depending on employer arrangements.

Creating a new scheme in a market with a lot of existing provision has caused tensions. The existing providers – pension funds and insurers – have lobbied hard for PAs to be “ring-fenced” for the target market, so that PAs do not eat into existing business. Partly as a result, access to PAs is restricted, contributions are capped (at £3,600, or around $10,000 a year), and transfers from existing pension savings to PAs are not allowed. Only the self-employed will be able to opt in to PAs. A worker for an employer outside the PA scheme will have to join his or her employer’s scheme and/or take out an individual pension product, but cannot join the PA scheme even if he or she would prefer to  However, outside the existing employer relationship, a worker may use his or her PA set up by a previous employer.

These distortions keep complexity in the system and will mean people end their career with lots of different pots from different jobs. KiwiSaver starts in a simpler market, with fewer and smaller existing providers, but it should be an advantage that, from the individual saver’s point of view, all retirement saving over a lifetime can be collected in one KiwiSaver pot. KiwiSaver can be the organising account for an individual’s lifetime savings.

KiwiSaver builds on existing provision and providers

Given the size and sophistication of the pre-existing UK market compared to that in New Zealand, it seems counterintuitive that the New Zealand government chose to build on existing provision and providers to create KiwiSaver, whereas the UK government is pioneering a new model for delivering PAs. Yet this is indeed the case.

The KiwiSaver Act 2006 sets out what the KiwiSaver product should look like and the conditions that have to be met to provide a KiwiSaver product. Any provider can apply to become a KiwiSaver provider, through a registration process handled by the Government Actuary. KiwiSaver products are run by 31 different providers, most of which are existing financial services companies such as banks, insurance companies and investment managers. Six providers have been chosen through a tender process to provide KiwiSaver default funds, to which auto-enrolled members are allocated with equal probability.

Existing workplace retirement savings provision has continued since the introduction of KiwiSaver. Such schemes can apply to convert to a KiwiSaver scheme. Employees are only auto-enrolled into a KiwiSaver scheme, so if an existing scheme does not convert it has to run alongside access to KiwiSaver. The exception is for some very large employers, who offer such good schemes they have been granted an exemption from auto-enrolment. So far, survey evidence indicates that while some members of existing schemes have joined KiwiSaver as well, none have transferred to KiwiSaver and most have carried on in their existing scheme (Inland Revenue 2008a).

In contrast, the UK scheme of PAs has been designed to take an entirely new approach. The Personal Accounts Delivery Authority (PADA) is a new non-departmental public body accountable to Parliament and reporting, through a board, to the appropriate Secretary of State. The new body will be responsible for oversight of the new systems – contribution collection, account administration and fund management – but will outsource these functions to the private sector.

PAs will offer a limited choice of funds for savers and a default fund for those who do not choose a fund. The expectation is that the vast majority of members will not move out of the default fund, so that in effect the PA scheme will be one large fund, invested at arm’s length by the investment managers chosen by PADA. A new system for administering membership and allocation of contributions to investment will be built and operated by external suppliers, again chosen by PADA.⁸

So instead of using traditional savings product providers like banks and insurance companies to run individual products (as in KiwiSaver), PADA will contract wholesale investment managers and administration companies to run, essentially, one large fund. Competition will be through contracts set up by PADA rather than by individual consumers making their own choice of product. It is worth noting at this point that although the PA scheme is just part of the UK pension market, and targeted at new savers, it could have up to 7 million active members (PADA 2008:13). So the PA scheme alone is several times the size of KiwiSaver.

The policy process that decided on this “one large fund” model set up a straight contest between two alternatives: this centralised model and a so-called “industry” model, which was very similar to KiwiSaver. The reasons given for choosing the winning model were that it would have the lowest cost and simplicity for the target members, and that private market suppliers would have the skills, expertise and capacity to deliver (Department for Work and Pensions 2007:25).

Different policy decisions

Choice

As a result of the different policy decisions, KiwiSaver and auto-enrolment in New Zealand are very different from PAs and auto-enrolment in the UK in three key ways. First: choice. The policy approach to choice in KiwiSaver can be illustrated by the following quote from Inland Revenue (2008a:27): “Choices in schemes, providers and contribution rates provide [KiwiSaver] members with opportunities to exercise control over their investment choices and to choose a savings scheme that is appropriate for their circumstances.”

Savers can choose which KiwiSaver provider they want to administer their product and which investment fund they want to be in. An auto-enrolled member does not have to make a choice as he or she will be allocated to one of the six default funds. Alternatively, employers can nominate a KiwiSaver provider that acts as a default for its employees (opt-in or auto-enrolled).

UK policy for PAs is different, based on the view that PA members will not want to choose and do not have the financial literacy to choose well, and that too much choice can cause decision paralysis (Department for Work and Pensions 2007:102–103). A PA member will have no choice on which company administers his or her account, because administration is contracted out. The PA scheme will focus on one default fund for efficiency and offer few other funds. As in KiwiSaver, a PA member does not have to stay in the default fund, but all default PA members are in one fund rather than being distributed across six providers.

Since KiwiSaver started, 35% of KiwiSavers entered via one or other of the six default funds, 13% via an employer-nominated scheme and 53% via one of the non-default active choice schemes (Inland Revenue 2009a:34). This understates the proportion of members who have made an active choice, as some of the members in the default fund chose that rather than being allocated to it, and some members would have made a choice subsequent to entry. It seems, therefore, that KiwiSavers are exercising more choice than in the assumption underlying the UK’s centralised model.

But the “whole market” of KiwiSaver would be expected to exercise more choice than the restricted and largely auto-enrolled target market for PAs. The UK view that most people will not switch out of the default fund seems to be supported – so far – by the 90% of KiwiSaver auto-enrollees who entered via the default or employer-nominated scheme (Inland Revenue 2009a:35). However, it is not known how many of these considered moving out but actively chose to stay in the default fund or how many will choose to move out of the default scheme in future. Also, some KiwiSaver members making active choices have chosen a default fund. This will not be a choice open to British savers.

The KiwiSaver market seems to have avoided the threat of paralysis due to too much choice. KiwiSaver providers do not offer many funds each. For example, it is typical for a major bank to offer five funds only, described by risk type. The Sorted website⁹ helps with a “risk recommender” to help people find their preferred risk category out of a possible six. Information on fees for each fund is available by category. The smallest category has just over 20 funds, the largest has under 40. This may be a consequence of a small and less developed market, but an impression of the information disclosed on KiwiSaver funds is that it is easier to understand than the more complicated disclosure information required for UK pensions.

Cost

The second way that PA policy decisions are very different from KiwiSaver is in cost. A cost target has not been explicit in the development of KiwiSaver. Reasonable fees are required to pass the registration test for a KiwiSaver provider, with a stronger test for providers of the six default funds. So what providers charge is up to them, as is how they spend the money they collect through fees (on marketing, or advisers for example). Competition between providers would be expected, and Inland Revenue (2008a:54) reports from a survey of KiwiSaver providers that they have had to adapt to a new “low-cost, low-touch delivery model”.

A very strong operating cost target has been the driving force behind the policy decision for a centralised model for PAs. A key assumption of the Pensions Commission was that pre-existing UK pension provision had not extended to the target market for PAs (generally people on low to medium incomes), because it was not profitable to service that customer segment except at such high fees that returns were unacceptably reduced. The Pensions Commission suggested the centralised model so that operating costs could be kept low by economies of scale in investment and administration, and by making competition work through centralised contracts (rather than through individual products), which should reduce marketing expenditure and account switching.

The Commission’s first estimate of a target annual management charge of 0.3% of funds under management for PAs was later amended to “possibly as low as 0.5% in the short term and below 0.3% in the long term” (Department of Work and Pensions 2006:94). There has been no further information from PADA on latest best estimates for 2012 and beyond, so whether this target can be met remains uncertain.

These target costs cannot be compared directly with those of KiwiSaver, largely because KiwiSaver is available to every New Zealander, so government (that is, the taxpayer) has financed part of the operations. For example, government has paid to advertise the brand and for generic KiwiSaver information and education. It has also set up a centralised clearing house for contribution collection within Inland Revenue.¹⁰

In the UK, auto-enrolment also affects every worker, but there will be no clearing house. Instead, individual employers have to manage contribution collection.¹¹ Compliance for contribution collection will be undertaken by the Pensions Regulator, a public body currently funded from taxation. Further, the PA scheme is ring-fenced for a targeted segment only. There can be no government subsidy to only part of the saving population, so future PA members have to finance the whole cost of PAs, including development costs.

Because the PA scheme is essentially a default fund, it is appropriate to compare it with default KiwiSaver funds. KiwiSaver fees are not reported in a standard way, but analysis of the instruments of appointments of these funds shows a range from around less than 0.4% to around 0.65% of funds under management. This early evidence therefore suggests that the KiwiSaver model transplanted to the UK could result in fees to the saver at a level within the target range for PAs, especially given the greater scale economies in the UK and that competition over time may reduce KiwiSaver fees.

Given that the aim of reducing cost proved critical for the decision on a centralised versus a provider model in the UK, it is worth questioning why costs mattered so much for PAs but have been relatively overlooked in KiwiSaver commentary. It could be that the New Zealand market is not as cost aware as the UK market. For example, there has been mandatory disclosure of fee levels using standard measures for long-term savings products in the UK for many years, whereas providers participate in fee comparisons for KiwiSaver voluntarily. But the KiwiSaver experience provides no evidence that the fee levels are off-putting to potential savers.

Risk

The third axis on which the UK and New Zealand models differ is in risk. The system risk in KiwiSaver has been managed. Providers were able to adapt or build what product systems they needed in time for the KiwiSaver launch. The clearing house for contributions was built within Inland Revenue, partly using existing systems. KiwiSaver systems – provider and Inland Revenue – are generally accepted to be working well, with some glitches mainly dealt with in the first year (Inland Revenue 2008a). The systems were set up very quickly: it took less than three years from the first report suggesting something like KiwiSaver (SPWG, 2004) to receiving the first contributions.

If the same pace of development had been achieved in the UK, the UK’s first auto-enrolled contributions would already have been made, instead of waiting until 2012. Partly this is due to longer decision-making and legislative processes. However, there is also a long period of supplier procurement and new system build in the PA model that was not needed in the KiwiSaver provider model. The PA “one large fund” model concentrates the system risk in one new build. Building from scratch is itself more risky than the KiwiSaver model, which built off existing systems and gave the challenge of getting products ready to providers.

The risk of any system failure or dissatisfaction could, in any national scheme, rebound on government. This is true in the UK, despite PADA being a public body organisationally separate from government: the perception is still that the PA scheme is a government initiative. The KiwiSaver provider model is also a national government initiative, and some risk of system failure or dissatisfaction exists for Inland Revenue, although some risk is also spread among a number of private providers.

The risk to government from a national auto-enrolment savings scheme is not just the risk from systems. There is also the risk of saver dissatisfaction with investment or product performance. In the PA “one large fund” model, every default fund member will be exposed to the same investment strategy as chosen by PADA. The investment performance of that fund will therefore matter to millions of people, and it will be under intense scrutiny. Investment strategy has yet to be developed, but it has been assumed that the best investment strategy for this fund will be to diversify across asset classes, and to “lifestyle” so that a member’s exposure to equities reduces in favour of bonds as he or she approaches retirement age (Department of Work and Pensions 2007:106).

KiwiSaver default funds are invested mainly in fixed interest, with at least 15% but less than 25% in growth stocks. This has been criticised, from a UK perspective, as being too conservative (Pensions Policy Institute 2008b). Yet, especially after 2008 financial markets events, is it possible to say which is the best investment strategy for so many people? And with millions of people investing in one fund, could the UK government afford not to make some compensation if some regulatory failure occurred in the fund?¹²

The KiwiSaver default model is not without risks. A problem with one of the six default providers could challenge the tendering decision, but the risk to savers is spread by having six default funds rather than one. The Retirement Commissioner (2007) raised the concern that the different KiwiSaver funds will give different returns to savers who have been allocated to them by chance, because the funds invest differently and have different costs. However, the government view is that the potential for different outcomes from the different default providers was accepted as part of the KiwiSaver design and that members could switch out of default funds without additional cost (Office of the Minister for Social Development and Employment 2008). This illustrates the underlying preference in KiwiSaver policy that individuals should make their own choices on what is appropriate for them.

Both the New Zealand and the UK models allow everyone to make some choices, although access to schemes is more restricted in the UK than in New Zealand. Both models also enable people not to make a choice, by having defaults. But the underlying preference in the UK model is different. The UK model recognises that people mostly will not make choices, and expects to make significant choices for them.

So a key question is this: even accepting that not many people will want to make a choice on fund or provider, is it then appropriate to make the same choice on behalf of all non-choosers? The UK model asserts that this is so because of the perceived cost advantage of the “one large fund” model, and the opportunity for superior investment returns by operating competitive contracts for investment managers.

Three arguments can be made for the contrary view. First, making decisions for savers shields them from the realities of the investment risks to which they are nevertheless exposed. It gives up on trying to improve financial understanding. Instead, KiwiSaver strategy has been, since an early Cabinet paper (Office of the Minister of Finance 2005:5) “to support a cultural change in New Zealanders’ attitudes toward savings to increase self reliance and forward planning”. The need for initiatives to increase financial literacy was recognised at that stage. Further ways to help people make appropriate choices (not just at auto-enrolment but throughout life as circumstances change) could include mandatory disclosure and standardised comparison of investment strategies, returns and costs.

Second, the supposed cost advantage from the UK model compared to the KiwiSaver model is not yet proven. Early evidence suggests that the KiwiSaver model can also operate within or close to the target fee range for PAs. Further, it is possible that any cost saving between the models is negated by poor investment performance, or some other detriment to the saver such as poor administration, delay in changing investment contracts or dissatisfaction with the level of ethical investment in the fund. Even if cost to the saver were higher in one model, the overall benefit to the saver from the package of level of choice, risk and cost may still be higher.

Third, the risk to the taxpayer also needs to be considered. The more decisions are made on behalf of savers, instead of being clearly each saver’s own responsibility, the higher the risk that government has to step in if anything were to go wrong. This is a contingent cost for the taxpayer.

Conclusion

Is KiwiSaver a good model for other countries?

Any country looking to introduce a national auto-enrolment retirement savings scheme can look to KiwiSaver or to the UK plans.

The success of any similar scheme in another country, and the appropriate scheme design, will depend on the particular context. Countries will have different issues with the extent and type of existing provision, savings incentives, levels of financial literacy and the interface with the public pension. For example, in the UK, the success of any policy to encourage retirement saving is threatened by the prevalence of asset- and income-tested benefits for people over pension age. On this point New Zealand cannot offer any experience, because the prevalence of such benefits is much lower, largely because NZ Superannuation is not means tested.

KiwiSaver will have five years' experience before the UK scheme starts. The early progress of KiwiSaver appears to confirm that auto-enrolment helps to bring in new savers, and that a simple, nationally branded and incentivised product can be attractive. The delivery method has worked. The economic and distributional impacts of KiwiSaver are not the subjects of this paper. It is too early to tell, but membership trends look promising. Some issues and uncertainties remain.  The decisions on KiwiSaver incentives made after the 2008 general election show that changes can always be made.

Two major differences have been identified in this paper as rationale for the different delivery models chosen for the national auto-enrolment savings schemes in the UK and New Zealand. One is that KiwiSaver is a "whole market" solution, which has sought to involve existing providers in delivering a nationally branded product that can be an individual's single account for lifetime retirement saving (as well as saving for a first home, a feature which is not discussed in this paper). The UK model allows existing providers to participate in auto-enrolment through a continuation of existing varied products, but effectively prevents them from participating in a new ring-fenced scheme for a particular market segment. Therefore, KiwiSaver is a model for countries seeking to create a simple and unified national scheme for lifetime saving. The UK scheme is a model for countries wishing to include a further option in an existing market - but the KiwiSaver delivery model could work in this case too.

The other major difference is the UK's policy focus on low cost for the target new saver. This has allowed the expectation that people will not willingly make choices to result in the "one large fund" model. Less emphasis on low cost has allowed the KiwiSaver model to be consistent with a preference that people will make their own choices. It remains to be seen which model will result in better investment performance or more satisfied savers. But it is worth mentioning that there are other ways to control costs for the saver: cost caps could be imposed by regulation or, as in the early KiwiSaver model, government could subsidise fees.

In summary, therefore, KiwiSaver is a working model for other countries seeking a national auto-enrolment savings scheme with policy preferences in line with a balanced package of choice, risk and cost.

References

Department for Work and Pensions (2006) Personal Accounts: A New Way to Save, The Stationery Office, London.

Department for Work and Pensions (2007) Personal Accounts: A New Way to Save: Summary of Responses to the Consultation, The Stationery Office, London.

Inland Revenue (2008a) KiwiSaver Evaluation: Annual Report 1: 1 July 2007 – 30 June 2008, Inland Revenue, Wellington.

Inland Revenue (2008b) KiwiSaver Evaluation: Six-monthly Report 1: 1 July 2007 – 30 June 2008, Inland Revenue, Wellington.

Inland Revenue (2009a) KiwiSaver Evaluation: Six-monthly Report: July - December 2008, Inland Revenue, Wellington.

Inland Revenue (2009b) KiwiSaver Statistics as at 30 June 2009

Madrian, B.C. and D.F. Shea (2001) “The power of suggestion: Inertia in 401(k) participation and savings behaviour” Quarterly Journal of Economics, 106(4):1149–1187.

O’Connell, A. (2006) NPSS Policy and Design Choices, Pensions Policy Institute, London.

Office of the Minister for Social Development and Employment (2008) Government Response to 2007 Review of Retirement Income Policy

Office of the Minister of Finance (2005) Budget 2005 Savings Package: Work Based Savings Scheme, memorandum to Cabinet Policy Committee, 6 April 2005, http://www.treasury.govt.nz/

PADA (Personal Accounts Delivery Authority) (2008) Discussion Note Supporting the Prior Information Notice (PIN) for contracts to run the personal accounts scheme from 2012

Pensions Commission (2005) A New Pension Settlement for the Twenty-First Century: The Second Report of the Pensions Commission TSO London.

Pensions Policy Institute (2008a) Pension Facts, Table 14

Pensions Policy Institute (2008b) PPI seminar: Lessons from KiwiSaver

Retirement Commissioner (2007) 2007 Review of Retirement Income Policy, Retirement Commission, Wellington.

Savings Product Working Group (SPWG) (2004) A Future for Work-Based Savings in New Zealand

Footnotes

¹ Correspondence
Email alison@oconnellnz.com Alison O’Connell was Director of the Pensions Policy Institute (PPI) in the UK to 2006. This paper is based on a presentation at a PPI policy seminar in London in July 2008.

² “Defined contribution” means that contributions are invested and the amount available when taken out depends on the investment returns achieved. Each saver bears the investment risk in his or her own account. Defined benefit schemes used to be more prevalent than today in both the UK and New Zealand. In these schemes the pension benefit is defined by a formula usually related to number of years worked, with no investment risk passed to the individual member.

³ This is NZ Superannuation in the case of New Zealand, Basic State Pension and other second-tier public pensions such as State Second Pension in the UK.

⁴ Estimates are based on Government Actuary data for New Zealanders aged 18−64 and Pensions Policy Institute 2008a for people aged 15−64 in the UK.

⁵ These figures are as at 31 July 2008; KiwiSaver started on 1 July 2007.

⁶ Author’s own calculations; details available on request.

⁷ Author’s estimate; details available on request.

⁸ After PAs start, PADA will cease to exist and PAs will be run by a non-profit Trustee Corporation. However, this paper continues to refer to PADA for simplicity.

⁹ www.sorted.org.nz, run by the Retirement Commission.

¹⁰ The largest cost to the taxpayer from such a scheme is the cost of incentives. As already noted, the cost per saver in the UK from the incentives there (in the form of marginal tax relief) is probably higher than the cost per saver of the incentives in KiwiSaver. However, the costs being considered here are operating and risk costs rather than tax incentives.

¹¹ This decision may have been taken because the Pay As You Earn systems in the UK could not be amended to allow for savings contributions as easily as the equivalent systems in New Zealand could be for KiwiSaver.

¹² There are two recent (2007/08) precedents for this: the UK government has offered compensation to workers of troubled firms with under-funded pension schemes, and the Parliamentary Ombudsman has recommended that government compensate policy holders of pension provider Equitable Life for losses due to regulatory failures.