Retirement Policy
Ponzi schemes do not invest their investors’ money but rather rely on succeeding generations of investors to pay out the previous generation and deliver the unexpectedly high levels of return. They are unsustainable, because to pay each previous investor you must find at least one replacement, which is difficult to do as the population which has not yet been fleeced shrinks. Unfortunately, however, this scam was not merely used by Bernie Madoff to get obscenely wealthy (and arrested), it was also Robert Muldoon’s vehicle to election victory: New Zealand Superannuation is the largest Ponzi scheme to ever operate in this country.
Pensions in New Zealand operate under the intuitive PAYGO system: Each succeeding generation pays for their parents’ retirements, trusting that their children will pay for theirs, ad infinitum. That system works reasonably well so long as the working age population is growing; that ensures that the individual burden on workers of this system is not excessive. However, as the dependency ratio (the number of retirees each worker is required to support) increases, it begins to break down.
This is the very problem we face today. The baby boom, the generation born after the Second World War, and their children are coming into retirement and bringing with them a truckload of pension liabilities. After the War, our development cut infant mortality rates significantly but this did not flow through to lower birth rates for a while: In essence, families still had plenty of children in anticipation some would die in childbirth or infancy, but they didn’t, leading to an exceptionally large generation. Unfortunately for us, their children and grandchildren, they didn’t have many of us. The Sexual Revolution, the development of the contraceptive pill, our growing wealth, and the increasing involvement of women in work all collapsed the birth rate in their generation. This means there are fewer of us to share the burdens of paying for their retirement. This problem is compounded by the fact their retirement is likely to be incredibly long and expensive: Life expectancies have increased significantly, and with them levels of chronic illness have increased too. Expenditures on hip replacements and dementia care, previously so small as to be immaterial, are increasing at a rapid rate.
The worst part of it all is that the pension doesn’t even work for many who receive it. The basis of any retirement scheme is surely to provide security in retirement to all. No civilised society would or should accept the elderly living as paupers. The present system fails on this account. The $20,000 of the pension is insufficient to cover reasonable living expenses for many retirees, especially those without a freehold house who must pay rent or a mortgage. For the most prosperous pensioners, however, the $20,000 from the pension is an afterthought, a nice-to-have which funds luxurious holidays abroad.
This leaves us with a significant conundrum. Our generation should be assured that, if we’re going to be paying into such a system to pay for previous generations’ retirements, some similar system will exist to provide for us.
The Solution
Some would suggest merely increasing the age of entitlement for Superannuation, in line with life expectancy. This was Sir Bill English’s attempted solution to the problem as Prime Minister. This is certainly a significant improvement on the status quo – which is fiscally irresponsible. It is not, however, a panacea. Firstly, it probably punishes some manual labourers excessively. Though scientific advancements and improvement in medical care have extended life expectancy, we have not necessarily increased the length of one’s working life where one can work in physically demanding roles. Pushing out the date at which those who work in such jobs can receive state support seems cruel. Secondly, it doesn’t repair the mechanistic flaws in the status quo – it simply kicks the can down the road. In the future, another birth cohort might bubble through and screw over their children while the Ponzi scheme structure remains in place. Thirdly, it doesn’t end the uncertainty which faces those planning for their retirement: Now they may be assured that Super will survive until they retire, but they may not be sure when they will be eligible for it. Once such a change happened, altering the eligibility requirements for Super once again might become a plausible option available to Finance Ministers looking to balance the books
Better, in my view, to just get radical changes over and done with. We should shift to a prefunded superannuation system. Sir Michael Cullen’s attempts to head down this path, when he created the New Zealand Superannuation Fund and KiwiSaver as Minister of Finance, were admirable, but they did not go far enough. But nor should we simply go back to the New Zealand Superannuation Scheme of the Third Labour Government. KiwiSaver should become the basis of this new system, supported by a system of public retirement insurance.
The changes to KiwiSaver I suggest would be relatively simple. Firstly, and most crucially, it should become compulsory for all workers. The compulsory rates of contribution from both employers and employees should be increased [1]. Secondly, all government subsidies and tax advantages for it should be removed. Thirdly, every KiwiSaver account should be split in two: Compulsory contributions and discretionary contributions. As the name suggests, compulsory contributions, from employers and employees, should be paid into that account, whereas all contributions in excess of the statutory rates should be paid into the discretionary contributions account.
This system of privatised retirement accounts should be supported by a guaranteed retirement income scheme. The Government should determine a base income below which they are not willing to let any over-65 fall – clearly this should be higher than the present Super amount. Just as now, this should be indexed to wage growth. For the sake of convenience, I’ll call this scheme Retirement Insurance. It should be the replacement for New Zealand Superannuation.
This system of Retirement Insurance should be operated by an independent Crown Entity – the NZ Retirement Insurance Fund – and be principally self-funding. It should take ownership of the assets of the Super Fund and receive an annual premium from every working New Zealander, as a small percentage of his income. It should invest those premia in long-term securities, just as the Super Fund does today, to pay for future retirements. The crucial distinction here, however, is that the NZRIF should be sustainable. Capital depletion should not take place: The NZRIF should only pay-out the returns it earns in excess of the amount necessary to protect against inflation and build some capital. To pay for the remaining amount of retirement support (of which there will be much in the early years, as the NZRIF builds its capital), there should be an unlimited Crown guarantee, raised from general revenue. This mechanism would ensure that the Fund would be well-equipped to handle bubbles of retirees in future, because they would have paid-in to the Fund and accumulated a significant principal, from which Retirement Insurance would be funded.
The amount superannuitants receive in Retirement Insurance should be based only on the balance in their compulsory contributions account of KiwiSaver. Each retiree should have calculated for him a “sustainable income” he could be expected to be able to draw from his KiwiSaver compulsory account balance each year. This should be based not merely on the interest received but also a sustainable rate of capital depletion. The exact value of this “sustainable income” should depend on three factors: The balance of the retirees’ account at the beginning of the year, the expected rate of return on conservatively-invested assets [2], and his life expectancy (based, for simplicity’s sake, only on his age and gender). Retirees with a “sustainable income” up to twice the guaranteed retirement income should receive RI payments. As your “sustainable income” increases by $1, the RI you receive should fall by 50¢ [3]. As a mathematical consequence of this setup, every retiree would receive at least the guaranteed retirement income.
An abatement rate of 50% is quite high. However, it is unlikely to have a significant disincentive effect, because contributions would be compulsory. One might therefore reason, why not have a 100% abatement rate and make the plan more steeply progressive and affordable? Firstly, because that would significantly discourage earning during the working lives of savers. In that scenario, until savers had more in their compulsory contributions account than would be required to deliver a “sustainable income” in excess of the guaranteed retirement income, every dollar they saved would be in effect giving them net-0 personal benefits. This would mean the entirety of the compulsory contributions would be, in essence, a tax and create a larger disincentive to work. Secondly, because an 100% abatement rate would encourage excessively risky investments with the compulsory account. The amount in the compulsory account would be immaterial to retirees without the amount required to deliver a “sustainable income” in excess of the guaranteed retirement income, because whether it was there or not, they would receive the same retirement income. They would be more likely, therefore, to put these funds in high risk/reward investments than otherwise – because the upside would be significant, while the downside would be 0. The creation of such moral hazard would make the system significantly more expensive to operate. With a 50% abatement rate, both problems still exist to some extent, but they are reduced.
Some, especially libertarians, will disagree with compulsory superannuation on principle. Prima facie, it seems paternalistic to force people to save for their retirements. I do not believe it is. If we begin from the premise that the state should provide support for retirees, taking it as true for now for the sake of argument, we have two options: Provide a universal benefit to all, regardless of need (our present method), or means-test the pension and provide it only to those who need it least. The first option has significant flaws. The first, about the risks of a PAYGO method when birth cohorts fall in size, has already been discussed. But it is possible to imagine a prefunded universal state pension (for instance, through a beefed-up NZ Super Fund) which is invulnerable to such criticisms. Still, however, compulsory individualised retirement accounts with a state top-up are preferable. Firstly, because they vest control of people’s money with themselves. With a prefunded state pension, workers would have no choice where their money was invested. This means they could not prefer, for instance, investments which better meet social objectives they support. Second, it also diversifies risk away from a single point of failure. A poor decision by the Guardians of the NZ Super Fund would affect all retirees. By contrast, a much rarer systemic error would be required for the same effect to propagate through individual retirement accounts [4]. Third, it is simply unfair. Why should those who are well-able to fund their own retirements receive just as much support from the state in their old age as those who are unable? They simply need it far less.
Means testing is the only fair option. With the same allocation of funding, means testing allows the level of support to needy individuals to be much higher. Unfortunately, means testing cannot really work in the absence of compulsory superannuation. This is because it would create a significant disincentive to save.
There is, however, a clever technical solution to this problem which allows the freedom to choose not to save and can create sustainable means-tested pensions. This is to base means-testing not on actual monies saved but on a shadow account, which would simulate the returns from taking a percentage of the individual’s income (but not actually taking it, of course) and investing them in reasonable investments (e.g., a shifting balance of shares and bonds, in accordance with wages). This model would change the means-tested support you receive based on your lifetime income, not your savings – still encouraging individuals to save as much as they like. This approach has significant appeal to me. Unfortunately, however, the variability in retirement income across individuals is not solely driven by different lifetime incomes, but also different returns on investment. To take an extreme example, if a lawyer had been a top-10% earner for his entire life, but his retirement fund was poorly invested, leaving him penniless at age 65, while the stock market in general had continued to perform acceptably, in a shadow account model, he would receive no state support at all – despite being essentially a pauper. That does not seem fair to me. Your retirement support should be based on your ability to afford retirement right now, rather than the hypothetical ability to afford retirement you would have had had you invested in a benchmark portfolio.
I do, however, note that many would disagree with taking the premise that the State should guarantee at least some level of retirement income as a given. This is certainly a small group, but that doesn’t dilute the fact they may well be right. They might contest that retirement is an entirely predictable outcome and does not meet the threshold of unpredictable happenstance where social insurance might be valuable. Note, however, that while retirement is entirely predictable, your lifetime earnings and the success of the superannuation fund you invest in are not. The case for social insurance at this stage of life is strong – certainly stronger than the case for social insurance for working-age individuals. Working age individuals might be reasonably expected to work their way out of poor decision-making or bad luck, but the oldest in society do not have that luxury.
Overall, it seems clear that if we remain committed to providing for the elderly, as we ought to, the only equitable and efficient way of doing so is via a system of compulsory retirement savings.
Policy Recommendations
- Abolish New Zealand Superannuation
- Make KiwiSaver compulsory for all employees and increase the required contribution
- Implement a retirement insurance fund scheme, funded by an annual surcharge on compulsory retirement savings
[1] The distinction between “employer” and “employee” contributions to KiwiSaver is, in an economic sense at least, false; both are factored into employers’ decisions on net pay and the economic incidence between employees and employers is equivalent in both cases: Entirely on employees. If I was designing a system from scratch – and basing my design purely on economics, not politics – I wouldn’t include any distinction. I cannot, however, see any harms to the status quo continuing either.
[2] New Zealand Government bonds seem an obvious ‘conservative’ vehicle for investment – given they have an effective 0% default risk – and, therefore, an obvious benchmark on which this rate-of-return could be based. However, the rate of return on other assets, more realistically representing the investment of retirees, like AAA-rated corporate bonds, foreign government debt, and term deposits at local banks, might be more appropriate to provide some level of realism.
[3]
Let I
be the Retirement Insurance payments received by this retiree,
S
be his “sustainable income”, and G
be the guaranteed retirement
income. In this case, I = (2G-S)/2
[4] Of course, I am aware that the same ‘single point of failure’ argument applies to the new NZRIF. Notably, however, the costs of failure would be much lower in a system of insurance, rather than full state-funded pensions, because the required shortfall the State would have to make up would be smaller.