Economist Bill Rosenberg runs the ruler over NZ Super cost predictions, and argues it's not as scary as we've been told; Calls for a proper debate on population policy

By Bill Rosenberg*

Last month, Prime Minister Bill English announced that his Government now favoured raising the age of eligibility to receive New Zealand Superannuation from 65 to 67 by 2040, a turnaround from its previous denial that any change was necessary. Labour has also reversed its 2014 election policy and now opposes the raising of the age.

The usual reason given for raising the age of eligibility is affordability of the scheme. For example the Commission for Financial Capability (formerly the Retirement Commission) in its 2016 Review of Retirement Income Policies justified it by asserting that in 2015/16 New Zealand Superannuation (NZS) cost 4.1 percent of GDP, and that “Treasury predict that it will rise to 7.1% (net) of GDP in 43 years”.

I have a look at these numbers and find that they are misleading. The picture looks very different if we take into account the tax paid by the New Zealand Superannuation Fund, the contributions both main parties say they will start making to the Super Fund, and the contributions the Super Fund will start making to the cost of New Zealand Superannuation in 2032/33 (according to current plans).

There are further arguments to be considered about the affordability to the public purse of current levels of payment and economic affordability.

The affordability to the public purse is frequently presented as an absolute. But affordability is a matter of priorities, what New Zealand society wants and what it is prepared to pay in the way of taxes. Many other countries are facing the same problem and many have considerably more retired people, and costs, in proportion to their populations than we do.

Economic affordability centres around the ‘dependency ratio’ – the number of people who are working, thereby generating income to be taxed and shared with superannuitants, compared to the growing number of superannuitants. But this does not take into account either other ‘dependants’ such as children, nor Treasury’s most recent long-term projection which did not show a looming economic problem.

Finally, it will become apparent that the question of New Zealand Superannuation cannot be seen on its own: we need to think about matters like New Zealand’s population and other forms of retirement income. I don’t cover these in detail but they are important to gain a full picture.

How much will New Zealand Super actually cost into the future?

Treasury looked at this question as part of its regular statement on New Zealand’s Long-term Fiscal Position. It released its latest one in November. In it they project current policies into the future to look at the ‘fiscal’ consequences of continuing those policies indefinitely – that is, the effect on government spending and revenue needs.

It shows that it is important not to rely on the headline figures of what the government spends on New Zealand Super. In the year to June 2016, payments to New Zealand superannuitants cost 4.9 percent of GDP or $12.3 billion. The projection shows that by 2060, the period of Treasury’s long-term Statement, it would be costing 7.9 percent of GDP – an increase of over 60 percent, which sounds a little scary.

But New Zealand Super is taxed – so it is actually its net cost after the claw-back of superannuitants’ tax payments that is important. (A Government could easily reduce its apparent level of spending by simply making New Zealand Super tax-free at its current net levels!) Its net cost to the government in 2016 drops to 4.2 percent of GDP or $10.4 billion. In 2060 it would cost 6.7 percent of GDP.

In addition, the Super Fund, set up by the 2000s Labour-led Government to partly fund New Zealand Super in the future, also pays tax. It doesn’t make much of a difference right now (though the $0.5 billion in tax paid in 2016 and $4.6 billion since the Fund started is not to be sneezed at) but by 2060 the tax is projected to be $8.0 billion a year. That brings the net cost by 2060 down to 6.1 percent of GDP.

Then there is the direct effect of the Super Fund. Firstly, both main parties (and the Retirement Commissioner) want contributions to the fund to resume – some earlier than others. The current Government won’t restart contributions until the year to June 2021. From then until the year to June 2033, when withdrawals are started from the Fund, there is an additional cost of up to $3.0 billion (which incidentally isn’t counted in Core Crown expenses because it is regarded as capital spending). In 2021, the total cost to the government of New Zealand Super plus these contributions is projected to be 5.0 percent of GDP. It would have been the same in the year to June 2016 had contributions resumed that year. By 2060, when the Fund is projected to be contributing $4.0 billion to that year’s New Zealand Super costs, the net cost to the government of the day amounts to 5.9 percent of GDP.

So the true comparison of the fiscal cost now and the cost in 2060 is more like this: 5.0 percent of GDP soon, compared to 5.9 percent in 2060. That increase doesn’t look nearly as scary. If it is affordable now (as all seem to agree) then it is likely to be affordable in 2060. Here’s a table summarising the situation with the total impact on government finances in the bottom line:

Affordability to the public purse It is wrong to present affordability to the public purse as an absolute in the way that English and some others put it: he was raising the retirement age to “ensure the scheme remains affordable into the future” . Affordability is a matter of priorities, what New Zealand society wants and what taxes we are prepared to pay.

The Retirement Commissioner points out that there are other costs that rise as the population ages such as health care. But to draw the conclusion that superannuation should be cut to pay for these is not logical, unless we are moving to a society where each generation is expected to look after itself and not concern itself about younger or older generations.

We do always need to consider our priorities and options, but one option is to raise more revenue. Many New Zealanders would be willing to pay more to maintain the financial security of their parents and themselves in retirement, and for other public services that they value. If the assumption is that the current level of taxation is fixed and can never increase as a proportion of GDP then there are many other problems we cannot address and the outlook for New Zealand is dim. Some of these problems cannot be addressed by individuals on their own (such as environmental problems, income inequality and poverty). In other cases such as health and retirement income, individuals could pay for it but it becomes inefficient, inequitable and impoverishing (as the US private health system shows). It is much better for everyone if the risks are shared and it is paid from government revenue. It might raise the government’s costs, but from an economy-wide and societal point of view it costs less and is fairer.

We are not a highly taxed country. OECD data shows that we have one of the lowest differences between what an employer pays and take-home pay among the high income countries that make up the OECD – we rank between 28 and 34 out of 34 countries and well below the OECD average. Our tax revenue as a proportion of GDP is low for a small country . We rank 19 out of 35 OECD countries and less than most similar size OECD countries. Our problem is not the level of taxation but its distribution.

The proportion of GDP New Zealand spends on pensions is also low. The OECD put it at 4.8 percent of GDP in 2013 (a misleading figure but it is just a basis for comparison). The OECD average was 6.4 percent of GDP and only 10 out of 33 countries had a lower proportion, including Mexico with zero and others that also have private compulsory contribution schemes. There were 14 already above 7 percent of GDP. New Zealand’s ratio is low partly because we are fortunate to have a relatively young population.

Clearly affordability is a decision that societies make in terms of their priorities.

Economic affordability

Much of the economic debate centres around the ‘dependency ratio’ – the number of people who are in paid work and thereby generating income and tax revenue, divided by the number of older people. This is projected to fall: fewer people will be working to generate the income required for each retired person. Treasury’s population projections show it falling from 7 working age people to every person 65 or over in 1972 to 4.3 in 2017 to 2.1 in 2060. (That’s taking the working age population to be aged 15 to 65. It’s unlikely that adjusting the lower limit of 15 up a little would make a big difference to the analysis. Statistics New Zealand defines it to be aged 15 years and over.) . Of course the effect of the fall in the ratio will depend on how many people of working age are working (the participation rate), and how many of the over 64s are working. The participation rate is currently rising more rapidly in this age group than any other.

But consider this: people over 64 are not the only ‘dependants’ in society (and an increasing proportion of them are not dependent either). Children make up the other main group of dependants. In 1972 children under 15 made up 31 percent – almost a third – of New Zealand’s population and the 65+ age group only 9 percent. The working age population made up 60 percent. The whole ‘dependency ratio’ was 1.5 working age people to every dependant. The population is aging in two ways: we have a greater proportion of over-64s and a falling proportion of children under 15. In 2017 the children made up only 19 percent of the population, people of working age made up 65 percent and the over-64s 15 percent. The ‘dependency ratio’ is 1.9. By 2060 the projection reduces children to 16 percent of the population, people of working age 57 percent and the over-64s to 27 percent. The ‘dependency ratio’ would be 1.4 – not much lower than the 1.5 it was in 1972. So in 2017 we are in a sweet spot – the highest dependency ratio since 1972 was in 2.0 in 2006 and we are not far from that. Perhaps this is the unusual time rather than 2060! The effect of this all depends on the cost of raising, educating and looking after the health of children compared to the costs of old age.

It is interesting that Treasury’s economic projections for the size of the economy do not show an economy struggling to pay for New Zealand Super – otherwise its cost as a proportion of GDP would be much higher. It is of course dependent on its assumptions which may be unrealistic. These include a high proportion of people continuing to work, and in particular among the 65+ age group. It also assumes that labour productivity grows at an annual rate of 1.5 percent and that real wages (the average hourly wage adjusted for rising prices) grow at the same rate. Productivity has been struggling well below that level for a decade. Wages since the early 1990s have failed to keep up with productivity.

However if the link between productivity and wages were achieved, Treasury observes that raising productivity is not the answer to paying for New Zealand Super: raising productivity raises wages, which raises the cost of Super because it is linked to wages, and we are no further forward without more progressive tax rates.

All of this means that this modelling cannot be the final word on the subject. But it is not immediately obvious that there is an economic reason to reduce the cost of New Zealand Super.

Finally

This discussion raises many questions: the question of New Zealand Superannuation cannot be seen on its own. What would be the impact on its affordability of increasing our future working age population by encouraging people to have more children or a somewhat higher level of immigration (better managed than now)? We could encourage more children by paying a universal child allowance, making child care better quality and free, and reducing working hours. Treasury says that if it raised its assumed net immigration rate from an average of 12,000 per year to 25,000 per year in the long run (both much lower than at present), “population ageing slows and the population is younger and approximately 928,000 higher in 2060. The higher net migration lowers the ratio of expenditure-to-GDP” and reduces net core Crown debt. These questions add to calls for a proper think about where we want for New Zealand’s future population to head – a population policy.

New Zealand Super is not the only income retired people rely on. We should be thinking about boosting Kiwisaver, and the Retirement Commissioner recommends raising contribution rates. The CTU has proposed making Kiwisaver compulsory if the employer contribution rate was raised to 6 percent, there was a 2 percent contribution from both workers and the government, the minimum wage was increased at the same time, and the government contribution of 2 percent (of minimum wage or benefit level or another amount) applied to all those of working age who are not earning for a period.

We should also be thinking of fair ways to pay for the increasing cost. Susan St John in the Auckland University Retirement Policy and Research Centre has made proposals for a progressive tax on those receiving New Zealand Super but more universal solutions may be less contentious given our history.

There will never be a last word on this subject. We should continue to review the situation, keeping a watch on both the adequacy of our people’s retirement income and the cost of it. But New Zealand is lucky enough that we don’t have to make urgent decisions to manage the cost of New Zealand Superannuation.

* Dr Bill Rosenberg is the Policy Director and Economist for the New Zealand Council of Trade Unions. This piece was first published in the CTU's March Economic Bulletin and is republished here with permission.