A new look at setting retirement income policies
by Michael Littlewood
Former co-director, Retirement Policy and Research Centre, University of Auckland
The Retirement Commissioner is currently reviewing the country’s retirement income policies and has called for suggestions to assist her. This kind of review has happened regularly (1992, 1997, 2003, 2007, 2010 and 2013) but governments of the day have essentially ignored all but the first one in 1992. There have been contemporary political reasons for this relative indifference but I wonder if we need to step back a piece from the detail and think about the public policy issues from a wider perspective.
My submission to the 2016 review (accessible here) suggests such an approach. We need to discuss what governments can do and, based on the evidence, what they seem unable to do.
I suggest that there are just four main things that only the government can do with respect to retirement incomes and saving:
- Tax and redistribute: Using its power to tax and redistribute, a government can reduce or, perhaps even eliminate poverty in old age. New Zealand Superannuation presently does a reasonable job of that but the question is whether NZS can fulfil a similar role over coming decades?
- Regulating and taxing investments: Its power to regulate means that only a government can set disclosure requirements for investment offerings and tax equivalent investments equivalently. That does not happen at the moment. For example, the tax treatment of ‘collective investment vehicles’ is a complex, inconsistent mess.
- Impeccable, deep data: Only the government can demand access to information and produce deep, impeccable data to help us understand what is actually happening out there. For example, despite Census data, we do not really know what proportion of New Zealanders own their own homes; nor whether that is changing. Nor do we really know whether New Zealanders are under-saving for retirement. Are either of these issues things we should actually be concerned about? Without impeccable data, we cannot answer those questions.
- Information and education: Finally, a government’s information and education programme is more likely to have traction than information from financial service providers.
International evidence suggests that governments cannot force or encourage people to save more for retirement than they want to save. That should eliminate compulsory private provision and tax-incentivised saving programmes as effective policy options. That’s mainly because they seem not to work. However, that is just one of a long list of problems with both compulsion and tax incentives.
The four ‘first principles’, if accepted, mean that we have quite a lot to do before New Zealanders can have a proper discussion about, for example, the size and shape of New Zealand Superannuation in the coming decades. There is so much that we don’t know and only the government can fill those information gaps.
In summary, I suggest there is a range of things that only the government can do – it should do those things. There is another range of things that, based on the evidence, the government seems unable to do – it should stop doing those. Finally there are things that the government is doing but, based on the evidence, seem not to be effective – it should also stop doing those.
This is evidence-based policy-making – if it works, based on the evidence, then do it; if it doesn’t work, stop doing it. If we do not know whether it works, gather the evidence before deciding what to do. For example, do we really know whether KiwiSaver is ‘working’? Is anyone asking that question? What data do we need to gather to answer that? Counting the number of members and the amount of money saved does not answer the question because KiwiSavers will have changed their ‘other’ behaviour as a result of being members. That’s what a rational KiwiSaver should do.
For New Zealand, this suggested approach to policy–making on retirement incomes would be a change but I think it’s time we tried it. Before we can even start that process, we have a lot of data to gather.