Retirement nest-eggs for many

By BRIAN FALLOW

Contrary to the conventional wisdom, it may be that people are saving enough for retirement.

If, that is, you narrow "people" down to 45 to 55-year-olds and define "enough" to mean enough to maintain the same standard of living after retirement as before it.

That is the tentative conclusion of a study by Treasury economist Dr Grant Scobie and Waikato University's Dr John Gibson analysing data from the household savings survey.

The survey is a snapshot of the wealth of nearly 3000 couples and 2400 singles taken by Statistics New Zealand in 2001.

Confining themselves to the 45-to-55 age bracket, as those are the peak earning and saving years, they asked: given what the survey says about the net worth and earnings of those people, what percentage of their pre-tax incomes would they have to save to smooth consumption across the rest of their working lives and through their retirement years?

They then compared the theoretical savings rates their model said would be needed to smooth real consumption, or maintain a steady standard of living, with actual savings rates from another source, the 2001 household economic survey, and found a pretty close fit.

They concluded that "there does not appear to be strong evidence from the household savings survey of significant under-saving for retirement by New Zealand households aged between 45 and 55".

But this finding is wrapped in ifs and buts, qualifications and caveats.

Because of the wide disparities in the distribution of income and even more of wealth, they divided the age group into tenths or deciles by wealth and by income.

However, the result of this greater precision is generalisations from relatively small numbers of survey respondents.

Another issue is that the economists doing the model know the average life expectancy, and how it varies between men and women, Maori and Europeans. But the individual does not know how long he or she will live, and the model makes no allowance for "wasted" over-saving on precautionary grounds.

As Scobie and Gibson acknowledge, there are other possible definitions of adequacy. One would be a relative measure, like keeping retirement incomes at least equal to some fraction of the average incomes of the working population.

Peter Harris, a former economic adviser to Finance Minister Michael Cullen, is critical of the assumption that people just want to continue to consume in retirement what they were consuming while working.

That implicitly cements in the income distribution, he says, "as if those in the poorer deciles are used to being poor. They may even be better off on New Zealand Superannuation, so they would be irrational to save."

The analysis includes entitlement to NZ Super, assuming average life expectancy, in the calculation of people's wealth. For the poorest 40 per cent of couples, it represents the lion's share of their wealth, ranging from 90 per cent for the first decile to 58 per cent for the fourth.

Surprisingly perhaps, housing wealth (net of mortgages) is a smaller proportion of total wealth than NZ Super entitlement for every decile.

But housing is still more important than financial wealth (including private super) for all but the richest 30 per cent of the age group.

The study found that the savings rate people need to maintain their standard of living is very sensitive to how they deal with the money tied up in housing when they retire.

For example, individuals in the sixth decile (by income) who retire at 65 and consume none of their housing wealth will need to save 12 per cent of their income and can expect a gross retirement income 65 per cent of what they had before retirement.

If they trade down, and free up 50 per cent of their housing wealth, the required saving rate drops to 7 per cent and the income replacement rate rises to 72 per cent.

If they are willing to sell up altogether and rent, the required saving rate is only 4 per cent and the replacement income (which now has to cover rent as well) is 74 per cent.

Scobie and Gibson chose to err on the side of caution in the assumptions they build into their model.

They assume that the value of houses and farms does not rise in real terms, that is, faster than general consumer price inflation.

They assume pre-retirement incomes grow by only 1 per cent a year, to approximate the average annual growth in labour productivity and real wage growth.

They assume NZ Super payments are constant in real terms. Again this is a conservative assumption as super payments are linked to the average wage.

And they assume a rate of return of 2 per cent on business and financial assets, after tax and inflation.

On the other hand there is no allowance for the possibility of wealth-depleting shocks between now and retirement such as ill-health, marital breakups or redundancy.

For couples retiring at 65 and using up none of their housing wealth, the required savings rates range from minus 13 per cent for the lowest income decile (implying they will be better off on NZ Super) to a peak of 25 per cent for the fifth decile.

For the whole age group, the median required saving rate is 18 per cent, which would replace 54 per cent of pre-retirement income.

Actual savings rates from the household economic survey turn out to be close to those prescribed by the model for most of the distribution.

But that still leaves 10 to 15 per cent who can expect a retirement income of less than 60 per cent that of their contemporaries - one definition of a poverty line. Harris worries that people will draw false comfort from the study - "It is hard to reconcile this with aggregate data about what households are doing."

Measures of household incomes and outlays from the national accounts show saving rates falling and now in negative territory, implying that households overall are spending more than they earn. Other data show growing household debt.

"If they are right, that's fine," says Harris.

"But if they are not, what do you do in 40 years' time? It's a long way home."