The Investor: Retirement Spending Puzzle Solved

Conventional wisdom says that two people can live on much less than twice what one person lives on. This is reflected, for example, in NZ Super payments. For those in the lowest tax bracket, singles living alone get $349 a week after tax, while couples get $537.

But there are a couple of probable reasons why Matthews' data doesn't show the "economies" of living together.

An obvious reason is that her numbers don't include housing costs - rent, mortgage payments or rates. Unfortunately many people are still renting or paying off a home loan in retirement. And everyone pays rates, either directly or as part of their rent.

Clearly, accommodation costs per person are cheaper for two people than one, even allowing for somewhat more space for a couple.

Another possible reason - pondered during the coffee break after Matthews' presentation - is to do with age.

Matthews' research looked at everyone whose main form of income is NZ Super (which by the way must exclude some wealthy retirees). She didn't break down the data into younger and older retired people, and it will be interesting to see if that research comes next.

However, we don't need research to tell us that people aged 65 to 75 are out and about and travel more than the 75-plusers. A common response to that is that older people spend more on medical costs. But Matthews' data show health doesn't loom large on retired people's budgets.

What has this got to do with one-person and two-person households? Well, at the time of retirement, a certain proportion - let's say one third - of households will be one-person, and most of the rest will be two-person. But as partners die, a growing proportion of retiree households will be one-person.

It would be lovely to counter this by saying, "What about all the retired people who find a new partner and marry or move in together?" But that's not all that common.

Logic tells us that, on average, the people in one-person retired households will be older than the people in two-person households. And if we assume that older retirees spend less than younger retirees, that would help to answer the question at the conference. Two-person households spend more because they are younger.

Do the numbers support this?

Matthews presented results for "No Frills" households, which are a quarter of the way up the spending range, and households with "Choices", which are three quarters of the way up. She also looked at "Metro" households, in Auckland and Wellington, and "Provincial" households elsewhere.

Comparing one-person and two-person households in each category, we find that two-person households tend to spend considerably more than twice what one-person households spend on: restaurant meals and ready-to-eat food; alcohol, tobacco and illicit drugs; household appliances; transport; and recreation and culture. The differences were smaller on basics such as other food and household energy.

All of which supports the idea that the average two-person retired household is younger and more actively involved in all sort of activities.

If we were able to compare one-person and two-person retired households in the same age groups - and we added accommodation costs - the idea that two can live more cheaply than one would probably be confirmed.

One odd finding: one-person Choices households in big cities spend $14.74 a week on clothing, while their two-person counterparts spend $11.79. What's that about?