An alternative savings measure suggests households might be doing better than commonly thought at saving more than they spend.

But even the re-jigged figures cannot erase a noticeable plunge in saving rates during the mid-2000s' housing boom.

Westpac economists used an approach dubbed "equity injection" that was originally devised by the Reserve Bank and updated it with more recent data.

The result bolstered arguments that household savings might be higher than suggested, but the bank acknowledges the figures remain shaky and it can not say how much better the picture might be.

Like the alternative medal tables that became popular during the Olympics, there are several ways to measure household savings, each with their own problems.

The most commonly cited one - national income minus national expenditure - is criticised for underestimating income, including overseas earnings, while accurately measuring expenditure, resulting in a smaller savings rate.

The alternative measure Westpac used takes household investment in assets such as houses and managed funds and subtracts the household level of borrowing.

That approach, too, is imperfect and prone to miss private businesses and other sources of wealth.

Westpac senior economist Felix Delbruck said the figures were not a replacement for or an improvement on the official figures but they showed how sensitive measures of saving could be.

"It is the best among the cross-checks."

The alternative measure resulted in a similar trend, showing saving dipping in the mid-2000s and then starting to climb again.

But it was several percentage points higher at almost all times since 1990 and usually remained in positive territory, unlike the official measure.

If farm sales were included then saving barely dipped below zero at any point.

Both the common income-minus-expenditure measure and alternative Westpac/RBNZ approach have been criticised for taking a large, imperfect number that has been calculated at a national level and subtracting it from another national number to draw conclusions about individual households.

Even data based on actual household surveys can underestimate spending and miss income from sources such as family trusts and businesses, researchers have found.

However Waikato University economist John Gibson said survey-based measures were the most valuable because they shed light on individual households' saving and whether it was appropriate given their age and income. "If you use the aggregate and get a broad figure that the national saving rate is minus 5 per cent net you have no basis to know if that is the same for people who are already retired, and for whom it is quite sensible to be withdrawing savings, or whether it is people between 45-65, who would usually be doing the bulk of their saving," said Gibson.

Three out of the four most common savings measures resulted in a picture that was less pessimistic than the commonly published one, he said.

The uncertainty highlighted the need for better financial literacy.

"We know these are fairly complex decisions and people are making them with not great literacy so ploughing money into (retirement commissioner and Sorted-org.nz's) Diana Crossan and into improving people's ability to make these financial decisions seems like a great investment."

New Zealand is generally thought to have poor household savings rate compared with similar countries such as Australia. But attempts to measure income more equitably between countries have suggested New Zealand may be closer to other comparable countries than was previously thought.

Critics who disagree with costly Government measures such as KiwiSaver have pointed to measures suggesting households are doing better than supposed.

As a cross-check equity injection was the best because it was the most comparable to the usual RBNZ series and allowed users to pick and choose the most reliable sources of data, said Delbruck.

- © Fairfax NZ News