Confidence in the big cities and provincial towns fell too, but not as much.
It is a troubling indicator because the recovery has been export-led and underpinned by a 20 per cent rise in the terms of trade over the past two years to 37-year highs, boosting national income.
Forecasters expect the terms of trade (export prices relative to import prices) to deteriorate from here, though not sharply, effectively shutting down what has been a major impetus in the recovery under way over the past 2 years, plodding as it may have been.
Another factor underpinning the recovery is also being turned off - fiscal policy.
Right now it is still on balance stimulatory though much less so, of course, than during the recession and its immediate aftermath.
But over each of the next two years it will be more contractionary than at any time in the past 15 years and probably since the Mother of All Budgets period of the early 1990s.
The consensus among economic forecasters polled by the New Zealand Institute of Economic Research this month is that household spending will remain subdued.
Private consumption growth is expected to average 2.2 per cent over the next two years - better than nothing and a pickup from the past year, but still only half the average growth rate over the five years before the recession.
The slowdown in private consumption compared with the last boom period reflects three main factors: higher unemployment, lower wage growth and a shift in households' attitudes towards debt.
Statistics New Zealand says that in the year to March households spent less than their disposable income for the first time in 10 years, though only just.
The consensus forecast is for the unemployment rate to decline from its current 6.6 per cent to 6.3 per cent by next March, 5.8 per cent a year later and 5.1 per cent by March 2014.
That compares with an average unemployment rate of 4 per cent over the five years before the recession and 6.5 per cent since it technically ended 2 years ago.
In the Westpac survey only a net 5 per cent of respondents expect to be better off financially in a year's time.
Since the grim days of the early 1990s that indicator has only been weaker three times, during the 2008 recession and immediately after February's quake.
It is important because research by two Reserve Bank economists, Michael Reddell and Emmanuel De Veirman, indicates that actual and expected incomes have been the largest influence on household consumption over the past decade, rather than the housing boom and the associated wealth effect.
"For quite a few years, at least until around 2005, [gross domestic product] ran well above most expectations. Times were good, unemployment was very low, wage rates were rising. It would not have been at all surprising to see households consuming as if the good times would last. There was no hint, whether from official agencies or private forecasters, that they would not."
Since the recession of 2008/09 it looked as though people have been spending on the assumption that future incomes would be rather lower than they previously expected, they said.
Their work, reported in the latest Reserve Bank Bulletin, challenges the conventional story of the last decade, which goes like this:
A housing boom, fuelled by banks' ability to borrow offshore in a world awash with cheap money, encouraged homeowners to spend some of the associated increase in their wealth, allowing consumption to grow faster than incomes and stoking inflation to the point where the Reserve Bank had to push the official cash rate over 8 per cent, tipping the economy into recession even before the global financial crisis hit.
Between 2002 and 2007 house prices more or less doubled. The increase was fastest in 2003 and the bulk of the rise had occurred by 2005.
But De Veirman and Reddell point out that an increase in house prices does not amount to an increase in consumption possibilities for the economy as a whole.
"When house prices increase, the asset wealth of existing homeowners increases. But the cost of acquiring a house also increases for everybody who may ever wish to buy a house. So a rise in house prices does not make the population as a whole materially better off," they said.
"The relatively old, who might have been about to downsize, benefit from the windfall, and can spend more than otherwise over the rest of their lives. But the relatively young, coming into the housing market, have to pay the higher cost of housing, and find themselves worse off."
Landlords benefit because they can sell and bank a profit, but tenants are worse off.
Overall, there is some increase in the "collateral effect" - banks' willingness to lend - on the strength of rising home equity.
But Reddell and De Veirman point to the fact that household consumption, measured against national income, remained quite subdued during the period of most rapid house price inflation and never reached exceptional levels at any point in the decade.
They conclude that the surge in house prices did not raise private consumption so much as alter its distribution.
A more potent effect from around 2005 on, they suggest, was the loosening of fiscal policy.
After-tax incomes were boosted, not least by Working for Families, while the Government's own consumption of goods and services rose rapidly as a share of national income, albeit from its lowest level in decades.
Total consumption, private and public, jumped from around 80 per cent of gross national income in 2004 to 84 per cent two years later.
Government spending rose from 29 per cent of GDP to 31 per cent over those two years, but the impact of the recession has seen it climb to 35 per cent now. Getting it back down to 31 per cent over the next three years as the Government intends will be far from painless.
