Seven out of 10 New Zealanders believe the country has a social class system - based mainly on how much you earn, according to a Listener Omni Poll.
Although income was seen as the strongest driver of social class, Auckland residents were more likely than the rest of New Zealand
to see education and residence as factors.
Ethnicity, occupation and family background were ranked next as driving the class divide.
Christchurch residents were less likely than people elsewhere in the country to see race as a factor in class.
Those earning under $40,000 were more likely to strongly agree there was a class system.
But a quarter of all males surveyed disagreed with the statement that there is a class system, compared with 13 per cent of the total sample of 1000 people.
People with a technical institution education or who were in managerial and technical roles also strongly agreed with that suggestion.
Those with tertiary education were more likely to perceive themselves as fitting into the upper middle class.
The survey also probed voter ideologies.
It found National voters were less likely to strongly agree that a class structure existed.
New Zealand First voters and those who say they would not vote are more likely to strongly agree with the statement.
Labour voters were more likely to see education and income as determining class.
When asked what class respondents perceived themselves as fitting into, National voters were more likely to say the upper middle class, New Zealand First voters were likely to say lower middle class.
Fewer undecided voters (5 per cent) saw themselves in the upper middle class.
New Zealand First leader Winston Peters was preferred Prime Minister by a quarter of lower middle class respondents and a quarter of upper middle class voters preferred National leader Don Brash.
The numbers
What is the NZ class system based on?
* Income 74 per cent
* Education 37 per cent
* Where you live 35 per cent
* Ethnicity 35 per cent
* Occupation 28 per cent
* Family background 25 per cent
* Gender 11 per cent
* Other 9 per cent
* Don't know 2 per cent
(sample: 1000 people)
