The picture varies from area to area because of small numbers, but Manurewa central roughly reflects the national Maori/European smoking divide.

Nationally, 32.7 per cent of Maori aged 15 or older smoke, compared with 13.9 per cent of Europeans. Pacific people are on 23.2 per cent, Asians 7.6 per cent - and the whole population, 15.1 per cent.

Slicing the 2013 smoking statistics up by poverty reveals even greater differences. The smoking rate in the most-deprived areas is more than four times greater than in the least-deprived.

By 2013, all major ethnic groups' smoking rates had declined since the preceding Census, in 2006, but the percentage reduction was greater for Europeans than for Maori, a trend present since the 1980s when the national adult smoking rate was 33 per cent and the Government made its first big efforts to get Kiwis to quit.

Researchers say history predisposes Maori to having higher rates of smoking than other ethnic groups in New Zealand, and this is passed on from parents to children like a contagious disease.

Professor of public health Tony Blakely, of Otago University at Wellington, ties this back to the European colonisation of New Zealand and the use of tobacco to buy Maori land.

"A hundred years ago Maori females had very high smoking rates. For non-Maori females it started much later, post-World War II. There has been a longstanding history of high smoking among Maori, which comes back to the way trading was done initially.

"You've got a cultural norm and a contagion ... of smoking transmission."

Professor Blakely said the higher smoking rate of Maori was in part explained by lower socio-economic status - SES, a measure of income, educational level and occupational class - but only to a small extent.

Maori were "coming off a very different history of a very high smoking rate which means you've got further to go down".

"Normally what we see around the world is rich males take up the habit first and drop it first, followed by low-income males who take it up, drop it next, followed by high-income females, then low-income."

New Zealand is largely following the same pattern, although Maori females are an exception.

"There are these waves in smoking going up then going down, and the lower socio-economic groups tend to take it up last then take a long time to give it up. It's just out-of-phase epidemics.

"Why. Because when smoking initially comes into a society it's seen as glamorous and hits the higher socio-economic groups first. Then they realise it's bad for them and give it up first because they have higher levels of education, better knowledge and better income to get quit treatments, although we [the state] do try to subsidise them now."

In a 2003 paper, Professor Blakely and colleagues reported that although smoking rates had declined between 1981 and 1996, ethnic inequalities had widened under the indoor smoking restrictions at many workplaces, the controls on tobacco advertising and other mainstream tobacco control policies of the time.

"These mainstream interventions appear to have been more effective for those population groups who already had the lowest rates of smoking," they wrote.

"Thus the overall prevalence of smoking may have been reduced at the expense of growing inequalities in tobacco use and tobacco-related health outcomes."

It was several years after the 1996 Census before policies were introduced to specifically target Maori smoking (1998), and to make reducing health inequalities a major health goal (in 2000).

Doubts remain about whether New Zealand can meet the Government's target of being a largely smoke-free nation - widely interpreted as a smoking prevalence of less than 5 per cent - by 2025. The latest projections from Professor Blakely's group, published in the New Zealand Medical Journal, predict a European rate of around 7 per cent and Maori rate of 19 per cent, although these figures don't take into account the two 10 per cent tobacco excise tax rises scheduled for next month and 2016.

In an Otago University blog post last month, Professor Blakely and others said that to make achieving the 2025 goal a "reasonably high" probability , annual 10 per cent tax rises would be needed, plus one other big new policy, such as:

Regulating the tobacco market and gradually reducing the supply of tobacco.

Reducing the levels of nicotine - the addictive component - in tobacco to very low levels.

A large reduction in the number of tobacco sellers.

Reducing the number of points of sale could be particularly effective in poor areas, research from Canterbury University suggests.

For his Master of Science thesis in 2011, geography student Christopher Bowie compared Christchurch neighbourhoods on their densities of convenience stores and supermarkets.

"Individuals living in low SES neighbourhoods," he concluded, "have greater access to commercial sources of tobacco products than those living in high SES areas ..."

Analysing Health Sponsorship Council youth smoking research, Mr Bowie also found that while all young smokers overestimated adult smoking prevalence, young smokers at schools in poorer areas over-estimated it the most. Thirty per cent of them believed half of adults were smokers, and 43 per cent put the proportion at three-quarters.

Auckland University tobacco control expert Dr Marewa Glover said these young people's beliefs would have been shaped by what they saw around them in a similar way to how Maori smoking was perpetuated.