Asians, Pasifika and Maori: Auckland’s integration of its minorities.

Auckland city’s population has grown 40 per cent in the last 25 years, and its growth has been driven by immigration, specifically international immigration. Ethnicity data on Auckland over the period 1991 to 2013 show the profound effects of this immigration.  Auckland has had a reputation for being ‘the largest Polynesian city’ and the Pasifika population has increased by 70 per cent over the period 1991 to 2013. But the growth in the Pasifika community is far exceeded by the growth in the Asian community, whose population has grown 600 per cent (but from a very small base).  Being only 6 per cent of Auckland’s population in 1991, Asians had grown to form 26 per cent in 2013. Meanwhile the Europeans, who were 70 per cent of the city’s population in 1991, were by 2013 just 55 per cent.

So have the immigrants been well integrated into the city’s neighbourhoods, or is there a high degree of segregation? Population and urban analysts now recognise that ethnic residential patterns are too complex to be represented by a single index. The US Census Bureau analyses ethnic residential segregation in five dimensions: evenness, exposure, concentration, centralisation and clustering. Table 1 shows the first of these, evenness. for the three leading minority groups. The index measures the percentage of a group that would have to change residence for each neighbourhood (census area unit) to have the same percentage of that group as there is over the whole city. When the index equals 1, the group in question is largely found in neighbourhoods with a fairly homogenous composition. When the index equals 0, the group is completely integrated. As we can see, Maori are becoming more evenly spread across the city, while the concentration of Asians is increasing. Pacific people exhibit the highest level of concentration or segregation, and there is virtually no change over time.

Table 1: Evenness of Distribution of Ethnic Minorities in Auckland, 1991 – 2013

Ethnic group 1991 1996 2001 2006 2013
Maori 0.3408 0.2981 0.3120 0.3130 0.2885
Pacific 0.4852 0.4752 0.4927 0.5000 0.4982
Asian 0.2540 0.2708 0.3030 0.3292 0.3347
Index: Dissimilarity Index

Exposure, the second of our dimensions of residential segregation, refers to the extent to which members of an ethnic group are exposed only to one another. It differs from the concept of evenness in taking into account the size of the group. Thus a group, if small in size, may be concentrated in a few neighbourhoods, but not be numerous enough to dominate the neighbourhood. We measure exposure with the isolation index. This index ranges from 0 (a high level of exposure to the majority ethnicity) to 1 (a low level of exposure). The values in table 2 suggest that Asian people are becoming less exposed to the majority European population. Or, to put it another way, are finding more people of their own ethnicity in their neighbourhoods.

Table 2: Exposure of Ethnic Minorities in Auckland, 1991 – 2013

Ethnic group 1991 1996 2001 2006 2013
Maori 0.1819 0.1903 0.1823 0.1755 0.1625
Pacific 0.3157 0.3456 0.3780 0.3878 0.3895
Asian 0.0832 0.1537 0.2105 0.2910 0.3479
Index: Isolation Index

Concentration, the third dimension, refers to the amount of physical space a minority group occupies. Concentration is here measured with the delta index. A 0 value means that the group has a low residential density, and a 1 means that the group has a high density. As can be seen in Table 3, Maori and Pasifika people appear to have become slightly more geographically spread out over time.

Table 3: Concentration of Ethnic Minorities in Auckland, 1991 – 2013

Ethnic group 1991 1996 2001 2006 2013
Maori 0.5063 0.4800 0.4719 0.4569 0.4684
Pacific 0.6142 0.6000 0.5945 0.5820 0.5894
Asian 0.4884 0.4833 0.4736 0.4612 0.4816
Index: Delta Index

Centralisation refers to the propensity to live in either inner or outer suburbs. The absolute centralisation index ranges from -1 (mostly found in the outer suburbs) to +1 (mostly found in the inner city). The index values for all three ethnicities are moderately positive, with Asians having the highest values. However, concentration in the inner city is not really a feature of New Zealand’s largest city, given that minority groups in the United States have values above 0.7.

Table 4: Centralisation of Ethnic Minorities in Auckland, 1991 – 2013

Ethnic group 1991 1996 2001 2006 2013
Maori 0.0847 0.0879 0.0468 0.0261 0.0016
Pacific 0.1912 0.1591 0.1211 0.0744 0.1022
Asian 0.3807 0.3719 0.3556 0.3224 0.3279
Index: Absolute Centralisation Index

Clustering is the fifth dimension. It refers to the probability that areas with high proportions of minority group members will be adjacent to each other. The spatial proximity index ranges from 0 to infinity. if the index is less than 1, this means that on average, members of minority groups live nearer to members of the majority group than to people of their own ethnic identity. If the index equals 1, this means that there is no differential clustering between minority and majority ethnic groups. If the index is greater than 1, it means that members of each group on average live closer to people of their own ethnicity than to other ethnicities. In the United States, their major ethnic minorities (African-American and Hispanic) record values above 1.2. In Table 5, the index for Pasifika people rises to exceed 1.2 over time, but for Maori and Asians, index values remain close to 1.

Table 5: Clustering of Ethnic Minorities in Auckland, 1991 – 2013

Ethnic group 1991 1996 2001 2006 2013
Maori 1.0812 1.0738 1.0886 1.1039 1.0907
Pacific 1.0878 1.1982 1.2514 1.2984 1.3011
Asian 1.0194 1.0263 1.0413 1.0616 1.0755
Index: Spatial Proximity Index

With high levels of immigration and consequent declining proportion of Europeans in Auckland, it is pertinent to ask how this has affected the ethnic mix in the poorest and wealthiest neighbourhoods. Europeans have decreased their representation in suburbs of all socio economic statuses, but not in equal measure. If we look at Auckland’s most affluent CAUs, Europeans used to be 77 per cent of residents in 2001, but by 2013 had fallen to 71 per cent (Figure 1).

But in Auckland’s poorest neighbourhoods, Europeans in the same period have dropped from 30 per cent of the total to 20 per cent (Figure 2), an even bigger drop than in the wealthy neighbourhoods.

Asians appear to have settled in suburbs of all socio economic statuses. In the affluent suburbs, Asians in 2013 made up 19 per cent of local residents, an increase matching the decrease in Europeans since 2001. Asians are also seen in the poorest suburbs; in 2001 they were 12 per cent of the people in those neighbourhoods; in 2013 they were 15 per cent.

In contrast Pasifika people are highly unequally distributed across the city. In the wealthiest suburbs, they only account for 3 or 4 per cent of residents, but in the poorest suburbs, they now (as at the last census) account for 45 per cent.

In summary, New Zealand (specifically Auckland) doesn’t appear to have urban ethnic segregation to the same degree as the United States. But there is a large difference in the spatial and socio economic distributions of the Pasifika and Asian communities.



Posted in Urban issues | Leave a comment

Household incomes stagnant in Auckland 2006-13, some neighbourhoods experience income declines

The average household income in Auckland, has been more or less static in real terms between the 2006 and 2013 censuses. (Auckland mean household income 2006: $82,966 in 2013 dollars; in 2013: $80,370). But buried within this metropolitan average, are experiences ranging from real increases to declines. Here, we look at the geographical patterns of household income changes between over this seven year period. Figure 1 shows Auckland’s census area units from the 2013 census classified by amount of change in average household income from 2006 to 2013.


It can be seen from this map that most neighbourhoods have exhibited little or no change in average household income level (within ±10 per cent either increase or decrease for the whole 7 year period). There are very few neighbourhoods having either more than a 10 per cent real increase or more than a 10 per cent real decrease.

Figure 2 shows CAUs classified into ten income levels, or deciles. Decile ten neighbourhoods are concentrated in central and eastern Auckland and the North Shore, while decile one neighbourhoods are found in southern and western suburbs. A comparison with Figure 1 suggests that there is little or no correlation between socio-economic status of neighbourhoods and amount or direction of change in average income.


A scatter plot of income change by income level confirms the tight clustering within that ± 10 per cent income change band, and also little apparent association between average household income of neighbourhood and income change 2006-2013 (Fig. 3).

Income change vs income level 2006-13

A close inspection of the data does however show some differentiation. There were only nine CAUs experiencing a real increase in neighbourhood average income of 10 per cent or more; six of them were from deciles 8,9 and 10. If you take the seventy-six CAUs experiencing any real increase (even just 1 per cent in those 7 years); half of these were also in deciles 8,9,and 10.

Similarly, of the 57 CAUs experiencing a real decrease of 10 per cent or more, 37 of them (65 per cent) were from deciles 1, 2 and 3.

Averages, as they say, conceal as much as they reveal, and care has to be taken in interpretation of these changes. shifts in average income of a neighbourhood may reflect either changing fortunes of individual households resident there, or it may reflect changes in household composition of the neighbourhood, or it may mean that the geographical area has changed.

The analysis in this article had to take into account changes in the geographical areas of CAUs. In 2006, there were 394 CAUs in the Auckland isthmus, from Kumeu in the north, to Pukekohe in the south. In 2013, the number of CAUs was now 412, because of population growth. Meshblocks were taken away from some CAUs and given to others; some entirely new CAUs were created. An example is Stonefields, a new development in the vicinity of Remuera. It was created from meshblocks formerly part of the St Johns and Mt Wellington North CAUs, and became a new decile 10 CAU in its own right. It features as the extreme outlier on the right hand side of Figure 3. It records an income increase of 52 per cent, because the average CAU household incomes of Mt Wellington North and St Johns were about $73,000 in 2006, but Stonefields in 2013 had an average household income of $110,862. The explanation for the large rise is most probably that the meshblocks now making up Stonefields had households with incomes higher than the average in the CAUs they were part of in 2006.

But the number of cases of CAUs newly created between 2006 and 2013 were in the minority. Of the 412 CAUs existing in 2013, 351 of them had existed in 2006. In other words, it was only necessary to trace 2013 CAUs back to meshblocks belonging to other CAUs in the the 2006 census, in 61 cases.

Posted in Urban issues, Wealth and Poverty | Leave a comment

Trend towards concentration of affluence in Auckland has stopped, number of affluent suburbs in Christchurch has increased.

New Zealand’s affluent suburbs are overwhelmingly concentrated in Auckland,  but a comparison of data from the 2006 and 2013 censuses shows that Auckland’s dominance has slipped somewhat.

The Magnus affluence index evaluates the standing of census area units (CAUs) based on four attributes of households: having high income, income from investments, business or from rents, or a household member having high qualifications, or a managerial or professional occupation. Each CAU is assigned an average composite score based on these four attributes, the 10 per cent of the 1800-odd CAUs in the country with the highest scores (the tenth decile) then being deemed the affluent ones.

Between 2001 and 2006, the number of decile 10 CAUs in Auckland jumped from 86 to 98. But in 2013, this count stood at 91 (Table 1). If we include rural areas adjacent to Auckland’s urban CAUs, then in 2013 Auckland had 95 decile 10 CAUs.

Table 1 Geographical distribution of decile ten affluent areas

City or area 2001 2006 2013
Auckland 86 98 91
Wellington 54 47 48
Christchurch 14 13 20
Hamilton 5 6 6
Dunedin 4 3 4
Rotorua 1
Gisborne 1
Hastings 2
Palmerston North 3 3 1
Kapiti 1
Nelson 2
Queenstown 1
Rural areas 3 3 6
Total 169 174 183

Note: area defined as census area unit.
Source: NZ Census, 2001, 2006, 2013

The number of affluent CAUs in Christchurch has leapt from 13 to 20 (or 21 if we include the rural area of West Melton) (Table 1). It is difficult to know what to attribute the change to. Interpreting change in the affluence index is complicated by the number of CAUs increasing over time (due to population growth), and by the geographical definitions of CAUs also changing over time. The Christchurch CAUs elevated to decile 10 status in 2013 are ones that had been decile 9 in 2006. possibly what has happened is that following the earthquake in 2011, some affluent households have been forced to move, and have moved into the formerly decile 9 areas, thus raising the proportion of affluent households there.This hypothesis does not however explain how smaller cities such as Gisborne or Nelson have come to now figure in decile 10, being previously absent. Factors of change in number of CAUs or of definitional change do not at first glance seem to help explain why suburbs in cities like Gisborne or Nelson should have risen in status.

Posted in Wealth and Poverty | Leave a comment

Household income inequality has dropped slightly, but is still higher than it was prior to 2000

On the subject of household income distribution, the 2013 census show there is something for both sides of the political divide. The graph below shows the trend in gross household income inequality between 1996 and 2013, as measured by the Gini coefficient, the most well known measurement.


The political right can point to the fact that household incomes are now less unequal than they had been in 2006. The Gini index shows that household income inequality dropped by 2.8 points between 2006 and 2013.1

The left can counter with the observation that income inequality since 2000, whether of individuals or of households, is higher than it had been in the 1990s, 1980s or 1970s.2 The Statistics New Zealand publication “New Zealand Now: Incomes” (1999) states Gini index figures for household market income of 0.469 in 1991 and 0.384 in 1982. (Note ‘market income’ is different from what is declared in the census, as it excludes social welfare benefits).

It has to be pointed out that shifts in income distribution are not simply a matter of whether we have a Labour-led or a National-led government. Influences on household income distribution are a complex brew of demographic change, and change in the factor distribution of income, in the remuneration of different occupations, in the distribution of employment, as well as government policies.

A second point is that the census only provides us with pre-tax income from all sources. This measure of  income misses important tax and benefit influences on household incomes. Only the household economic survey (HES) can tell us about the distribution of disposable income.

Despite the difficulties of interpretation, income and wealth inequalities are certain to be an issue this election year.

More (and better) information on income distribution can be found in Brian Easton’s Users Guide published in the NZ Sociology Journal.3 Also the Ministry of Social Development’s own report on household incomes.4

  1. The trend in the Gini index is corroborated by two other indices, the Theil index and the Mean Logarithmic deviation. Also the Ministry of Social Development’s latest report on household incomes, mentioned in my previous post. 

  2. Inequality among individuals may show different trends from inequality among households; see my posting of January 15 



Posted in Wealth and Poverty | Leave a comment

2013 census suggests income inequality continues to increase

The NZ Herald ran an article in December 2013 stating that newly released incomes data from the 2013 census indicated that since 2006, there had been falls in real terms of the median personal income in deprived parts of Auckland. Meanwhile, median personal incomes of the affluent areas had been more or less static over that time.

A calculation of the Gini inequality index for personal incomes of the national population provides some corroboration for the view that inequality, after showing a drop at the time of the 2006 census, has now increased again (Figure 1). (Although inequality of both individual and household incomes had been lower in 2006, the level of inequality then was still higher than in the preceding 30 years. Income inequality has been increasing since the 1970s).


Source: NZ Census 2013

The latest Ministry of Social Development Household Incomes Report shows trends in household incomes up to 2012. The degree of household income inequality in the 2004 – 2012 period appears to have been volatile. Their analysis of the Household Economic Survey (HES) shows that inequality of household incomes was lower in 2012 than it had been in 2004-07 or 2001. Superimposition of a trendline over the volatility suggests that the general trend since the early 2000s is for a slight decline in inequality.1

On the other hand, the MSD report shows that the degree of household income inequality in the year 2011 was higher than in 2012, 2007, 2004 or 2001.2

Household incomes data from the 2013 census will not be available for some time to come. When we are finally able to piece together evidence from the 2013 census and the HES, we may have to draw the conclusion that the general trend over the past decade has been towards more inequality.

  1. Ministry of Social Development (2013), Household Incomes in New Zealand: Trends in Indicators of inequality and hardship 1982 to 2012, Wellington, Ministry of Social Development, pp. 10, 84-88 

  2. That is, when measured with the Gini index. When measured with a ratio of the 90th and 10th percentiles, it is lower 

Posted in Wealth and Poverty | Leave a comment

Is Auckland as low density as they say it is?

Auckland’s urban limits have been much in the news recently. On the one hand, there is pressure to free up ‘greenfields’ land on the outskirts for more housing. For example, the Productivity Commission’s 2012 Housing Affordability Enquiry highlighted the central role of land value in the price of a new home, and asserted that the easing of restrictions on housing developments at the metropolitan urban limit will go a long way towards solving Auckland’s housing crisis. On the other hand, planners at the Auckland Council are proposing increased density in existing suburbs to keep Auckland within the metropolitan urban limit.

Many New Zealanders will perhaps have heard the claim that Auckland is a spread out city, occupying almost as much area as say London or Los Angeles. And many will be familiar with the adage that ‘there’s no need to move to Auckland, all you have to do is wait, and Auckland will come to you.’

So how spread out is Auckland, compared to say, Sydney? The Demographia World Urban Areas handbook (2013 edition) 1 cites the Auckland urban area as covering 544 sq km (as at 2008)  with a population of 1,310,000 (2013), and an average density of 2,400 per sq km. In contrast, the Sydney urban area is measured as 2,037 sq km (2011) with a population of 3,956,000 (2013) and an average urban density of 1,900 per sq km. Auckland is therefore rated as having a higher population density than Sydney. Demographia’s figures were used in a 2010 blog to counter claims from John Banks that Auckland was too thinly spread out to justify public transport investment.

The Wikipedia article on Auckland stated the urban area is 482.9 sq km, the population is 1,397,000 (2012) and the density is 2,900 per sq km.2 The Wikipedia article on Sydney states that its population density is 2,037 per sq km.3

Figure 1 below shows the Auckland urban zone as defined by Statistics New Zealand in 2006, stretching from Hatfields Beach in the north, to Papakura in the south, a distance of roughly 65 kms. The urban zone covers an area of 1,085 sq kms. The usually resident population of this area was 1,208,094 in the 2006 census, with a density of 1,113 per sq km.4


Below is a map of the Greater Sydney metropolitan area. The Greater Sydney area is 12,145 sq km, containing 4,627,000 people (2010), and with a population density of only 381 per sq km. The Australian Bureau of Statistics definition of the Sydney urban area (cited in the Wikipedia article referred to above) is considerably less than the area of Greater Sydney, measured as 1,788 per sq km (2006), and with a population of 3,641,421 (2006), rendering a population density of 2,037 per sq km. If we further restrict our definition Sydney to just the 25 local government areas closest to the Sydney central business district, that is with the outermost areas being Manly, Willoughby, Ryde, Parramatta, Bankstown, Hurstville and Kogarah, the aggregate of these 25 areas amounts to 575 sq km, with a distance from north to south of approximately 20 kms. It contains a population of 1,772,088 (2006), or an average density of 3,081.1 per sq km.


Now, if we remove areas in Figure 1 with a population density of less than 400 per sq km (the areas shaded yellow), the population density rises from 1,113 to 2,039 per sq km. The demarcation of 400 per sq km was chosen because this is stated to be an internationally recognised threshold density for an urban area.5 This density of 2,039 is almost exactly the same as the official figure for Sydney in the same year (2,037) and a lot less than the density figure arrived at by Demographia.

Given, among other things, that there is a five-year discrepancy between the population as numerator and square kilometres as denominator of the density figure in Demographia’s estimate, it is not clear that Auckland’s population density is greater than Sydney’s.




  2. Wikipedia Auckland accessed 20 November 2013. The density figure appears to be the quotient of 1,397,300 / 482.9 = 2,893.6. 

  3. Wikipedia Geography of Sydney accessed 20 November 2013. 

  4. Statistics New Zealand (2007), Demographic Trends 2006, Table 6.03. 

  5. Demographia (2013) Demographia World Urban Areas (World Agglomerations) 9th annual edition, p. 2. 

Posted in Urban issues | Leave a comment

Europeans dominate Auckland’s affluent areas

Census-based affluence and deprivation indices provide us with a snapshot of the distribution of Auckland’s ethnic groups across the city’s poorest and wealthiest neighbourhoods. Maori and Pacific people constitute only 9 per cent of the population of the most affluent areas, but 58 per cent of the population in the most deprived neighbourhoods.

On the other hand, European individuals constitute roughly three quarters of the population of Auckland’s affluent areas, but only a quarter in the most deprived areas (Figure 1).


Source: NZ Census, 2006; Magnus Affluence Index, Magnus Consulting Ltd; NZDep2006, Wellington School of Medicine and PHI Unit, Ministry of Health.

Posted in Wealth and Poverty | Leave a comment

Auckland has the most affluent suburbs and households, but Wellington is the more affluent city.

The 2001 and 2006 censuses show that affluence is disproportionately concentrated in New Zealand’s largest cities. While Auckland and Wellington had 33.2 per cent and 9.1 per cent respectively of the national population in 2006, half (56%) of New Zealand’s decile ten census area units (CAUs) from the Magnus Affluence Index were in Auckland, with another 27 per cent in Wellington.

Auckland may have the largest share of the nation’s affluent suburbs, but more of Wellington is in the affluent class than is the case in Auckland. In Auckland, 40 per cent of its CAUs are classed as affluent but in Wellington it is 54 per cent. The difference is even more stark if we take the top fifth (quintile) of CAUs from the Magnus Affluence Index. In Auckland, 60 per cent of its neighbourhoods are in the top quintile, but in Wellington it is 89 per cent (Figure 1).

Auck & Wgton Affluence

Posted in Wealth and Poverty | Leave a comment