National Ethnic Population Projections 2023-base

General methodology is outlined in the National Ethnic Population Projections data collection. Projection assumptions for 2023-base national ethnic population projection update Base population These projections have as a starting point the estimated resident population (ERP) of each ethnic group at 30 June 2023. The level 1 ethnic populations (‘European or Other’, Māori, Pacific, Asian, and MELAA) are directly estimated using results from: the 2023 Census of Population and Dwellings the 2023 Post-enumeration Survey (a sample survey to measure census coverage) international travel and migration data (to measure residents temporarily overseas on census night) births, deaths, and migration data (to measure population change between census night and 30 June 2023). The derivation of the 2023-base ERP is described in Estimated resident population (2023-base): At 30 June 2023 and Estimated resident population 2023: Data sources and methods. The level 2 ethnic populations (Indian, Chinese, and Samoan) at 30 June 2023 are indirectly estimated using the ratio of the level 2 preliminary ERPs at census date to the respective level 1 preliminary ERPs at census date applied to the level 1 ERP at 30 June 2023 by age-sex. The estimated resident population (ERP) is the best available measure of the number of people usually living in New Zealand. The ERP at 30 June 2023 includes estimates of uncertainty from contributing data and estimation processes, where possible. The main source of uncertainty in the ERP was the estimated net census undercount measured by the 2023 Post-enumeration Survey. However, the estimation did not incorporate uncertainty around data quality in individual sources or from adjustments made to the ERP at 30 June 2023. As not all sources of uncertainty in the ERP are measurable, some wider uncertainty in the base population has been assumed for projection purposes. The uncertainty is assumed to vary by age and sex, broadly consistent with that measured in the June 2023 ERP, and reflects the combined uncertainty from two sources: Census enumeration and processing: Coverage errors may arise from non-enumeration and mis-enumeration (for example, residents counted as visitors from overseas and vice versa), either because of deliberate or inadvertent respondent error. Errors may also arise during census processing (for example, misrecognition during scanning, uncertainty from age imputation). Adjustments in deriving population estimates: This includes the adjustments applied in deriving the ERP at 30 June of the census year (for example, net census undercount). It also includes uncertainty associated with the post-censal components of population change (births, deaths, and net migration between the census date and 30 June). Simulations of the base population are produced by drawing a random number sampled from a normal distribution with a mean of zero. For each simulation, a random number is multiplied by the assumed standard error for each age-sex then added to the base ERP. Fertility and paternity Fertility assumptions are applied to the female population of childbearing age (13–50 years) to add new birth cohorts to the population. Similarly, paternity assumptions are applied to the male population (15–54 years) to allow for births that men of a given ethnic group have with women not of that ethnic group. The fertility assumptions are formulated using birth registrations, period fertility rates, census data on ‘number of children ever born’ (including rates of childlessness), and the fertility assumptions of the National population projections: 2024(base)–2078. Period total fertility rates (TFRs) and total paternity rates (TPRs) are assumed to vary throughout the projection period. The median assumption is not an assumed trajectory, but the median of the fertility/paternity simulations, which vary every year around that median. Age-specific fertility rates (ASFRs) and age-specific paternity rates (ASPRs) are assumed to vary throughout the projection period. Under the median assumption: ASFRs generally decrease below 35 years over the projection period, and generally increase above 35 years ASPRs generally decrease below 40 years over the projection period, and generally increase above 40 years. For each ethnic group, simulations of TFRs and TPRs are produced using a simple random walk with drift model. Random errors for TFRs and TPRs are sampled from a normal distribution with a mean of zero and different standard deviations for each ethnic group: 0.048 and 0.005 for 'European or Other' 0.076 and 0.022 for Māori 0.065 and 0.014 for Asian 0.153 and 0.017 for Chinese 0.124 and 0.023 for Indian 0.112 and 0.040 for Pacific 0.162 and 0.055 for Samoan 0.149 and 0.040 for MELAA. The drift function shifts the median of the TFR and TPR simulations to follow the assumed median TFR and TPR. Median ASFRs and ASPRs are scaled to sum to the simulated TFR and TPR. The projections allow for births to parents of each ethnic group that are not registered as children of that ethnic group. Simulations of this loss factor for each ethnic group and year are produced by drawing a random number sampled from a normal distribution with different means and standard deviations based on historical data for: ‘European or Other’ – a mean of 1.7 percent and standard deviation 0.2 Asian – a mean of 1.5 percent and standard deviation 0.5 Chinese – a mean of 3.5 percent and standard deviation 0.6 Indian – a mean of 0.9 percent and standard deviation 0.6 Māori – a mean of 2.9 percent and standard deviation 0.4 Pacific – a mean of 2.5 percent and standard deviation 0.5 Samoan – a mean of 2.7 percent and standard deviation 0.4 MELAA – a mean of 5.0 percent and standard deviation 1.5. The projections then allocate births between male and female. Simulations of the sex ratio at birth for each ethnic group and year are produced by drawing a random number sampled from a normal distribution with a mean of 107.4 males per 100 females for Samoan births, 105.5 males per 100 females for other ethnic group births, and different standard deviations for each ethnic group based on historical data: 1.1 for 'European or Other' 1.6 for Asian 2.8 for Indian 2.8 for Chinese 1.6 for Māori 2.3 for Pacific 3.5 for Samoan 5.0 for MELAA. Future fertility trends are uncertain and depend on a range of factors, including: changes in population composition and different trends in subpopulations (including ethnic groups) trends in ideal family size and the strength of individual desires for children trends in the patterns of education and work, including the timing, duration, and proportion of time dedicated to those activities changing macro-level conditions (for example, government policies, childcare facilities, housing) that influence the cost of children in a broad sense changing nature and stability of partnerships, including rates of partnership formation (including re-partnering) and dissolution changing biomedical conditions (for example, female fecundity, new methods for assisted conception). Mortality Mortality assumptions are applied to each age-sex group to allow for deaths. The mortality assumptions are formulated using death registrations, period life tables, and the mortality assumptions of the National population projections: 2024(base)–2078. Under the median assumption, life expectancy at birth (e0) increases over the projection period for both males and females for all ethnic groups. The median assumption is not an assumed trajectory, but the median of the mortality simulations, which vary every year around that median. As with the national population projections, death rates change at different rates at different ages, and age-specific survivorship rates (ASSRs) are assumed to vary throughout the projection period. For each ethnic group, simulations of e0 are produced using a simple random walk with drift model. Random errors are sampled from a normal distribution with a mean of zero and the standard deviations are based on period life tables for males and females in each year. The drift function shifts the median of the e0 simulations to follow the assumed median e0. Median ASSRs are scaled to sum to the simulated e0. Mortality reductions are expected to continue in the future, although the extent of the trends is uncertain and depends on: changes in population composition and different trends in subpopulations (including ethnic groups) changes in biomedical technology, regenerative medicine, and preventative methods including monitoring, treatment, and early intervention changes in health care systems including effectiveness of public health changes in behaviour and lifestyle (for example, smoking, exercise, diet) changes in infectious diseases and resistance to antibiotics environmental change, disasters, and wars. Migration Inter-ethnic mobility (IEM) assumptions are applied to each age-sex group to allow for the net effect of people changing their ethnic identification over time. The IEM assumptions are formulated using two approaches: Net levels derived as a residual from intercensal demographic estimates. This assumes the population change between the census year population estimates, and not explained by the other components (births, deaths, migration), is due to IEM. Age-specific rates derived from the New Zealand Longitudinal Census (NZLC), with emphasis on the most-recent linked census pairs (1996–2001, 2001–2006, 2006–2013, 2013–2018, and 2018–2023). IEM rates represent the net propensity for individuals to enter or leave an ethnic group during the following year, relative to the ethnic population at the start of the year. The rates are based on linked records where ethnicity was specified in both census years. Smoothed male and female rates were broadly similar, and are assumed to be the same in the projections. In the median IEM assumption, there is an average annual net change to the population between 2023 and 2048 due to people changing their ethnic identification of: -0.06 percent for 'European or Other' +0.42 percent for Māori +0.10 percent for Asian -0.03 percent for Indian +0.15 percent for Chinese +0.08 percent a year for Pacific +0.09 percent a year for Samoan +0.12 percent a year for MELAA. The median IEM assumption is not an assumed trajectory, but the median of the IEM simulations, which vary every year around that median. For each ethnic group, simulations of IEM rates by age are produced using an autoregressive integrated moving average or ARIMA (1,0,1) with drift model. Random errors are sampled from a normal distribution with mean of zero and different parameters for each ethnic group. The drift function shifts the median of the IEM simulations to follow the assumed median IEM. Net migration assumptions are applied to each age-sex group to allow for the net effect of people changing their country of residence over time (migrant arrivals less migrant departures). Ethnicity is not collected directly in international migration data, but the migration assumptions are formulated using international migration data (including migrant arrivals and migrant departures by country of citizenship, birthplace, and age), linked administrative data (by ethnicity), observed intercensal ethnic population change, and the migration assumptions of the National population projections: 2024(base)–2078. Net migration simulations are assumed to vary throughout the projection period. The median net migration assumption gradually increases over the projection period for the Asian, Indian, Chinese, Pacific, Samoan, and MELAA ethnic populations, with an annual average net migration between 2023 and 2048 of: +2,880 for 'European or Other (including New Zealander)’ +32,420 for Asian +10,740 for Indian +7,150 for Chinese -1,500 for Māori +1,530 for Pacific +530 for Samoan +3,450 for MELAA. The median assumption is not an assumed trajectory, but the median of the migration simulations, which vary every year around that median. For each ethnic group, simulations of net migration are produced using an autoregressive integrated moving average or ARIMA (1,0,1) with drift model. Random errors are sampled from a normal distribution with mean of zero and different parameters for each ethnic group. The drift function shifts the median of the net migration simulations to follow the assumed median net migration. Net migration by age-sex is interpolated between a high and low pattern, to sum to the simulated net migration level. Future migration trends are uncertain and depend on a range of factors in source and destination countries, including: changes in immigration policy (in New Zealand and other countries) changes in the main motives for migration (for example, work, family reunification, education, asylum, retirement) changes in migration pressure in source countries (for example, population growth, economic growth) changes in the attractiveness of New Zealand as a place to live (for example, work opportunities, economic conditions, wages relative to costs and other countries, settlement and integration practices) costs of migration, including cost of travel and existence of networks and pathways that facilitate migration environmental change, disasters, pandemics, and wars. Inter-ethnic mobility Inter-ethnic mobility (IEM) assumptions are applied to each age-sex group to allow for the net effect of people changing their ethnic identification over time. The IEM assumptions are formulated using two approaches: Net levels derived as a residual from intercensal demographic estimates. This assumes the population change between the census year population estimates, and not explained by the other components (births, deaths, migration), is due to IEM. Age-specific rates derived from the New Zealand Longitudinal Census (NZLC), with emphasis on the most-recent linked census pairs (1996–2001, 2001–2006, 2006–2013, 2013–2018, and 2018–2023). IEM rates represent the net propensity for individuals to enter or leave an ethnic group during the following year, relative to the ethnic population at the start of the year. The rates are based on linked records where ethnicity was specified in both census years. Smoothed male and female rates were broadly similar, and are assumed to be the same in the projections. In the median IEM assumption, there is an average annual net change to the population between 2023 and 2048 due to people changing their ethnic identification of: -0.06 percent for 'European or Other' +0.42 percent for Māori +0.05 percent for Asian -0.03 percent for Indian +0.15 percent for Chinese +0.08 percent a year for Pacific +0.09 percent a year for Samoan +0.12 percent a year for MELAA. The median IEM assumption is not an assumed trajectory, but the median of the IEM simulations, which vary every year around that median. For each ethnic group, simulations of IEM rates by age are produced using an autoregressive integrated moving average or ARIMA (1,0,1) with drift model. Random errors are sampled from a normal distribution with mean of zero and different parameters for each ethnic group. The drift function shifts the median of the IEM simulations to follow the assumed median IEM. en-NZ