SA2 household projections 2018-base

Period-specific information Reference period This release constrains 2018-base household projections by statistical area 2 (SA2) areas. This is the first release of SA2 household projections, which were previously available as customised projections on request. The new projections incorporate the latest demographic information, including: SA2 population projections: 2018(base)–2048 assumed ratio of census household counts to census usually resident population count for each SA2 from 2018 census territorial authority and Auckland local board (TALB) household projections: 2018(base)–2043. These projections have the estimated households at 30 June 2018 as a base, and cover the 25 year period to 2043 at 5-year intervals. Projection Three alternative household projections (low, medium, and high growth) are available for each SA2, indicating possible outcomes given the respective alternative SA2 population projections. The medium household projection, for example, is derived by applying the assumed ratio of census household counts to census usually resident population to the medium SA2 population projection. At the time of release, we consider the medium SA2 household projection suitable for assessing future household changes, which are consistent with the medium projection of Subnational population projections: 2018(base)–2048 (released 31 March 2021), Statistical area 2 population projections: 2018(base)–2048 (released between April and December 2021), and Subnational family and household projections: 2018(base)–2043 (released 15 December 2021). However, the low and high growth projections represent plausible alternative growth scenarios. Geographic classification The household projections in this release are based on the territorial authority and statistical area 2 (SA2) classifications and boundaries at 1 January 2021. Summary of methods For each SA2, the projected population at 30 June for the base year and each 5 year period is multiplied by the assumed ratio of census household counts to census usually resident population counts (discussed in the ‘Projection assumptions' section). The medium SA2 household projections are constrained to the respective territorial authority and Auckland local board area (TALB) household projections for each period from the official subnational family and household projections. The high and low projections represent plausible alternative scenarios for each SA2, so are not constrained to any higher-level projections. Check plausibility and make manual adjustments if needed. Projection assumptions Projection assumptions are formulated after analysis of short-term and long-term historical trends, information provided by local planners, and government policy. The subnational population projections, projection assumptions, 2018(base)–2048 NZ.Stat table provides a summary of the low, medium, and high population projection assumptions for each SA2 area. More detailed subnational family and household projection assumptions are available on request ([email protected]). Base population These SA2 household projections use the estimated number of households at SA2 level at 30 June 2018 as a base. This is equivalent to the census household count, with implicit adjustments for: net census undercount households temporarily overseas on census night change between census night and 30 June census year households temporarily absent from their usual residence but elsewhere in New Zealand on census night. The estimated number of households are not directly comparable with census counts because of these adjustments. Assumed ratio of census household counts to census usually resident population The ratio of census household counts to census usually resident population counts are derived using the 2018 census data for each SA2 area. The following steps summarise how the assumed ratios for each SA2 area have been formulated: Derive the initial ratio initial_ratio = census household counts (excluding visitor-only households)/census usually resident population counts The census household counts are adjusted down for a small number of SA2s (< 30), so they are not higher than the census usually resident population counts . The adjustment is done by excluding the households coded to ‘household composition unidentifiable’ category. See census household composition standard classification for more details Apply the adjusted initial ratio to the respective SA2 population estimates at 30 June 2018 to get the initial base SA2 household estimates. The assumed ratios are then adjusted so that the SA2 estimated households within a TALB add up to the respective TALB household estimates One specific manual adjustment has been made for Christchurch Central. The initial census ratio for Christchurch Central was 4.1. Given the large number of apartments being built in the area, we manually set the ratio to 1.5 These ratios are used to derive medium projections which are assumed to stay constant throughout the projection period For the low and high projections, further adjustments have been made to the ratios, so that the sum of low and high SA2 household projections are lower and higher than the respective TALB household projections for each projection period. For example, for each SA2 in a TALB and for each 5-year period, adjusted low ratio=(initial SA2HH * sum of SA2PP within TALB/TALBPP)/SA2P, where SA2HH is SA2 household projections, SA2PP is SA2 population projections, and TALBPP is TALB population projections Final adjustments have been made to these ratios if the low, medium, and high SA2 household projections do not follow the rule of low < medium < high. The low and high projections are manually set to be 10% lower or higher than the medium projections. The final ratios are adjusted accordingly. Usage and limitation of data The projections should be used as an indication of the overall trend, rather than as exact forecasts. The projections are updated every 2–3 years to maintain their relevance and usefulness, by incorporating new information about demographic trends and developments in methods. One limitation of the overall method is that the average household size for the three alternative projections does not convey the true uncertainty. These relate a low population growth scenario to a low household growth scenario, a medium population growth scenario to a medium household growth scenario, and a high population scenario to a high household scenario. It is possible that low population growth could combine with high household growth to produce a higher average household size scenario. Or high population growth could combine with low household growth to produce a lower average household size scenario. Users are advised to use their own judgement on the suitability of these projections for their specific use. en-NZ