Stats NZ Rebases Labour Survey to 2023 Population

This report outlines the effect of estimated resident population (ERP) revisions on the Household Labour Force Survey (HLFS) for the September 2018-March 2025 quarters.

Key points

We have revised the historical HLFS data from the September 2018 quarter to the March 2025 quarter and investigated the effects of revised national population estimates (NPE), Māori population estimates (MPE), and subnational population estimates, on our series.

While there were substantive changes to high-level estimates, the effects on key rates were negligible at the national level.

The main effects of the revision between the September 2018 and March 2025 quarters are set out below:

• overall decrease in the working-age population, from 4,335,000 to 4,297,000 in the March 2025 quarter, with both male and female working-age populations decreasing by a similar amount
• overall increase in the Māori working-age population, from 649,700 to 658,300 in the March 2025 quarter, with the wāhine Māori working-age population increasing more than the tāne Māori working-age population
• the number of employed people decreased slightly faster than the number of people in the working-age population, leading to a downward revision in the seasonally adjusted employment rate in the March 2025 quarter, from 67.2 percent to 67.1 percent
• the working-age population was revised downward for men and women in most age groups in the March 2025 quarter, with the only upward revisions for teenagers (aged 15-19 years, men and women) and women aged 20-24 years
• all regional working-age population estimates were revised down. The largest percentage decreases were in Otago (down 2.3 percent, 5,100 people) and Southland (down 2.1 percent, 1,800 people) in the March 2025 quarter.

Background

HLFS sample data is weighted to the working-age population, using national population estimates based on the latest census and post-enumeration survey (which measures census coverage). The total population is subsequently updated each quarter using estimates of births, deaths, and net migration for the quarter. As we move further from the base census-year estimates, the level of uncertainty in our estimates is likely to increase. Rebasing allows us to use the most accurate population estimates for the HLFS.

In this rebase, we are updating our weights based on the following benchmarks:

• total population by sex by five-year age groups
• total population for 12 regional groups
• Māori ethnic group by sex for the 15-29 and 30+ age groups.

These benchmarks are the data source for the HLFS working-age population estimates and allow our HLFS estimates to align with published population totals.

The recalibration of population estimates after a census is a normal revision that occurs every five years. The revision reflects the combined inaccuracies in:

• population estimates at the start and end points, which are based on census counts and adjustments (for net census undercount and residents temporarily overseas)
• estimated components of population change (births, deaths, and net migration).

Exclusions are applied to the estimated resident population to produce benchmarks reflecting the HLFS target population. They exclude long-term residents of homes for older people, hospitals and psychiatric institutions, and people in prison. These exclusions have been updated from June 2023 onward based on 2023 Census data.

Following this rebase, all historical HLFS estimates were revised from the September 2018 quarter.

HLFS estimates for the June 2025 quarter (release date 6 August 2025) onwards will use the population estimates based on the 2023 Census, and the outcomes-based measure of migration.

Effects on the March 2025 quarter

Revisions to the NPE have caused the following revisions to the March 2025 quarter HLFS series.

The seasonally adjusted working-age population was revised down by 38,000 people (19,000 men and 19,000 women). The number of people in the labour force was revised down by 29,000, which consisted of 28,000 fewer employed people. There was no change to the number of unemployed people. The number of people not in the labour force was revised down by 9,000.

The employment rate was revised down by 0.1 percentage points to 67.1 percent, as shown in table 1.

The underutilisation rate was revised up 0.1 percentage points to 12.4 percent.

Table 1

Impact on the working-age population

The working-age population has been revised downward from the September 2018 quarter onwards, with the gap between the previously published and revised series widening over time (figure 1).

Figure 1

The revised working-age population affected both men and women to a similar degree (figure 2).

Figure 2

The Māori population estimates affected both the male and female working-age population (figure 3). The Māori working-age population estimates for men and women have been revised up by a similar degree from the September 2018 quarter to the December 2022 quarter. From the March 2023 quarter onwards, there was a larger upward revision for women. People with multiple ethnicities are included in each ethnic group they identify with.

Figure 3

The revisions affected all ethnic groups (figure 4). The Māori working-age population estimates were revised upwards for both men and women (as above), while the European and Asian working-age population estimates were revised downwards for both men and women. Proportionally, Pacific peoples were less affected by the revisions to the working-age population.

Figure 4

Certain age groups in the working-age population saw larger revisions than others (figure 5). The largest revision was a decrease in males aged 70 years and over. The largest increase was in males aged 15-19 years.

Figure 5

The working-age population was revised down across all regions in the March 2025 quarter (table 2). The largest percentage decreases were in Otago (down 2.3 percent, 5,100 people) and Southland (down 2.1 percent, 1,800 people).

Table 2

Impact on key HLFS series

Revisions to the working-age population had the following effects on the main HLFS series.

Similar to the decrease in the number of people in the working-age population, the number of people employed was revised downward (figure 6). This resulted in a downward revision of people in the labour force.

Figure 6

There were minimal changes in the number of people unemployed across all quarters, with no effect on the unemployment rate at a national level (figure 7).

Figure 7

The population revision also resulted in a small decrease in the number of people not in the labour force (figure 8).

Figure 8