Source: MIL-OSI Submissions
Net migration 48,000 in December 2018 year – Media release
15 February 2019
The provisional estimate of annual net migration in the year ended December 2018 was 48,300 (± 1,800), Stats NZ said today.
Migrant arrivals were provisionally estimated at 145,800 (± 1,700) and migrant departures at 97,500 (± 1,400).
“Provisional migration estimates are a timely measure of migration, which we revise each month until they are finalised after 16 months,” population insights senior manager Brooke Theyers said.
Net migration for the year ended November 2018 is now estimated at 48,000 (± 1,200), compared with 43,400 (± 1,500) when it was first released last month. This is driven by changes in estimated migrant departures, from 100,600 (± 1,200) to 96,200 (± 1,000) for the year ended November 2018. Migrant arrivals remained relatively unchanged.
Why estimates now change
As new data becomes available, the model has more information about the border crossings it is trying to estimate. So, with an extra month of data available, this causes shifts in the estimated number of migrant arrivals, and departures, thus leading to a change in the net migration numbers.
For example, the extra data can contain information indicating the departure of travellers who were still in the country, but left in the extra month, or people who were away that have since returned.
“The revisions show the challenge of estimating migration independent of passenger cards. With nearly 14 million border crossings in 2018, small differences in classifying travellers to short-term or long-term can affect the provisional estimates.”
Compared with total border crossings, the number of migrants is very small. Of every 50 people crossing our border, typically 49 are short-term movements and only 1 is a migrant arriving or departing.
Of the 14 million border crossings in the December 2018 year, 81 percent of border crossings are currently classified with certainty. The remaining 19 percent represent 2.6 million border crossings, so a small change can affect the migration estimates.
The migration estimates become more certain after each subsequent month. In the December month, 1 in 4 arrivals are classified with certainty. This increases to 9 in 10 after four months. Therefore we expect the monthly revisions to become relatively small after about five months, as we can calculate the duration of stay/absence more definitively (according to the 12/16-month rule).
The provisional estimates have 95 percent confidence intervals (±) alongside them – the wider the interval, the greater the uncertainty about the estimate. However, these intervals reflect the model uncertainty, not the extent of future revisions to provisional data.
Our model will evolve as we get more data. A key next step is to take the new series further back in time to quantify the variability in the provisional estimates.
This release is the second to publish official provisional migration estimates using outcomes data – after the departure card was removed.
Text alternative for figure 1, Migration estimates (thousands) by direction, rolling annual, year ended December 2001 to year ended December 2018
The new series in figure 1 shows the final estimates from May 2015 to August 2017, and provisional estimates from September 2017 to December 2018, for the outcomes-based measure of migration. An experimental outcomes-based series is shown from December 2001 to June 2017 to give a longer time series.
The outcomes-based migration measure is a more accurate measure of migration than the previously used intentions-based permanent and long-term (PLT) measure, which concluded in October 2018 with the removal of the departure card. This is because the outcomes measure reflects actual – not intended – durations of stay/absence. The greater accuracy is also supported by historical population data.
International migration uses new official measure compares the two migration measures. At times the new outcomes measure is lower than the old intentions-based measure, at other times it is higher, and sometimes there is no significant difference.
Net migration totals 270,000 over last five years
The last five years – 2014 to 2018 – had the largest net migration gains ever in New Zealand’s history with an estimated 270,000 more migrant arrivals than migrant departures. An estimated 700,000 migrants arrived and 430,000 migrants departed over this period.
Over the five years, most migrants arrived on work, visitor, or student visas. However, by definition, they stayed for at least 12 months after extending their visa or transitioning to other visa types, including residence visas.
“Even though many migrants arriving only stay for a year or two, it’s important to count them as migrants and not short-term visitors. They are part of our resident population, which has implications for infrastructure and service provision,” Mrs Theyers said.
Net migration down from peak
Annual net migration has gradually fallen from the record peak of 63,900 in the year ended July 2016. This reflects an increase in migrants leaving New Zealand over this period – in particular, non-NZ citizen departures.
Annual net migration for the December 2018 year (48,300 (± 1,800)) is estimated to be lower than a year earlier, when it was 52,700 (± 200).
Text alternative for figure 2, Estimated migration (mean estimate), year ended December 2018
Country of citizenship
In the December 2018 year, migrant arrivals were mostly citizens from Asia or Oceania. New Zealand citizens were the largest group, with 36,500 (± 600) migrant arrivals, followed by China with 15,700 (± 400) arrivals, and India with 14,400 (± 200) arrivals.
Migrant departures in the December 2018 year were mostly citizens from Oceania or Asia. New Zealand citizens were the largest group, with 44,700 (± 900) migrant departures, followed by:
- China – 10,500 (± 400) departures
- United Kingdom – 5,900 (± 200)
- India – 5,400 (± 200)
- Australia – 4,000 (± 300).
There was a net loss of 8,200 (± 900) New Zealand citizens in the December 2018 year – more left the country long-term than returned. The annual average since 2001 is a net loss of around 18,000 New Zealand citizens. The 2011 year had the biggest net loss of New Zealand citizens (36,200) over this period.
Experimental series and new series
Since May 2017 we have been publishing migration estimates using an outcomes-based approach. These monthly estimates are available from January 2001 to June 2017; they are no longer being produced. They are described here as an ‘experimental series’.
From January 2019, monthly migration estimates are published using the new migration system. This includes a statistical model to produce provisional migration estimates shortly after each reference month. The method to produce final migration estimates has also been enhanced by, for example, using day of travel rather than month of travel. These new migration estimates are available from June 2014 to December 2018 and are described here as ‘new series’.
The new series estimates will be extended back to 2001 during 2019. In the meantime, the experimental series is comparable to the new series and is suitable for longer historical comparisons. Note: there are some level shifts between the two estimates series.
The latest migration estimates for the December 2018 quarter will be used in the national population estimates being released on 18 February 2019. Population estimates back to the September 2013 quarter will be revised in the near future, but no date has been set. These will incorporate new migration estimates and 2018 Census data; however, the revisions for each of these may be done separately. We will advise the timetable for these revisions when it’s available.
See Outcomes versus intentions: Measuring migration based on travel histories.
See Migration Data Transformation or email email@example.com for more information.
Text alternative for figure 1
Three time-series line graphs show migration estimates for migrant arrivals, migrant departures, and net migration, from rolling annual years ended December 2001 to December 2018 – for experimental series and new series.
The graphs show final estimates from May 2015 to August 2017, and provisional estimates from September 2017 to December 2018, for the new series outcomes-based measure of migration. An experimental outcomes-based series for December 2001 to June 2017 gives a longer time series. The new series follows from the experimental series and shows similar ups and downs for migrant arrivals, migrant departures, and net migration.
Text alternative for figure 2
Diagram shows arrivals of non-NZ citizens were 109,200 (± 1,500) and departures were 52,800 (± 900), making a net migration gain of 56,400 (± 1,500) non-NZ citizens.
Arrivals of NZ citizens were 36,500 (± 600) and departures were 44,700 up (± 900), making a net migration loss of 8,200 (± 900) NZ citizens.
Result is total net migration of 48,300 (± 1,800).
Note: The provisional estimates have 95 percent confidence intervals (±) alongside them – the wider the interval, the greater the uncertainty about the estimate. However, these intervals reflect the model uncertainty, not the extent of future revisions to provisional data.