Skip to content

Immigration NZ data via eolas

Immigration New Zealand (INZ, part of MBIE) administers visa decisions and operates the Migration Data Explorer — the canonical public dataset for who's getting which visas, by which category, over time. eolas serves 16 datasets from Immigration NZ — 13 from INZ's GitHub-published .rda files (the same data behind the Migration Data Explorer Shiny app) and 3 from MBIE's monthly RSE (Recognised Seasonal Employer) statistics workbook.

If you're doing migration analysis, labour-market forecasting that depends on visa flows, or RSE-scheme research — this is the source.

What's in the catalogue

Residence

Dataset Description
immigration_residence_decisions Residence visa decisions (approved / declined / withdrawn) over time.
immigration_residence_occupations Residence approvals by stated occupation.
immigration_residence_apps_accepted Applications accepted (started processing).
immigration_residence_apps_on_hand Applications in the queue at each month-end.
immigration_returning_residence_decisions RRV (Returning Resident Visa) decisions.

Visitor

Dataset Description
immigration_visitor_decisions Visitor visa decisions by source country + visa subtype.

Work

Dataset Description
immigration_work_decisions Work visa decisions (essential skills, accredited employer, etc.).
immigration_work_occupations Work approvals by stated occupation.

Student

Dataset Description
immigration_student_decisions Student visa decisions.
immigration_student_first_time First-time student visas (excludes renewals).

Arrivals / departures

Dataset Description
immigration_arrivals_by_visa Border arrival counts segmented by visa type.
immigration_departures_by_visa Departures by visa type.
immigration_limited_purpose_decisions LPV (Limited Purpose Visa) decisions — niche but useful for specific-event tracking.

RSE (seasonal worker scheme)

The Recognised Seasonal Employer scheme brings workers from Pacific Island nations for horticulture/viticulture seasonal work. eolas serves three RSE statistics streams from MBIE's monthly published workbook:

Dataset Description
immigration_rse_onshore_by_region_sex Current RSE workers on-shore, by region × nationality × gender (snapshot).
immigration_rse_onshore_by_year_sex Same breakdown but as a time series back to 2011/12 (~4,500 rows).
immigration_rse_atr_processing_times Monthly P50 and P90 processing times for Agreement to Recruit applications. SCD2 stream — each refresh appends a new month rather than overwriting, so this dataset accumulates a real time series.

Refresh schedule

Weekly, Wednesday morning NZ time. The RDA .rda files on the joh024/migration_data_explorer_public GitHub repo update monthly (after INZ's monthly data drop); our weekly refresh catches new releases promptly. The RSE workbook also refreshes monthly.

meta = client.info("immigration_residence_decisions")
meta["last_refreshed_at"]
meta["source_last_modified_at"]

License

All Immigration NZ data is published under CC-BY 4.0. Commercial use is fine; attribution required.

Recommended attribution: "Source: Immigration NZ, served via eolas (eolas.fyi). CC-BY 4.0."

Common patterns

Residence-visa flow over time

import matplotlib.pyplot as plt

res = client.immigration("immigration_residence_decisions", start="2018-01-01")
# res has columns: Date, Count, plus several classifier factors

# Total residence approvals over time
by_date = res[res["Decision"] == "Approved"].groupby("Date")["Count"].sum().reset_index()
by_date.plot(x="Date", y="Count", title="NZ residence approvals (monthly)")
plt.show()

library(dplyr)
library(ggplot2)

res <- eolas_get_immigration("immigration_residence_decisions", start = "2018-01-01")
by_date <- res |>
  filter(Decision == "Approved") |>
  group_by(Date) |>
  summarise(Count = sum(Count))

ggplot(by_date, aes(Date, Count)) + geom_line() +
  labs(title = "NZ residence approvals (monthly)")

Top source countries for visitor visas

visitor = client.immigration("immigration_visitor_decisions", start="2022-01-01")
top = (visitor[visitor["Decision"] == "Approved"]
          .groupby("Country")["Count"].sum()
          .sort_values(ascending=False).head(10))
top.plot.barh(title="Top 10 source countries — visitor visa approvals (2022+)")

RSE workforce composition

rse = client.immigration("immigration_rse_onshore_by_region_sex")
# 156 rows: Region × Nationality × Sex × Count + Suppressed flag

# Total by nationality
by_nat = rse[~rse["Suppressed"]].groupby("Nationality")["Count"].sum().sort_values(ascending=False)
by_nat.plot.barh(title="RSE workers on-shore by source nationality")

ATR processing-time trend

This is the SCD2 stream — accumulates over time so each refresh adds new month-end data points:

atr = client.immigration("immigration_rse_atr_processing_times")
# Columns: YearMonth, Description, Applications, P50, P90

atr.plot(x="YearMonth", y=["P50", "P90"], title="ATR processing time (business days)")

Source-specific notes

  • Two upstreams, one namespace: 13 of the 16 datasets come from MBIE/INZ's GitHub-published .rda files (the Migration Data Explorer source). 3 come from MBIE's monthly RSE Excel workbook on immigration.govt.nz. Different file formats; eolas normalises both to the same schema convention.
  • Privacy suppression: cells with counts of 5 or fewer are suppressed to protect individual privacy. The eolas response has count_suppressed=true and count=null for those rows; raw "<=5" strings have been replaced. See data conventions §6 for the full rule.
  • Rounded counts: per MBIE policy, counts are rounded to base-3. Totals will not equal exact sums of cells.
  • RSE workbook has 3 sheets: Sheet 1 (current snapshot) → immigration_rse_onshore_by_region_sex; Sheet 2 (year × sex time series) → immigration_rse_onshore_by_year_sex; Sheet 3 (ATR percentiles) → immigration_rse_atr_processing_times. The ATR stream uses SCD2 because the source workbook only ever contains the current month's percentiles — appending lets us build a longitudinal record.
  • Not currently in eolas: sector data reports (health/tourism/primary industries — Cloudflare-protected parent page), Migrant Employment ZIP (large multi-CSV quarterly), refugee resettlement (PDF-only), Migration Trends 2016/17 historical XLSX (one-off snapshot). On the backlog if there's demand.

Where to find more