Public data can show New Zealand's youth jobs problem. It cannot show a firm its own pipeline.

In its Insight 131 analysis, NZIER warned that youth unemployment in New Zealand had reached levels last seen around the Global Financial Crisis, despite a broader recovery.

The figures bear that out. What no public source can do is tell a firm whether its own entry-level intake will give it the people it needs in ten years.

That second question is narrower than the headline, and it is the one a board actually holds.

We set out to test whether a diagnostic could show a New Zealand firm, by sector and size, whether its intake will sustain its future skills and succession needs. We traced the public evidence stage by stage, from hiring entrants through training, retention and progression to succession cover, using official statistics, regulator data, programme evaluations and what large employers publish. The test was simple: at each stage, could public data read the pipeline by sector, by firm size, and for an individual firm?

The pressure is real, and it falls on entry. In the June 2026 quarter, unemployment was 25.3% among 15–19-year-olds and 3.5% among those aged 25 and over. Hiring of teenagers fell 30.5% between March 2023 and March 2025, against 21.9% for all ages. The pipeline's largest measured leak comes next: five-year apprentice completion fell from 58% in 2019 to 40% in 2025. Then the public record stops. No public source we found measures progression or succession at any resolution, apart from a few large employers that report voluntarily. Training records carry no employer size. National employment projections ran only to 2026.

What public data does well

Sector data describe intake and turnover reasonably well, and size data proved better than we first expected. Firms with 1–19 employees hold 36.4% of teenagers' jobs, against 28.4% of all jobs. Sector differences are large and consistent across independent official series. And a diagnostic built on a firm's own records, read against that public context, is achievable now. The firm already holds most of what it needs: its estimate of future need, how many entrants stay, how many skilled roles it fills from inside, and how much it buys from outside.

Two things surprised us

Our view, for what it is worth

A reliable diagnostic built on public evidence is premature, and we doubt that will change quickly: the bodies and surveys that might have filled the gaps were disestablished, paused or allowed to expire between 2024 and 2026. A transparent diagnostic built on a firm's own data is a different matter. Cutting an entry pathway is quick. After the GFC, apprentice numbers took about eleven years to regain their 2008 level. To us, that makes entry-level intake look more like a capability decision than a cost line. Three questions a board could ask before deciding:

These are questions, not advice. The decision, and the judgement about what the answers mean for a particular firm, stays with the board.

Our findings cover the public record as at 12 October 2026. If it misses something you know, such as an employer series, a survey or a dataset we did not find, we would like to hear it.