When Census does not go as planned

Joel Hernandez
Insights Newsletter
18 April, 2019

A shocking 700,000 individuals - or 14.3 percent of New Zealand’s population - either partially completed or did not complete the 2018 Census.

More shocking than the numbers themselves was how this information was made public. It was only under threat of parliamentary contempt that Liz MacPherson, Chief Statistician of Statistics New Zealand, finally revealed that last year’s Census was in worse condition than previously identified.

To put this bungle into perspective, the 2018 Census counted 85.7 percent of our population, while the 2013 Census counted 97.6 percent; India’s 2011 Census counted 97.7 percent of their 1.2 billion population.

The 2018 Census clearly did not go as planned.

The implications of this failure are far-reaching for research and policymaking – and will be felt by New Zealanders all around the country.

Data from the Census is used to make decisions on school decile funding and DHB funding, set electoral boundaries, and collect iwi data.

Both the Ministry of Health and the Ministry of Education have resorted to using the outdated 2013 Census to determine future funding as a result of the more than a year-long delay.

The full details of all those who were not counted in the Census have not yet been released, but Maori, Pacifica and the elderly are likely to be disproportionately represented in the missing 700,000-person cohort. Historically, this has always been the case.

This poses a significant problem for researchers in Statistics New Zealand’s Integrated Data Infrastructure (IDI) using Census data for their research. Such a large data gap could lead to biases in results, which could then lead to biased policy recommendations.

This is where the Census is going to affect New Zealanders the most – policy, whether it is school funding or DHB funding, is designed to improve the outcomes for the most disadvantaged communities in New Zealand, the very people most likely missing in the 2018 Census.

Even after Statistics New Zealand fill in the data gap using imputation from existing Ministry administrative databases, researchers can never be certain that they are analysing the correct information and therefore making the correct policy recommendations.

Economists around New Zealand are rightly calling for the next Census to be brought forward to 2021. However, it is uncertain whether Statistics New Zealand could meet this target and produce a successful Census.

New Zealand needs reliable and robust data for good evidence-based policy. A reliable and robust Census is how we get it.

Stay in the loop: Subscribe to updates