Stats NZ scraps survey gathering key child poverty data

Stats NZ has scrapped its groundbreaking Living in Aotearoa survey, as it looks to find efficiencies.

The survey was set up in response to the former Labour-led government’s child poverty reduction plan.

Though data for nine of the measures of child poverty, set down in the 2018 Child Poverty Reduction Act, can be gathered from the existing Household Economic Survey, the tenth measure of persistent poverty requires longitudinal data.

This led to the establishment of the country’s largest panel survey, which began gathering longitudinal data from families in 2022.

The survey also collected information on how employment, housing and natural disasters have an impact on people’s wellbeing.

But after just two rounds of annual interviews, Stats NZ has decided to drop the survey, amid efforts to cut costs by an agency facing increasing financial pressures.

This has left experts questioning what will replace the first-of-its-kind national survey. It also raises questions about this Government’s focus on targeting spending based on robust data, and tracking outcomes.

In a quiet, Friday afternoon update on its website, Stats NZ noted the survey would not be continuing.

In his letter to survey participants, Government Statistician Mark Sowden said the survey was gathering important information “for communities, charity groups and businesses across Aotearoa New Zealand, and for us too”.

He continued: “As part of our work to ensure we’re operating as efficiently as possible, Stats NZ has been reviewing the collection and production of household living cost statistics.”

Sowden said he had made the decision to “consolidate” surveys, and instead “enhance” Stats NZ’s Household Economic Survey, which only required participants to be interviewed once, and using information already collected by other government agencies.

‘The unsustainable nature of the survey led to my decision to discontinue the Living in Aotearoa survey‘

Mark Sowden, Government Statistician

Stats NZ chief executive and Government Statistician Mark Sowden told Newsroom in a statement that since implementing the survey, the agency had faced “significant and ongoing challenges in data collection, participant retention, as well as increasing costs”.

Aotearoa was not alone in facing these challenges, as people became more time-poor and mobile, it was harder to connect with people, and retention rates had fallen. 

These problems have been well-traversed in regard to the high costs and low uptake of the census. In the latest census, the blowout was about $37m, pushing the total above $300m.

Sowden said these matters were exacerbated by the longitudinal survey model.

Despite these costs and challenges, other countries – including Australia, Germany, the UK and the US – all continue to gather longitudinal data.

Sowden said Stats NZ was working to streamline and modernise the way it worked to ensure the ongoing sustainability of robust statistics for people and communities across New Zealand.

As part of that work, the agency had been reviewing the collection and production of child poverty statistics.  

“The unsustainable nature of the survey led to my decision to discontinue the Living in Aotearoa survey,” Sowden said.

The Household Economic Survey – whose participants were only interviewed once – didn’t face the same retention problems as the Living in Aotearoa survey, he said, adding that the Household Economic Survey would be enhanced by adding and updating questions.

‘In short, fewer or noisy data will lead to less effective and efficient policies, exacerbate inequities, and cost more in the long run‘

Kate Prickett, Victoria University of Wellington family sociologist and demographer

This would be used in conjunction with administrative data already collected by other agencies to ensure Stats NZ was producing all of the child poverty measurements required.

“I can assure New Zealanders that Stats NZ remains fully committed to providing quality child poverty data and statistics to meet our obligations to report on child poverty under the Child Poverty Reduction Act 2018.”

Victoria University of Wellington family sociologist and demographer Kate Prickett said the Living in Aotearoa survey was designed to collect better information on wellbeing, and enabled the government to measure persistent income poverty and persistent material hardship.

Without the longitudinal data, it would be difficult for the Government to establish whether policies were making a causal difference for people’s wellbeing.

Longitudinal surveys collected information from the same people at different times (in this case, annually), making it possible to track changes to their wellbeing before and after a policy or service was provided, Prickett said. 

Cross-sectional data such as the Household Economic Survey didn’t have the ability to track changes in the same way. 

“The better our data are, the better we can make decisions,” she said.

Less precise data, or data that missed parts of the population, led to less confidence that government resources were getting to the right people and making a real difference.

“In short, fewer or noisy data will lead to less effective and efficient policies, exacerbate inequities, and cost more in the long run. In this way, it’s important to think of the data we collect as infrastructure. It needs continued investment and, if we don’t, we’ll end up paying more for it in the long run.”

The decision to drop the survey comes as Stats NZ faces increasing cost pressures. 

Quality data collection has become increasingly expensive, and also Finance Minister Nicola Willis has directed Stats NZ to cut 7.5 percent from its baseline spending.

Since the child poverty measures came into law, Stats NZ has received on average $20.7 million in funding each year to carry out the surveys required for measuring child poverty. 

However, Stats NZ told Newsroom the actual costs for the longitudinal survey were substantially higher than anticipated and would have required significant additional ongoing funding.

There is also a cost associated with not gathering the most robust data.

This Government has grounded itself in the promise to set targets, measure outcomes, and create policy based on evidence and data. But it is hard to take a targeted, social investment approach to policies such as child poverty reduction without the most robust data.

Prickett acknowledged that surveys like this were expensive, but that reflected the high-quality, robust and equitable data collection efforts.

“One may argue, though, that the cost of this data collection is minimal compared to costs of leaky sieve policies – one’s that are poorly targeted and that we can’t tell are providing the wellbeing and associated fiscal benefits we expect because we’ve failed to collect robust information to inform or evaluate our policy decisions.”

Without this longitudinal data, Stats NZ might be able to “cobble together” something from administrative data. But this would likely miss an important portion of Kiwis, Prickett said.

Though the bulk of Kiwis will respond to a survey or census with little extra input from interviewers, a small – but important – group will take a lot of effort to engage.

Interviewers will walk across the countryside and build relationships and trust to ensure they capture data from New Zealand’s most vulnerable people and children.

Māori communities, or hard-to-reach communities such as gangs, are often hesitant to share their information and personal data with the state. And those who are homeless, transient, or don’t have an income, are harder to track through other datasets such as tax data.

These are the people the Living in Aotearoa survey was designed to capture.

“Maybe we can get a persistent income poverty measure from the administrative data we have, but it will be noisy, imprecise, and miss our most vulnerable families – presumably those ones who are more likely to be living in persistent poverty,” Prickett said.

“It means our persistent poverty numbers will look rosier than they actually are.”

Though Stats NZ would find a way to measure persistent poverty – because the legislation demanded it – the lack of transparency and consultation about what would replace the Living in Aotearoa survey led to speculation about the Government’s plan to measure persistent poverty using the current data resources, she said.

“And right now, those other data resources do not appear to be in a position to give us a good and robust measure of poverty persistence.”

Labour Party statistics spokesperson Reuben Davidson said the decision to stop this survey suggested measuring child poverty was not a priority for this Government.

“And that, I think, sends concerning signals around how committed the Government is to address an issue, if they don’t even want to meaningfully measure it.”

Stopping the longitudinal survey in favour of using the Household Economic Survey was like taking a photo but not developing it fully, Davidson said.

“You’ll get a sense of what the image is, but you won’t get the ongoing picture of whether … some of our most vulnerable New Zealanders … are moving out of poverty and back into poverty; whether they’re moving out of poverty; or whether they’re staying in poverty.”

Child Poverty Reduction Minister Louise Upston said it was the role of the Government Statistician to decide how Stats NZ collected and produced statistics, and Sowden had assured her the agency remained fully committed to meeting its obligations to report on child poverty under the Child Poverty Reduction Act.

“The coalition Government is committed to doing everything it can to turn around the worsening child poverty rates that we inherited from the previous government, and having good quality data is a key part of that work,” Upston said.

Statistics Minister Andrew Bayly has been approached for comment.