Unemployment as measured by Stats NZ is sitting at a record low of 3.2 per cent. But how is that figure reached and how reliable is it?
Liam Dann looks at those questions as well as why it differs from other measures - like the numbers on Jobseeker benefits - and why economists still put more weight on it.
What is the Stats NZ official unemployment figure?
The Stats NZ unemployment number is based on responses to the Household Labour Force survey (HLF).
Every quarter, Stats NZ surveys 15,000 households - catching data for more than 30,000 people - to get a representative sample of Kiwis' employment status.
"The household survey allows us to ask people questions about all the aspects of their employment, if they're unemployed or how long they've been unemployed. It really allows us that in-depth insight into New Zealanders," says Malak Shafik, Stats NZ's manager for the labour market release.
"We do it as a survey because we're aiming to be representative of New Zealand."
The HLFS dates back to 1986, making it highly effective for illustrating trends in the labour market across time.
The methodology is regularly checked and adjusted against best practice so the figure can be benchmarked accurately across time and against other OECD countries.
To be classified as unemployed by Stats NZ, a person must not have a paid job, must be available to start work and must have been actively seeking work in the past four weeks or be due to start a new job in the next four weeks.
Part-time workers who want more work are captured by StatsNZ in the underutilisation rate. That is currently at 9.3 per cent - the lowest it has been in 14 years. It peaked at 15.6 per cent in 2012.
Is it accurate?
The figure is as accurate as any statistic collected and used to gauge economic progress.
It is statistically robust as a means of capturing the direction of travel for the labour market.
When we consider the sample size, bigger would always be better, says ASB senior economist Mark Smith.
But relative to political polls, which use samples as low as 1000 people, the HLS survey is comprehensive.
There is a degree of subjectivity with regards to how unemployment is classified by Stats NZ.
However, this is true of most economic statistics.
For example, Stats NZ has influence over inflation figures because it chooses which consumer goods are counted in the Consumer Price Index.
Economic growth is measured based on choices about which activities are counted towards gross domestic product.
What's important is that the methodology remains consistent and free from political interference.
Why is it different to Jobseekers benefit numbers?
Jobseeker benefit numbers sometimes diverge from the official unemployment figure based on different definitions.
For example, Stats NZ will not classify someone as unemployed if they have part-time work, even though they may be on a benefit.
But conversely, an issue with Jobseeker benefits is that you can be unemployed and looking for work but if your partner earns a lot then don't you don't qualify for a benefit, so you don't get included in those numbers.
The fact that benefit categories are subject to government policy whims makes them less reliable across time, and around the world.
That's why, though, they do consider benefit numbers in their analysis, most economists put more weight in the Stats NZ data.
However, economists stress the importance of analysing the totality of the labour force data rather than relying on just the headline figure.
That data includes figures one employment growth and underemployment.
How does the unemployment rate differ from the Jobseekers right now?
Stats NZ currently counts 94,000 Kiwis as unemployed. The latest comparable benefit figure is 100,000 (Jobseeker work ready).
"Right now the groups of people that are unemployed under HLF and those eligible for unemployment benefit are heading in the same direction," says Smith.
There is sometimes confusion around the broader Jobseeker benefit figure (180,000), which includes those who are sick and disabled after changes made by the National government in 2013.
This broader figure has risen since 2017, prompting some critics to suggest the current government has made it easier to move on to a sickness benefit.
If this was true it would again highlight the importance of the Stats NZ figure, which is not influenced by government policy changes.