This post is inspired by the conversations Miekes, Chief Data Officer, has been having with Whetu, a third year engineering student. It led us to the hilarious website Spurious Correlations — where Tyler Vigen correlates completely unrelated data on a single graph. We love it, but it highlights how some things might look related on the surface but need to be considered in context.

“If I showed you a chart where ice cream sales and drowning rates spike at the same time, does that mean ice cream is dangerous?”

Correlation means two things change together, while causation means one thing directly causes the other.

Our brains get fooled because we are pattern-seeking and prefer simple stories — so when events line up in time or trend together, we instinctively assume cause and effect, even when a hidden factor or coincidence explains the relationship. In this example you could reason that Ice cream sales and drowning rates both rise in summer because of a third factor (hot weather and swimming), not because eating ice cream causes drowning, which is why interpreting charts requires thinking carefully about context and underlying drivers.

“How often do people see two things moving together in data and assume one is causing the other?”

Very often — this is one of the most common mistakes in interpreting data.

Humans are wired to look for patterns and simple cause-and-effect stories, so when two trends move together we frequently infer causation unless we deliberately slow down and check for alternative explanations.

“What’s an example of a correlation in New Zealand data that people might misread?”

The matching increase in the number of road cones and traffic delays could certainly be misrepresented as cones causing congestion. But most people are sensible enough to understand that they both appear as a result of roadworks or other major events.

Or, more seriously, the correlation between rising house prices and increased homelessness or emergency housing use, which people may misread as house prices alone causing homelessness.

But in reality, data shows this relationship is influenced by multiple factors such as income growth, population change, housing supply, and welfare settings, not a single direct cause.

“So when I see a headline that says ‘X linked to Y’ — should I be suspicious?”

Yes — a healthy dose of curiosity is the right response when we’re presented with anything as “fact” these days. Especially as misinformation and AI integrate deeper into everything we interact with.

When a headline says “X linked to Y” it’s often saying “these two things move together,” not “X caused Y,” so take a beat, look for context or alternative explanations, and remember that even good data needs careful interpretation. Especially important is to check the source of the data. Even when you do a quick google these days you are not given a list of sources to investigate yourself, you are given an AI generated answer presented as a fact. But in reality there is so much nuance to what exact survey or data source most accurately answers any given question that taking it at face, or LLM value, puts you at risk of perpetuating that cycle of misinformation.

Causation and correlation — what is it? was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.

“Wait — that can’t be right! I know lots of people who don’t have jobs.”

You’re not alone! It may sound counterintuitive, but that number doesn’t mean everyone else is working.

In fact, the unemployment rate counts just one specific group of people:

1. Don’t have a job, and
2. Are available to start working, and
3. Have been actively looking for work in the past 4 weeks.

So if someone isn’t actively looking, or can’t start work immediately, they don’t count in that number.

Who’s not counted?

There are lots of people without jobs who aren’t counted as unemployed, like students, stay-at-home parents, retired folks etc.

These people are considered “not in the labour force” — not employed, but also not “unemployed” by the official rules.

What else can we look at?

To get a fuller picture of joblessness in New Zealand, we recommend using some of the measures below.

Employment rate

The lower this rate is, the larger the proportion of people without a job.

Underutilisation rate

This includes: Unemployed people, people working part-time who want more hours, and people who could work, but aren’t currently job-hunting.

NEET Rate

This means: Not in Employment, Education or Training.
This is used to measure young people (ages 15–24) who aren’t doing any of those three things.

In conclusion

The unemployment rate is not a perfect measure. But it’s simple and useful for spotting big changes.
The main takeaway is to dig deeper and find other stats that give you a better sense of how the job market is actually doing!

Try using the keywords jobs or unemployment on Figure.NZ.

Why is the unemployment rate lower than you may think was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.

Inflation is an increase in the average price of a range of goods and services. It means our money loses its purchasing power over time. Prices don’t tend to come down again (though the rate of price increases may slow), rather our income — ideally — increases to maintain our quality of life. However, these two things don’t tend to happen at the same time which results in ebbs and flows of how people feel about the economy and their cost of living.

Things that impact national inflation figures can include seasonality, demand, and import costs. Inflation impacts people and households differently based on income levels, the value of savings and investments, and the importance of adjusting budgets to account for rising costs of essentials like food and fuel.

Most people have had a shock at the supermarket checkout or fuel pump recently (let’s not mention those electricity bills!) — let’s use data about New Zealand to have a look at what might be contributing 💸

With the high inflation rate visible here in 2022/2023 it’s easy to see why so many kiwi are feeling the pinch. Household utilities are the main driver of inflation figures — this includes electricity ⚡️

We often feel the impact most with our regular spends — such as petrol and groceries 🚗 ⛽ 🍓🍞

Some items also show very strong seasonal prices — such as tomatoes being much cheaper during their growing season when the supply is much more abundant 🍅

Have incomes kept up? Well, it’s a much steadier growth rate

Thinking back — Did you feel the impact of those higher inflation rates? Are you still feeling it now? What items do you notice it in the most? 🥚☕️🥑

That cost how much?! was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.

Welcome to the Figure.NZ May update.

This month:

• retail fuel charts are being updated weekly
• seasonality explained
• look ahead to Budget 2026 through data
• Tertiary language education charts
• Average life expectancy in Aotearoa New Zealand

Retail fuel charts

With the ongoing uncertainty around the retail cost of fuel, we’re now updating fuel charts as soon as MBIE releases the latest data.

Previously these charts were only updated every 6 months. But with fuel prices changing so frequently, we know timely, visualised data is more valuable than ever for understanding trends and tracking changes over time.

You can sign up to our weekly (ish) fuel chart update email and receive the latest data straight to your inbox.

A look at seasonality

As the seasons change, Chief Data Officer Andrea explains seasonality over on the blog. We’re leaving behind the abundance of tomatoes and heading towards increased power consumption.

Tomatoes, bond lodgements and power usage have really clear seasonal trends, as we have long-term data for them. Other seasonal trends may be present, but the data hasn’t been collected for long enough to show a clear seasonal trend.

It’s possible your work or interest area has seasonal data — great excuse to have poke around the site and see if you can uncover some new evidence for making decisions.

Budget 2026

Budget 2026 is set to be released on Thursday 28 May. It’s an important date on the calendar, as it reveals the government’s spending plan for the next 12 months.

We’ll be sharing a range of charts and information in the two weeks leading up to Budget Day. If you want to get a head-start, take a look at the chart series based on Treasury’s government spending data from 1997–2025.

Te reo Māori update

Data released by the Ministry of Education shows an increase in formal learning of te reo Māori over the last decade. And it’s not just locals choosing to learn the reo — international students are picking up the wero (challenge), too.

The full set of charts for the Tertiary Education Participation: Courses in Te Reo Māori 2025 dataset is now available online, after being updated in March this year.

Life expectancy data

According the data, average life expectancy in Aotearoa has risen steadily over time, reflecting improvements in healthcare, living standards, and public health. However, the gains are’t shared equally across all groups or regions tracked by Stats NZ.

Our recent chart updates for life expectancy cover Māori, Pacific peoples and Asian populations across the regions.

Figure.NZ May update was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.

He was Figure.NZ’s original Chair back in the day, hosting the start-up (then known as Wiki New Zealand) in his studio as it found its feet. He departed in 2019 and returned to the Board in 2025.

We know you were part of the origin story of Figure.NZ, hosting our Founder Lillian in your office and providing support as the first Chair. Having left in 2019, what called you back?

Haha, actually Lillian (and Vic) called me back. They were talking about how availability of data was no longer the issue: we all are (or can be, whenever we wish) awash with data. But the ability to navigate and thrive using data is as challenging as it ever was — maybe more so given all the slop that blocks the arteries of information now.

You’ve mentioned that you see right now as another significant moment in time for data. Can you talk a bit more about that?

Yeah. In 2012 we were seeing the emergence of a tribalistic “post-truth / alternative facts” vibe, driven by social media echo chambers. But we never imagined it would get this bad. The one bright spot seems to be that the situation has become so dire that the notion of trustworthy sources can become fashionable again. If we can ride that wave and double down on providing compelling, accessible data without an agenda, that would be a massive asset for Aotearoa.

How do you see your role as interim CEO? Is there anything you’d like to achieve?

I think we often ask too much of CEOs. They need to be strategic geniuses, operationally excellent, loveable people leaders, and builders of new external relationships. If I can work with the team to chart our direction in a world that looks a lot different to 2019 (not to mention 2012) and put in place some systems and processes to keep us on course, then someone can come in and help us get to the next level.

You wear a lot of hats outside of Figure.NZ — how do you balance all the demands on your time? Do you have any advice for others who do the same?

The Hitch-Hikers Guide to the Galaxy beat out the older Encyclopaedia Galactica for two reasons: “First, it is slightly cheaper; and secondly it has the words DON’T PANIC inscribed in large friendly letters on its cover.” I think that’s really good advice. Panic sucks up a lot of time and energy, and things usually look a lot more manageable in the morning. Empowering exceptional people to do their things helps a heck of a lot, too.

Is there a segment of data that you find particularly interesting? Or something that Figure.NZ publishes you think more people should know about?

There’s a lot of talk about New Zealand having a productivity problem. Figure.NZ charts from across all different sectors show boom times for productivity growth in the nineties. If policy settings contributed, what were those settings? If that was capturing an “internet dividend” what allowed that to happen, and how can we get a payoff from this decade’s tech revolutions? I’d like to see a more sophisticated debate about that, taking into account the productivity gains across Left and Right administrations, and across public and private sectors.

What would your zombie apocalypse skills be?

I’d probably be the guy leading chants and breathing exercises as the hordes closed in on us, trying to awaken from the nightmare.

(Re)Introducing Baruch ter Wal — interim CEO was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.

As we contend with the ongoing impact of the Iran war on fuel prices, increased access to this data is a top priority at Figure.NZ.

While MBIE publishes fuel price data weekly, in the past we’ve updated the relevant charts about every six months. This gives a useful overview of trends longer-term. However, we recognise that current events demand more reactive updates so alongside these standard charts we will now be releasing weekly updates.

Weekly premium petrol 95R prices, levies and taxes in New Zealand.

To save you checking in to see if we’ve done the mahi, you can sign up to a dedicated email list and we’ll send you the charts as soon as updated them.

This does depend a bit on when MBIE updates the data, so it may not be on the same day each week.

These charts are free to use for personal or professional purposes — all you need to do is link back to Figure.NZ where appropriate.

Keen? Sign up now!

Weekly fuel chart updates was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.

Tash: It’s true, The West Wing lives rent-free in my head, and I can almost always find a reference point with an episode for a current policy issue. In this case, it’s Season 1 Episode 6 — Mr Willis of Ohio. (yes, we acknowledge there are plenty of flaws in this TV show, as with all media!)

In the wise words of CJ Cregg, ‘I have to say that the census is starting to sound to me like it’s…well, important.’

For sure it’s important — and it’s changing to be based on something called ‘admin data’.

Miekes:

The New Zealand Census is evolving… so what is “admin data”?

Administrative data refers to information that government agencies collect in the course of their day-to-day business. It could include birth and death registrations, tax records, immigration data, school enrolments, and health system data. The data is not gathered through surveys, it already exists because people interact with government systems as part of daily life.

Tash:

So, what you’re saying is that the forms I fill out result in admin data — that makes sense. That data is gathered by various government agencies for a variety of reasons, like making sure my kid is a real person (by issuing a birth certificate) and I pay the right tax rate.

But….why? I love filling out the census forms!

Miekes:

How the New Zealand Census is evolving through the use of admin data

Not everyone loves forms as much as you, Tash! With costs going up, response rates going down, and quality of data getting impacted, New Zealand’s looking at different ways to collect census information.

So, we’re moving towards an “admin data first” model. Where administrative data provides the core population counts and household information which is then supplemented by smaller traditional surveys to fill gaps and perform quality assurance checks on the admin data.

Estimated median age in New Zealand

Figure from Stats NZ

Tash:

Okay, fine, I guess one less form to fill in means less life admin. You know, this brings me back to The West Wing. Aside from the fact it’s based on the American census from 1999, they talk about how many people are missed from the census due to where they live, and how expensive it is to run a ‘head count’ census. They talk about sampling rather than admin data, and how it’s more accurate. Is that right?

Miekes:

The benefits of the admin data first model

Compared to the traditional approach of asking everyone in the country to fill out a paper or online form for a specific date an admin data census model can offer more complete coverage, greater accuracy and consistency, lower cost, reduced respondent burden, and more frequent updates.

There is also the opportunity to use the smaller household survey to collect more detailed data about specific topics that couldn’t previously be squeezed into the full enumeration census.

Average Weekly household expenditure on rent in New Zealand — data from the Household Income and Living Survey is one example of this type of data that can be collected outside of the regular census.

Photo by Ketan Kumawat

Tash:

As a data lover, and knowing how valuable the census data is, I’m still wondering about what might be missed.

Miekes:

Yes, there are some cons

Because a lot of the data which will need to be used is not collected for the purpose of a census there needs to be some work to either change how it is collected or processed to make it fit for purpose. There are also some areas where the government does not collect data administratively (such as ethnicity and iwi affiliation, religion, and disability data) which is why a household survey will still be required, just at a much smaller scale.

Most common religious affiliations in New Zealand — which is a regular source for media stories about how many people identify themselves as Jedi or Church of the Flying Spaghetti Monster.

Tash:

I guess we have to wait and see what happens. One of my concerns that we haven’t mentioned is how government agencies will connect all this data to form a full picture of our country. It’s reassuring to see this mentioned in one of the Cabinet papers about the issue:

“The paper also seeks approval to amend the Data and Statistics Act 2022 to reflect the change in approach. This includes changes to allow for the full census to be taken in 2030, rather than in 2028 as currently required, to provide time for cross-government action to improve admin data.” — Modernising the census by taking an administrative data-first approach: Approval of policy decisions

Back to The West Wing, Toby Ziegler jokes that the logic of this approach means maybe voting should be based on a poll of about 1100 people with a 3 person margin of error — and it’s a valid point. No solution is perfect and will always present issues. You can be sure Figure.NZ will keep a close eye on how a different census structure delivers New Zealand’s data.

You can read more about the new approach to the Census on the Stats NZ website: https://www.stats.govt.nz/census/modernising-the-census/.

What is ‘admin data’? was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.

Our team and our Trust will miss her, but we know she will always be a champion for open data, and for democratisation of publicly available data for public good.

We wish Ngapera well for the future, and of course to remain friends and data nerds forever.

Ngā manaakitanga,

The Trustees of Figure.NZ

Tenā koutou,

After nine incredible years with Figure.NZ, working alongside the most intelligent and ethical people in the world, I’ve made the decision to resign as Chief Executive of Figure.NZ Trust and look for new opportunities.

When I first joined in 2017, our goal was simple but ambitious: make New Zealand’s data easy to find and, use so better decisions could be made for our communities and our future.
Our mission? “Get the people of Aotearoa using data to thrive”. Well we certainly achieved that goal.
Since then we have reached over 10 million people with our data, with thousands of people participating in our Community Data Series over the years. I am so proud of what we’ve achieved.

It’s been a privilege to help grow a platform and a movement that helps put people at the centre of the data by making it more accessible. Over the past nine years, we’ve built something truly special — a trusted home for data in Aotearoa, used by students, journalists, researchers, policymakers, businesses, and everyday New Zealanders who simply want to understand the world around them a little better. I helped design and launch platforms for Maori, Pacific, Disability, Youth, and Ethnic data, and taught thousands of people how to find and use data safely. I’ve had the privilege of watching our country progress with open data, helping agencies and organisations safely open and share data, whilst guiding others on how to protect their personally identifiable data with the many advisory roles and boards I have ended up on. Data leads us to many places, but always at the centre of this work, is using data for good, and in ways that can help our country.

None of this would have been possible without the extraordinary people behind the work. I’m deeply grateful to the Figure.NZ team (past and present), our trustees, partners, funders, and the many champions of open, understandable data who have supported the mission along the way. A special acknowledgement to our long term partners Internet NZ and Stats NZ. It’s been an honor and a privilege to work alongside the clever people in your organisations over the years, and I genuinely thank you all for believing in our work. And of course to our Founder Lillian Grace, I’m so proud to have steered your waka this far, and will pass it over in good faith to the next Navigator.

After nearly a decade, it feels like the right time for new leadership to guide Figure.NZ into its next chapter. I leave incredibly proud of what we’ve built together and optimistic about what comes next for the organisation. I remind everyone that the democratisation of data is now more important than ever for the benefit of our country, and the world. To help build public trust and confidence in data, the public needs to be able to trust it. Which is why a trusted, independent source like Figure.NZ Trust can help.
We exist only for public good, and only partner with others who genuinely care about New Zealand data too.

Thank you to everyone who has been part of the journey so far, I’m going to take a little break and see what doors open for the future. All I know is, I want to continue to make an impact for our future, and I am looking forward to joining a new mission. I hope to see you all soon!

Ngā mihi nui,
Ngapera

A graceful transition — farewell Ngapera was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.

As with median and mean, Tash and Andrea have different approaches to communicating. Pointing out the difference is important in communicating data, so we’ve pointed out our different sections here, too.

Tash:

Median and mean are at the heart of presenting statistics, and we often see comments from people saying ‘ugh someone needs to explain to [person] what the difference between median and mean are!’.

We wholeheartedly agree, so at long last — here is the official Figure.NZ word on the concepts.

MEDIAN

Median represents what is right in the middle. Think of a median strip — that bit in the middle of the road. A median number is the same — it sits right in the middle of what is on either side.

The thing with a median strip though, is that it can be anything. It could be grass, it could be painted, it could be a raised platform for people to wait on before crossing the road. Even if it’s made of roading materials, its only relationship with what is on either side of it is that it’s in the middle. It’s not really representative of what’s either side. Except the middle.

The same goes for data. A median number only tells you one thing — it’s the middle. It doesn’t really give you any information about the size of the numbers on either side.

Andrea:

Median is useful when extreme values would distort the picture if you used a mean. One example is income. If you look at the chart showing median weekly in New Zealand, you see what sits right in the middle of all earners.

A small number of people earning very high incomes do not pull this number up. That makes median a better choice when you want to understand what a typical person makes.

Tash:

MEAN

Mean always struck me as a strange thing to call an average. We could be mean and say, well, if you’re mean it means you’re average. Sorry.

To be fair, the mean has worked a bit harder to tell you the data story. It has taken all the data from a set of numbers, added them up and divided them by the number of data items you have.

For example, in my immediate whanau, we are aged 9, 13, 16, 42 and 44. The median in this set is 16. Whew, what a young family! But the mean tells another story. The mean is 24.8 — so not that young after all. Although youth is a relative statement…

Andrea:

It is more common to see means used to visualise scientific measurements, such as those about the environment and health. Figure.NZ has a chart presenting the mean systolic blood pressure by age bracket and gender.

It shows that, on average, blood pressure tends to go up as we get older. This is a useful insight in clinical settings as it provides patients with a benchmark that is relevant to their demographic group.

Tash:

There are clear differences between median and mean, which is why we need to be clear when communicating which one we’re using when presenting data. Figure.NZ charts will always state which is being presented, so you can trust what you’re looking at.

Median v Mean was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.

Welcome to our last newsletter for 2025. We hope the end of year period is treating you well. It’s been another 12 months of challenging times, and we appreciate you taking the time to read about what we’ve been up to.

This month:

• A note from Ngapera, our CEO
• Survey results
• Our holiday habits

What a year 2025 has been!

Recently I had my 8 year anniversary with Figure.NZ, and I had time to reflect on what a transformative time it has been in the NZ data eco system.

In those eight years, Figure.NZ’s data has reached over 10 million people.

We’ve launched 6 new platforms:

• business figures
• Maori
• Pacific
• Disability
• Youth
• Ethnic communities

to help scale our impact and reach, and we have held hundreds of workshops educating people on how to find and use NZ Data.

2025 has been a tough year for the public sector, social services and many charities. But we remain committed. Figure.NZwill be right here ensuring NZs data is accessible and trusted. We will continue guiding everyday users how to use this wealth of information to improve the lives of NZers.

At the time of sending, our people have updated 28,025 chart instances in 2025.

We process updates of existing datasets at a rate of around 16 tables per week. That’s a massive amount of data we’ve made accessible, for free, to New Zealanders.

A notable thank you to our core partner and funder Stats NZ — we could not have done this work without their support.

Also to Internet NZ who have supported our mission for over 10 years to help use the internet and data for good. Thank you.

Have a wonderful and safe Christmas and holiday season from me and the team at Figure.NZ.

Email Survey Results

Thank you to everyone who responded to our survey last month.
As you can see in the chart above, you want to hear about latest data releases most of all.

Reflecting on your feedback, here’s what we’ll focus on sending you in future:

• Monthly list of data releases
• Interesting data stories
• Examples of real-life use of Figure.NZ data

We’ll still let you know about upcoming data releases and events. Our processes will feature on the blog section of the website.

You can also expect to see weekly data releases shared on Facebook and LinkedIn.

Search functionality on the website is a noted issue — and one we’d love to fix. In the meantime, we’ll post some tips and tricks for improving search outcomes. Look out for these in early 2026.

Our holiday habits

If you’re off camping or staying at a camp ground holiday park, chances are you’ll be there for 2 nights:

As a nation, the trend shows more people are choosing to stay at holiday parks and campgrounds — although the data is only from 2020 so may be a gradual return to previous numbers.

Wherever you travel this summer, or even if you are working through or are having a staycation, have a safe and restful break.

We’re looking forward to sharing more data and stories with you in 2026.

Figure.NZ December update was originally published in It Figures on Medium, where people are continuing the conversation by highlighting and responding to this story.