Digging out information about warrantless searches has been a drawn-out process. Two years ago, Stuff asked for a breakdown of warrantless searches by which section of the act was used, and by ethnicity.

A reply came back, with a stack of numbers, and this comment from Detective Inspector Craig Scott: “Please note that Police do not hold data on ethnicity in the context of your request.”

We asked for more information about those numbers. The reply: Police, apparently, could not give a geographical breakdown of the numbers initially provided, but they could with a whole new set of data.

It seemed to be hard to analyse what was happening in much detail.

So, instead, we asked for details of every warrantless search carried out in 2018 and 2019, including where those searches happened, to see if there was a way of matching it with population data.

Police would not release exact locations of each search, citing privacy reasons, nor which searches resulted in an arrest, saying that information did not exist.

Other stations which were associated with hundreds of searches were Wellington, Hastings, Whangārei, Timaru, and Hamilton Central. Four of the top 10 were in Auckland: Henderson, Manurewa, Ōtāhuhu and Mt Wellington.

Aside from the increase in the overall number of searches between those two years (90.9 per cent), the other big change was a dramatic rise in searches where police had not even recorded which station the officer came from.

In 2018, about 7 per cent of the searches listed “unknown” for station, district and area. By 2019, that was 24 per cent - it was unclear where more than 2200 warrantless searches had even happened.

Unknown station searches in 2019 can, alone, explain half of the increase between years.

Asked why this would be the case, police said formal notifications are recorded only in relation to the officer who carried out the search, and that the information is subject to change “due to internal movements in our workforce”. 

As to why “unknown” would have increased so much, Wilson said: “There is no known reason that such errors would have increased.”

A blurry picture was emerging, with lots of missing and confusing information, making it hard to tell who the powers were being used against and whether certain communities were being targeted.

John Tamihere describes the gaps in the data as a “big hole that needs to be closed”.

“You've got to count the numbers: the numbers of kids you've turned over, numbers of cars you've turned over, what time you turn them over, and where you turn them over. And all that information has got to be logged.

Roimata Smail, a lawyer who has taken claims to the Waitangi Tribunal about the Crown’s treatment of Māori, cannot understand how there can be so many holes in the data, especially about ethnicity.

“It's not just relevant for Māori, but for all New Zealanders because we believe that our systems work properly, but if the information isn't collected, there's no way for anyone to check that.”

It is, however, particularly important to be able to examine whether there is a disparity in the way Māori and Pākehā are treated, Smail says.

“There's growing awareness that actually something's off there,” she says. “If they're keeping that information a secret by not recording it, that's sort of something that is damaging to all of us because they're doing those things on our behalf, but they're hiding what they're doing from us.

“And we're believing that they're going off and doing things right and there's not undue focus on one group. But if they're hiding that information, we don't know - they might be doing things really, really badly that we would really disagree with.

“I think most New Zealanders wouldn't want to find out warrantless searches are being done on a particular group without good reason.”

Stuff’s search for data went on - and eventually, made a surprising discovery. But in the meantime, it was time to speak to some people about their experiences with police.

He’s just 16 but when asked if he’s been searched before, this Henderson school-leaver immediately responds: “Too many times.” 

The last time was just last week, though, he says, he could see why that time: he was a passenger in a stolen car. So he admits he’s no angel, and he can see that the police have a job to do.

But there’s one time, when he was just 15, that still really gets him.

It was before Christmas last year, he and a friend, who was 17, had caught the train to Mt Albert to go visit a friend. They were just leaving the station when a police car pulled up in front of them, and three officers got out.

“They said, ‘What are you boys up to’? We said, ‘Nothing’.”

And they weren’t, he says. They were on their way to visit a friend.

The officers told him there had been some crime in the area and they wanted to talk to them. But they didn’t just talk to them, he says. There was a search.

“They went through my pockets, inside my pants, lifted up my shirt, and then they made me take off my shoes.

“We kept arguing, ‘We haven’t done anything’.”

In the end, after about half an hour, the police let them carry on, and the boys walked off to their friend’s house, he says. But they were angry and frustrated.

Look, says the 16-year-old, he gets it, sometimes: in his neighbourhood, he’s known as a troublemaker. But this was different. 

“It was a s... thing to do because they didn’t know us, they just came up to us because we were brown boys.”

Why should he be searched?

The teenager is not alone in feeling targeted by police. 

One morning last September, Raki (who asked that only his surname be used) had just finished a regular morning loop through Cornwall Park, part of a commitment he has made to keep up his fitness in middle age.

After the walk, he returned to the car, his sister’s 1995 Honda Civic, an old car, sure, but with an up-to-date warrant and registration.

“I took off my shoes, put my jandals on, and was just having a bit of a rest and a bottle of water,” says Raki.

He noticed a car slow down as it passed by him. “I was like, ‘Oh, it’s the cops. OK’.”

The police car turned around and pulled up behind Raki’s. In his rearview mirror, Raki could see the officer looking in his direction and using the radio.

“I wasn’t concerned, so I just sat there, and then I was ready to go so I started driving down the hill.”

Then he noticed the police car approaching rapidly. 

“I saw him zooming up behind me with his lights flashing and sirens, and I was like: ‘Oh, what’s going on here?’ I actually thought he was trying to go around me.”

The officer wasn’t trying to go around Raki. The officer wanted Raki to stop.

So he did, pulling into a carpark by a children’s playground.

“He actually parked right behind me, making sure I couldn’t get out.”

Raki had no idea why he would have been stopped. So, when the officer came to his window, he asked.

“I don’t understand - have I done something wrong? Because I can't see any reason for you stopping me.

“And then he looks at me and he goes, ‘Oh, mate, there's been a few break-ins of cars up around there’.”

Raki was shocked. He could only think of one reason the police would have singled him out from all the people in the park at that time: his brown skin.

He said to the officer: “So you're saying that because I'm an Islander driving an old car - what do you think? I'm going to steal cars, break into cars?”

“I was really upset because I couldn't understand why he stopped me. But then I thought about it and I knew he pretty much stereo-typed me.”

Raki’s feelings are an echo of what Māori and Pasifika have complained about for decades.

Lawyer Kingi Snelgar, who works in Manukau, says many of his clients have weekly interactions with police.

“It's not an uncommon occurrence they're stopped for things like warrant and registration checks,” says Snelgar. “You don't really get those kinds of assumptions for Pākehā people, not in the same way. Pākehā don't get stopped and asked about being suspicious.”

Do the police stop and search Māori and Pasifika disproportionately? It’s apparently not something that’s easily checked: remember, we’d been told in that 2018 Official Information Act response the data does not exist.

It seemed odd. You can find ethnicity data for all sorts of other interactions in the criminal justice system - from imprisonment rates, to who gets diversion - so why not for warrantless searches? 

John Tamihere thinks he knows why.

“They know what they're doing and they know they are targeting Māori and Pacific Island communities a lot greater.”

Roimata Smail has encountered the problem of a lack of data before in her Waitangi Tribunal claims.

“Across all of our cases where we're looking at differential treatment of Māori and non-Māori, there always seems to be information that you would have thought that government departments should be holding and it's just not there,” she says. “I don't know about motive but there's certainly a pattern.”

Snelgar believes there should be publicly-available data to track how police are using their powers - powers which he points out are discretionary: it comes down to officers in the heat of the moment making a decision about whether to search.

“And I think there is a huge difference across the police force about the way those powers are executed. 

“You have some really good police officers, and some that just seem to be exercising their discretion to the fullest extent against Māori.”

With the dataset from police of all 2018-2019 searches, Stuff was able to look at where searches were being used in relation to Census data about ethnicity.

To get a measure of where there might be disparities, we introduced crime data too. Was there any correlation between reported crime and warrantless searches?

Our analysis showed some stations serving communities with a higher share of Māori and Pasifika residents were more likely to see higher numbers of warrantless searches when compared with stations of similar crime rates.

This was particularly the case when examining drug crime and drug-related searches (sections 19 and 20).

Henderson, an area of Auckland that represents the largest estimated community served by a single police station, was an outlier.

A vastly disproportionate number of warrantless searches have been carried out there, despite the area being proportionally one of the 10 safest in New Zealand.

Henderson is home to a considerable contingent of minorities.

Snelgar says data like that backs up his belief some communities, particularly those with higher Māori and Pasifika populations, are being targeted unfairly.

“You don't really see police officers roaming Parnell or Herne Bay stopping young people and saying: ‘What are you up to today?’ Whereas, in Manukau that happens.”

If police are patrolling some areas and not others, they’re bound to find more wrong-doing.

“In areas like Parnell or St Heliers, if you have a young person the same age as my clients, I would suspect that they probably not even once a year might see a police officer in their community, let alone have an interaction with them. And I think people forget that there is over-policing in our communities like Manukau and Manurewa.”

John Tamihere agrees: “We have targeted police operations in poor areas. If you over-regulate, over-police, you will find fault.”

Out west, he says, people see it all the time.

“And we see that cars of brown boys get picked up and white boys seem to drive past. You have to ask the question: why did you pick up the brown boys without any due cause and you went on a fishing expedition and then the white boys were allowed to drive past? What made you believe that they're better than these boys?”

He says he’s not arguing all Māori and Pasifika young people are “angels”.

In Auckland’s west, further west even than the unofficial westie capital of Henderson, is the suburb of Rānui.

A strip of main shops is surrounded by houses squeezed beneath the Waitākere Ranges. In the middle of the township is a train station and a local primary school. More than 80 per cent of the students identify as Māori or Pasifika.

Sit down with youths from the suburb and they’re almost perplexed why you would bother talking to them about being stopped by the police: for them and their whānau, it’s not an uncommon occurrence. So why the fuss? 

One, an 18-year-old tells us he’s been stopped about six times, often just while out walking with friends. The day before we spoke he was pulled over after doing a u-turn on his way to a course in the morning.

The police car had been following him for a while, when he realised he’d missed a turn and had to go back to get to a short-cut he knew.

“They turned the lights on, and I thought: ‘Oh, s...’ They thought I was up to something sketchy.”

The 10-minute conversation with the officer went nowhere in the end, but it did hold him up, making him late for his course. And it added to his perceptions of why he gets stopped by police.

“Probably the way I dress, or the way I walk or something - or my colour.”

A 15-year-old from Rānui, too, tells us he’s been stopped about “six or seven times”.

The feeling of being targeted just because of the colour of a person’s skin, and assumptions being made, was tested in the Court of Appeal in 2016.

The case traced back to late one Saturday night, in central Auckland, when a police officer called out to a man walking back to a car parked down a side alley.

Eventually, a warrantless search of the car ensued after police allegedly spotted a craft knife, and 1.5 grams of cocaine was found.

The lawyers taking the case, Maria Pecotic and Golriz Ghahraman, argued that the only reason the police initially paid any attention to the man and his associate in the car was that he was brown. The officer approached the man whom he described as a “large Polynesian male”, not because there was a legitimate investigative or law enforcement basis, Pecotic argued.

Pecotic says cases like this are hard to win. But in this one “there was just this really uncomfortable feeling” that a “brown boy” going to his car was what started the whole thing off. And that wasn’t right.

“It was just really, really interesting how the police attention was drawn to this young fella in the first place,” says Pecotic.

In court, police denied that race was the trigger for the search. 

The case was sent back to the District Court, for the hearing of more evidence around the initial contact, and the search itself, and eventually the case was dropped. 

Nelson lawyer Steven Zindel, a spokesman for the Criminal Bar Association, points out that human rights law prevents racial profiling. But that doesn’t mean it doesn’t happen. It’s just how things are worded.

When questioned about why they chose to stop someone, no officer will directly identify race.

“Policemen will swear in the witness box: ‘I didn’t stop him because he was Māori. It was just because I was concerned about his erratic behaviour’. Who notices erratic behaviour? A middle-class person looking a bit erratic might not be stopped.”

Snelgar, a former Crown prosecutor, has taken cases trying to have evidence gathered through warrantless searches thrown out, examining the reason for the initial interaction between his client and police.

“We're trying to pull the layers back,” he says. “Like often you'll hear police say the reason was there was some suspicious behavior. But when you pull the layers back about what is it that was suspicious, for me anyway, there is this underlying current of race being a ground for warrantless searches.”

Other words that get used are “shifty” or “jumpy”, he says. “But when you focus on what are the characteristics of the person besides that, it's often that they are young brown people.”

Every time he’s challenged a search, police have strongly opposed. But Snelgar remains convinced that, as an institution, the police culture is biased against Māori and Pasifika.

“I think that unconsciously there's still an assessment made by the police that young brown people are more likely to be committing crimes and of course that's a prohibited ground under the Human Rights Act.”

After months of reporting and analysis of the data obtained from Police under the Official Information Act, Stuff got the opportunity to interview the commissioner. Coster, who took over the job in April during the coronavirus lockdown, had had a busy day: a police dog had been shot in Northland, with officers shooting the alleged offender; and staff at headquarters were working through the Royal Commission’s report into the mosque terror attack. 

Policing is never easy.

Nevertheless, Coster says he appreciates the scrutiny of this Stuff investigation, though “hopefully coverage can reflect the complexity of the issue because there’s a lot of positional stuff that’s very simple, on both sides actually, but the truth is in the middle and needs to be understood”.

He defends his staff, as you’d expect: “Our people turn up to do a good job and they don’t turn up to deliberately disadvantage particular groups of people.”

But - and there is a big but - he acknowledges there is a problem: the system.

“We need to acknowledge that practice evolves over time and it doesn’t always evolve appropriately over time. And we need to be really open to challenge that our policing practices are appropriate,” Coster says.

“I guess at some level you have to accept the proposition that the system is getting very poor outcomes for some people, particularly Māori, and therefore the system is not operating appropriately. [So] what needs to shift?”

We talk about Henderson and Manurewa, areas that stand out, places where people in the community feel they are the victims of “fishing expeditions”.

Coster says he “acknowledges those feelings”. 

But he says that both of those areas are high crime areas, with weapons particularly an issue. “That doesn’t mean we’re out of the woods: it just means that when we’re analysing this stuff you have to really look at the local context.”

To the issue of areas being targeted - Snelgar’s point that you don’t see the same policing effort in Parnell or St Heliers as you do in Manukau, for instance - Coster says, look, yes, there is a risk of confirmation bias, and it’s an issue police around the world are aware of.

“But we need to be reminded that we don’t generate all the searches by proactive policing activity. There’s also a decent chunk of them which are connected with calls to service where police are called to a location because an assault has occurred or someone has been observed dealing drugs or whatever.”

It’s what he means when he talks about the “drivers” of searches. “Unpicking” what was behind a search, and then taking into account the possibility of crime in other areas not being detected because police aren’t looking in that direction means “it’s not a straight-forward situation to address”.

During the interview, Coster discloses something unexpected. It’s a complete turn-around from what we had previously been told.

Since March 2019, he says, police have in fact been able to analyse ethnicity data. Data analysts have discovered that the change in the reporting system means police are now able to cross-reference search data with other databases.

“We do have nine months of data we can share with you, from 2019.”

There are still some issues with the data, he says, and “there’s not 100 per cent reliability, but it will give you a broad outline anyway”.

When it lands, it sure does - the disparities are stark and immediately obvious.

Across the country, about 40 per cent of the warrantless searches carried out were of Māori, and 41 per cent were of Pākehā.

Those figures are significantly out of whack with population data: Māori make up just 16 per cent of the population, Pākehā about 70 per cent.

In other words, there is a 24 point difference between the proportion of searches of Māori compared to the proportion of Māori in the country.

In some parts of the country, that disparity is significantly worse. Excluding locations with very few searches, at 41 out of 284 stations - 14.5 per cent - the difference between searches and population is worse than the national 24 point difference.

More than 130 stations have a higher proportion of warrantless searches on Māori than the proportion of Māori in the community that station serves.

Stations that stand out include: Wairoa, Ōpōtiki, Flaxmere, Tūrangi, Gisborne, Whakatāne, Rotorua, Kamo, Te Puke, Lower Hutt, Havelock North, Johnsonville, Wellington Central and Auckland Central.

It’s the same for Pasifika (who make up only 8 per cent of the country’s population) in 26 stations, including Avondale, Onehunga, Glen Innes, Balmoral, Beachlands and Newmarket. For 56 stations around the country, Pasifika people are more likely to be searched than Pākehā. 

There is only one station in the country (Dunedin North) where Māori have been searched at a rate lower than their proportion of that community’s population.

An almost exactly opposite picture applies when examining the figures from a Pākehā perspective: almost all stations carry out warrantless searches of NZ European people at a rate lower than their population in the community.

There are 40 stations around the country where Māori are five times more likely to be subject to a warrantless search than Pākehā.

If we exclude stations with less than 100 searches, there isn't a station where Māori are less likely to be searched than Pākehā.

No matter which way you cut the figures, Coster’s assessment that the raw data is “appalling” rings true.