Data and Statistics Bill

Kia ora to the committee, and welcome back. I just wanted to make some initial comments that respond to some of the points that came up late last night, just so that they’re on record in terms of the purpose of the bill and where we go from here. The member David Seymour, who, I know, will take a call shortly, raised the question about the difference between research and official statistics, and I placed an answer on record. I’d like to put a more fulsome one down. The deceptive simplicity of his question caught me off guard at the time. So, there we go, it’s a perfectly valid and important question.

Official statistics are defined in the bill as statistics that are produced by the Statistician or a public sector agency or by an individual or organisation approved in writing by the Statistician to produce those statistics. They are the output of the organised collection, analysis, and presentation of data, usually to describe an aspect of the economy, population, society, or environment. They’re often internationally benchmarked, including measures such as GDP, labour force participation, or the price of a usual basket of goods and services. Research, in this context, describes a wider, analytical approach whereby underlying data are used in new ways to create new insight and knowledge. It often involves using data to identify trends and observations over time or across different populations, to understand their relationship between variables—classically, whether X causes Y to happen—or to evaluate programmes and policies.

I also wanted to respond to a couple of points that the member Joseph Mooney raised that I didn’t have a chance to respond to last night. He raised concerns about the Privacy Act not applying overseas—that is, essentially, a summary of the point he was making, as I understood it. While the New Zealand Privacy Act doesn’t apply to a person overseas, clause 52 requires the Statistician to take extra care when deciding whether to allow access from overseas, taking into account the laws that apply in that relevant jurisdiction—this is laid out in the Act—and the relationship between Stats NZ and the overseas person, and any means available to the Statistician to ensure that the overseas person complies with any conditions imposed in that context by the Statistician. So I think it’s important to emphasise what has already been said in the debate: that data will not be provided to the researcher; Stats NZ continues to hold the data, and it can only be accessed through a secure portal. The member also noted that there was no obligation on the researcher to delete the data. This would only make sense if the researcher held the data. But, again, I stress: the data continues to be held by Stats NZ.

There were a couple of points that Joseph Mooney raised. Another one that wasn’t addressed last night, as we ran out of time, was that the member suggested, implied, that the Statistician is able to compel other agencies to provide data for research, and this is a concern that has been more widely aired. It’s simply not true. There is no ability for the Statistician to require the provision of data for research. The mandatory collection powers only apply to collecting data for official statistics. Agencies choose whether to provide data more broadly to Stats NZ. Many agencies have done so because they have trust and confidence in Stats NZ to safeguard and protect that data appropriately. What all this means is that the agency, essentially, gets to choose whether they pass on data to Stats NZ, and they can do that in accordance with whatever agreement they’ve made with the people they’ve collected the data from. So they can honour any commitments they made to people when the data was collected, and that might be that the data is only to be used for certain purposes, and that might not include research. So that is, in that respect, the decision of the department.

One other thing, which was covered briefly in response to a concern raised by Joseph Mooney last night, was his concern about principle 6 of the United Nations Fundamental Principles of Official Statistics having no mention of research. As I stated yesterday, it’s laid out in the implementation guidelines for the UN’s Fundamental Principles of Official Statistics—the aspect that relates to research—and the implementation guidelines make it very clear that non-statistical organisations must have a legal framework setting out their professional independence, to allow for impartial and equal access to statistics, but that access to data for research may have a different threshold for access. And I quote: “A law or formal provision is in force, which specifies that statistical agencies are professionally independent and impartial, develop, produce and disseminate statistics following professional standards, and treat all users in the same way. But this does not mean that there is no distinction between different types of users. For example, only accredited researchers might have access to anonymized microdata for research, while the general public would not be allowed such access. Among categories of users, each user must be treated ‘in the same way’ in accordance with published guidelines.” That’s directly from those guidelines.

The implementation guidelines have got an entire section on how to ensure confidentiality when enabling access to data for research and, consistent with these guidelines, Part 3 of the bill specifies the conditions—we’ll come to that, obviously, when we debate Part 3—required to protect confidentiality. Part 5 of the bill clearly sets out the framework by which data held by Stats NZ may be accessed for the purposes of research, and these are provisions that have been carefully thought through over the past six years or more. This bill, we know, goes right back to Craig Foss’ time as Minister of Statistics and has been carefully and thoughtfully developed from that time on, including under the stewardship of the Hon James Shaw. So, with those few comments, I open the debate.

Before I call the member, I’d like to inform the House that we’ve received two new tabled amendments from Debbie Ngarewa-Packer. These are available on the Table and on the House papers page. If we receive new tabled amendments, I’ll advise members so that they can refresh the House papers page to see the new amendment. Finally, it would be helpful for members to ask multiple questions, if they have them, of the member in charge during their call. I call on David Seymour.

Thank you, Madam Chair. I appreciate the Minister engaging on this question, because, actually, the difference between official statistics and research seems to go to the heart of many of the concerns that have been expressed about the bill. It would seem that “official statistics” is a term that implies certain restraints and safeguards around the use of data, whereas “research” is a much more generic term, and, of course, the confusion between the two is much compounded by the official statistician’s ability to delegate his or her powers, thereby meaning that you combine the delegation of powers to other entities with the somewhat more diffuse definition of research, and suddenly the restraints around the use of data that imbue trust in the Crown’s use of it and allow the Department of Statistics to keep collecting data is suddenly eroded. And that really goes to the heart of the problem that people have raised.

While I appreciate the Minister giving a more fulsome definition of the official statistics and research, all he’s really done, in my view, is highlight what a problem and contradiction we have in clause 3(b)(i)—and, I just note for one member, it’s not 3(b) “I”, it’s 3(b) “1”. That’s what they meant, the Romans. They meant “1” when they said that. But in any case, it says, “(i) enable the trusted collection and use of data for the production of official statistics and for research;”. Now, it would seem, based on what the Minister’s given as a definition, the facts are if research is something that, by definition, is done with official statistics, then it was redundant to include the “purpose of research” in clause 3(b)(i). So if all research is just using official statistics, if official statistics have to be collected before that research can be done, then clause 3(b)(i) should just say, “collecting data for official statistics” and what happens subsequently is a separate matter. But by using the conjunction “and” and saying “and for research”, this clause implies that the purpose of the bill is to collect data for the purposes of research, separate from collecting data for the purpose of official statistics. And as soon as you bring in that confusion, the whole purpose of the bill means that we potentially have the collection of data for purposes that won’t be constrained by the usual restraints that we expect for official statistics. So, of course, the fact that there are two parts of the bill about this, Part 4 and Part 5, defining “official statistics” and defining “research”, I think brings home the point more clearly. So, in terms of a question to the Minister, would he consider amending clause 3(b)(i) to the effect that “the purpose of this Act is collect data for official statistics.”? If research happens subsequent to that, that’s another matter, but it shouldn’t have the purpose of collecting for both.

A few points just to cover off there. The Government Statistician is accountable for every decision made by a delegate. So when we’re talking about this delegation—I think it’s just important to stress this is a baseline understanding for the House. I’ll come to the member’s specific question. The bill hardwires in legal obligations regarding independence and confidentiality, which apply to any delegate, and it is a criminal offence to wrongfully obtain data while acting as a delegate of the Government Statistician—clause 79 of the bill. So there are some really clear protections, and what we’re setting up here is an ethical framework for the handling of data.

Now, Stats NZ has two internationally recognised integrated databases: the Integrated Data Infrastructure—the IDI, as it’s known—contains data about people in households, and the Longitudinal Business Database contains linked data about businesses. If the bill did not provide the Statistician the ability to collect data for research, then all the data collected from other agencies—like justice, education, training, health, benefits, and social services—could not be integrated, and that would make that database completely worthless. So it is important to have that in there.

I’m not sure I agree with the conclusion that the Minister’s made about the Integrated Data Infrastructure, but I certainly would agree with him about the importance of it. I want to just support the point that David Seymour made, briefly, and then sort of add my own clarification of what the problem is that we’re trying to solve. I think we all understand that the intention of the Government is to restrict the Government Statistician’s, and his or her agents’, powers to the collection of data for use in statistical analysis. So we have two things: the collection of the data and the use of the data. The collection of the data is restricted to statistical analysis. The data can also be used for research. Now, that’s fine, I think we’d all understand that and accept it. The problem is that’s not what the bill says, and that’s why the Council for Civil Liberties and Transparency International New Zealand are worried that there can be legislative overreach.

Even in the Minister’s first call last night—which I’ll quote back to the committee because I think it kind of highlights the confusion that we have. Dr Clark said in his first call: “Like the Act it replaces, the bill only allows Stats NZ to collect data for official statistics and research. To be clear, the Statistician can only make mandatory requests for data for official statistics, not for research.” So they can’t make mandatory requests for research, and yet that’s exactly what the purpose clause says.

So I think we’re in a situation now where we have a Government and Stats NZ knowing what they want to do and how they want to do it, within the legal boundaries, but the boundaries are too wide, because not only at clause 3(b)(i) but also—and we’ll come to this in Part 3, I’m sure, but it’s relevant to Part 1—clause 22(c) says, “if the Statistician considers the data is necessary or desirable to enable the Statistician to produce official statistics, or desirable for research”, it can collect it. And that doesn’t sound to me consistent with the Minister’s last call in response to David Seymour.

So the point, I think, both the ACT Party and the National Party are saying is: we understand what the Government wants to do—we get it—but we need to be very careful that in the black-letter law, we don’t provide an opportunity for the Statistician—and as the Council for Civil Liberties says, it is more worried about, I think, the Statistician’s agents under delegated powers to collect more data than they need on the basis that it could be helpful for research in the future. That’s not how Stats works, but the way Part 1 and Part 3 are worded, I agree with Mr Seymour.

So my question, simply, is this: have I understood the Government’s intentions right—that is, a restriction from collection of data for statistical purposes, but the use of it for both statistical analysis and research, and, if so, will the Government consider an amendment to the bill? My amendment is only to Part 3, but I think Part 1, at clause 3(b)(i), I think it is, should also be tightened to make clearer what the Minister just said, albeit it was a bit confused when he said it last night.

Thank you, Madam Chair. Just to pick up on the conversations that are taking place already, questions from Messrs Woodhouse and Seymour, and, to be fair, the Minister as well, engaging in the point regarding that distinction between official statistics, on the one hand, and research, on the other. My question is, ultimately, in a similar space, but going a bit further down the track of what it means to be collecting statistics for the purpose of research. Just looking at it from the perspective of a researcher and someone who would be undertaking academic work, to discover some great truth about New Zealand as it is now or where it’s heading and so forth, and understanding that in an academic context it’s hugely important that there are ethical considerations of the collection of data, the use of data, the publication of the data, and knowing that, in an academic context, there are very strict requirements, such that there must be ethics approval for research projects given prior to the undertaking of those projects. And, of course, in relation to the use of the data and the publication of the data, whether we’re talking from a privacy perspective or not to cause harm inadvertently to those who are providing the information or, indeed, in the way that it is used and presented.

It occurs to me to wonder how researchers might be obliged to justify the use of data—and I use that term generically—in their research when that data has been collected before the undertaking of that research, in the sense of the way that the data is collected. So it seems to be there’s a bit of a disconnect in time, as well as in relation to the purpose that it was collected, and I wonder if the Minister’s got any advice on how that could be resolved, such that a reasonable ethical oversight can be undertaken by a researcher, noting that they wouldn’t have had any hand in the collection of that data in the name of official statistics in the first place.

I say this a little cautiously: I think a lot of the conversation that we’re having now is about the intent, because the purpose is about the intent. And some of the questions that are being raised, rightly, actually will be dealt with in a more concrete way, as the former Minister Michael Woodhouse referred to, Part 3 and Part 5, and David Seymour as well, and this is a test about whether those things align, is, essentially, what I hear from members opposite. I think we’re agreed that we do want the data, we want the official statistics, and we also want the ability to research in a safe and protected way, and to ensure people’s data is protected, and that the right kinds of protections are around that, and we come to that in the bill. So I just say that for context, and I think some of the real debate and the detail is going to happen in those later parts of the bill.

But I just want to also say for the record, the purpose enables collection of data for official statistics and research. Data can—and I have covered this, but I’m just saying it in other words—be provided voluntarily for both purposes. The mandatory collection only applies to official statistics—the mandatory collection only applies to official statistics. The bill clearly provides for data to be provided voluntarily or by agreement, as well as the mandatory aspect of it.

Thank you. Look, we very rarely lodge Supplementary Order Papers (SOPs), but the gravity of this is so severe that we thought that we must. And I’d like to just capture that for a couple of minutes and then go into my questions.

We’ve put forward SOP 185 that would require that the Act be reviewed within three years, and that the Data Iwi Leaders Group (ILG) and the National Urban Māori Authority be consulted on the terms of reference of this on any other review of the legislation. A reminder, in 2019, Stats New Zealand and the Data ILG signed a Mana Ōrite Relationship Agreement to work together and develop policies to give effect to the Crown’s Te Tiriti obligations and Māori aspirations for Māori data and data sovereignty. In 2020, the Data ILG and Stats New Zealand engaged in a co-design process with the aim of designing a wider model for Māori data governance. So it’s really disappointing to see, and to hear from the Minister last night, that, in fact, that’s been ignored and the bill has progressed without incorporating any of that mahi, any of that goodwill, and the Mana Ōrite agreement, effectively, in the outcomes, has been ignored.

So the bill is being perceived by Te Ao Māori as a slap in the face and yet another example of this Government refusing to work in good faith. Actually, I’d also like to go on, at the same time, to say the expansion of the facial recognition technology, which is going on at the same time across Government—again with no engagement with tangata whenua. So we’ve also tabled two further amendments on behalf of the New Zealand Council for Civil Liberties. One would remove reference to research and delegation of powers and ensure that the data collected under the Act can only be used for statistical purposes, the other would require Government to communicate directly to each household that a census return is required and inform them to obtain a paper copy or access a computer.

So my questions to the Minister are two: what are you going to do, or what are you doing, to ensure that this bill honours Te Tiriti o Waitangi? And, secondly: what are you going to do to ensure that this bill mitigates tangata whenua’s concerns on State profiling and misuse and abuse of data and statistics as tangata whenua have endured and experienced? Kia ora.

Kia ora. A few things to say. Obviously, in terms of Part 1 of the bill, clause 4 puts Te Tiriti o Waitangi out for the first time in this legislation. This is an appropriate and timely update to actually have obligations for the Crown put into the legislation, completely absent from what went before.

I don’t accept the member’s characterisation of the relationship with the iwi leaders group and the Mana Ōrite agreement. I really do think that we have a good, honest, open conversation that has started there that simply did not exist before that Mana Ōrite agreement was signed. It’s a work in progress, as is, I think, all of the Crown’s work to better engage, incorporate, and work with Māori for better outcomes.

The process, as I said last night, the work that’s going on with the Data Iwi Leaders Group has not yet reached its conclusion. It will be appropriate, when that work is completed on governance, to revisit, to review where the bill is and what the place of Māori data governance in that will be.

Thank you. I think my question was quite clear—I know it’s early in the morning, for some. But I think what the question to the Minister was is: what is this bill; what are you going to do to mitigate? It wasn’t about whether there’s a work programme in motion or how you now feel good about the group after you’ve ignored them. It’s more about what the bill is going to do to mitigate that. I didn’t hear anything in this and we really would like to have that assurance that while you have ignored the advice and the extensive work that they’ve done for three years, there is some consideration to how you’re going to implement any of their recommendations or considerations.

Well, I don’t think there’s much point in engaging too far in name-calling and slanderous accusations on this matter. But I do think there is, and experienced, a very genuine engagement and a lot of value coming from that relationship and the improvement of the bill as a consequence. I do think that the member fundamentally misunderstands the bill. And if she thinks that removing the value of such reports like this one is a good thing, with the Supplementary Order Paper (SOP) she’s put forward on behalf of the Council for Civil Liberties, I think that’s a real shame because I do think there is huge value that can be added through the use of statistics for research into improving wellbeing for our peoples, for our Māori communities, for our Māori businesses, and for New Zealanders as a whole. So I simply don’t agree. Obviously, this is not the place to have the debate about that SOP; that comes into a later part of the Act. But I do want to put on record that I think the member fundamentally misunderstands the bill.

Before I ask my question, it’s not our job to do this but I find myself in strong agreement with the Minister on his response to Debbie Ngarewa-Packer’s question.

The Governance and Administration Committee heard from iwi data groups—one in particular that his name eludes me—and I’m pleased to support the Minister’s response, which is, “We can’t look over our shoulders.” Maybe we haven’t treated iwi with the respect they need in a statistical sense in the past. There is a remedy. I think this is very strong and they were very satisfied with the level of engagement and the protection of data as taonga. So I just want to put that on record.

But the Minister’s distinction of voluntary versus mandatory data collection, I think, was something of a red herring, because if data is provided voluntary, it’s not subject to this legislation. It’s almost a legal paradox to say, “We need a law to say you can voluntarily hand over information.” So the whole purpose of the bill is for a mandatory framework—a legal framework—for the collection, storage, and use of data.

So my simple question, and the reason this is relevant to Part 1, is the Minister pretty much rightly said, “Look, these questions of research are down in Part 3.” But the problem we’ve got is by the time we get there, if we recognise that the purpose clause is wrong, we passed Part 1. So I think it is worth dwelling on that bit a little longer. So my question to the Minister is this: when he said last night, “The Statistician can only make mandatory requests for data for official statistics, not for research.” Where in the bill is that made absolutely clear? Because it’s not clear to me.

I do think, as the member has alluded to, that that debate properly belongs in Part 3 and I respect what he’s saying. I do think, you know, we need to be agreed on the purpose and then we need to be sure when we come to Part 3 that we’ve got the right protections in place.

I, obviously, bring the bill here reasonably confident of that. I think the select committee—and the member sat on the select committee—have done a really thorough and robust job over a full select committee process of designing and developing a bill that strikes the right balance in here, that carries over what has worked and served New Zealand well over a period of time. But, of course, the devil is always in the detail and we do need to work through the specifics, I think, when we get to that part of the bill.

Look, I thank the Minister for his answer to my question a few questions and answers back. It seemed to me what he said in response to my question about whether or not it was necessary to have the word “research” in clause 3(b)(i), after he appeared to have previously said that research was done using official statistics—I made the point that maybe research was redundant as a purpose of the bill, because if you had official statistics, research could be done subsequently. The Minister gave an interesting example. He suggested that the Integrated Data Infrastructure (IDI), which a lot of people regard as an extremely useful and world-leading tool, was a good example of—well, it wasn’t clear if it was official statistics or research. But the fact that the IDI accesses data sets, administrative data, from a range of departments—education, health, welfare, and so on—and integrates them together would be an example of research that is not necessarily official statistics, or at least that’s what I thought he was saying.

So my question is: when he used that example, was he trying to say that the IDI is an example of research or is the IDI an example of official statistics? If he’s saying yes, the IDI is an example of research, then what are the constraints on the use of the IDI? If he comes back and says, “Well, all of the chief statistician’s usual requirements are placed on the IDI for the custodianship of official statistics.”, then how is it any different calling it research if all of the same constraints and protections are there—why don’t we just call it official statistics and save ourselves time, which gets us back to the original question of: is it not redundant to say the purpose of this bill is research?

Thank you, Madam Chair. Minister, I’m wanting to go back to the question that I asked you last night and your response to it, around the purpose of this Act being to enable the trusted collection and use of data for the production of official statistics and for research. That appears inconsistent with the United Nations’ Fundamental Principles of Official Statistics, namely principle 6, which says that, “Individual data collected as a statistical aid by statistical agencies for compilation are to be strictly confidential and used exclusively for statistical purposes.” No mention of statistical purpose and for research.

The Minister referred last night to the United Nations’ Fundamental Principles of Official Statistics Implementation Guidelines to justify the purpose of the Act, saying that it can be for official statistics and for research. I’m curious about where in the Implementation Guidelines that is. I’ve had a bit of a look through it. Look, I’ll note, for example, in page 14 of the Implementation Guidelines, it says that, “For example, only accredited researchers might have access to anonymized microdata for research, while the general public would not be allowed such access.” Then page 56 talks about “Statistical agencies [having] established appropriate procedures and processes (e.g. anonymization) before passing over the data to researches.”

Additionally, the use of microdata is monitored in order to immediately apply corrective actions from circumstances appear in which confidentiality is violated. On that point, I note that the Implementation Guidelines keep on talking about anonymised microdata, and there is an important distinction there that I’ll just touch on briefly. Statistics New Zealand points to personal data being de-identified rather than anonymised, which is a higher standard.

I’d be interested in the Minister’s thoughts on that, but I’m struggling to see how the implementation guidelines justify moving away from principle 6 of the United Nations’ Fundamental Principles of Official Statistics, namely that it should be used exclusively for statistical purposes—albeit with some caveats in the Implementation Guidelines for anonymised microdata. I’d appreciate the Minister clarifying that point.

I do think that the answer’s in the member’s question, because that use of that microdata is the research, right? The anonymised microdata. I guess I covered this off: it is in the guidelines. We’re just making it explicit in the bill what that is in the purpose.

So I want to come back to, though, the member David Seymour’s question. Just to be clear: the IDI is research and Part 5 of the bill applies to that data set, so we will come to a detailed debate on that. If the bill didn’t explicitly cater for mandatory and voluntary provision of data, there would be no protection of voluntary provided data that is also being canvassed in the debate. Stats New Zealand may not have a mandate to collect data by agreement were that not explicit in the bill.

The general point I’ll come back to is that in all of this, the bill hard-wires an ethical and cultural framework into the bill and that is the debate that comes up in Part 5. It’s right that we then work through whether that has been done in the right detail.

I move, That the question be now put.

The question is that the Minister’s amendment to Part 1 set out on Supplementary Order Paper 163 be agreed to.

Amendment agreed to.

The question is that the Hon Michael Woodhouse’s amendments to Part 1 set out on Supplementary Order Paper 197 be agreed to.

The question is that Debbie Ngarewa-Packer’s tabled amendments to Part 1 be agreed to.

Amendments not agreed to.

The question is that Part 1 as amended stand part.

Point of order, Madam Chair. I don’t think I’ve ever seen, in a committee stage, a presiding officer put a question after the first closure motion, while there were multiple members making a call. This is a bill that has attracted widespread and deep expert criticism. There were several more clauses that had not been debated that I know members wanted to move on to. You’ve taken a closure motion while there were people calling to add more to the debate, and I urge you to allow us to actually fulsomely debate this issue. Otherwise, I think members on this side will have to ask for the Speaker to be recalled—it’s a serious matter.

Speaking to the point of order. Thank you, Madam Chair. The member should be well aware that an ability to accept a closure motion by the presiding officer rests entirely with the presiding officer. To suggest otherwise, I think, calls for disorder of the House.

Speaking to the point of order. I guess you could put it down to inexperience but you can question a ruling without questioning that it’s their right to make it. I haven’t questioned the presiding officer’s right to make the ruling; I just think, in this instance, it’s the wrong one. When the member’s a bit more experienced, he might learn that members are actually allowed to comment on the proceedings of the House.

CHAIRPERSON (Hon Jenny Salesa): Speaking to the point of order, the Minister, the Hon Dr David Clark, and then I will be making a ruling.

Speaking to the point of order. Yep, I will leave the member’s condescension to one side, but I do think that we have had a fulsome debate, somewhat repetitious, and getting into the detail of clauses that members were interested in. There has been plenty of opportunities for new clauses to be debated but members have chosen to focus in on certain areas. We have chosen not to have closure motions on this side of the House because we also believe in a fulsome debate, but when the material does start to get repetitious, of course we reserve the right to bring a closure motion because there’s business to get on with in the House.

CHAIRPERSON (Hon Jenny Salesa): I will now make my comments, because there’s been a lot of point of orders and further points of order on this issue. I was actually on this seat last night when this bill began to be discussed, and I’ve just counted how many calls I’ve taken and how long we have actually taken to discuss just Part 1 of this bill—this is a seven-part bill. We have debated Part 1 for over two hours. So one of the reasons why I actually accepted the closure motion was that we have discussed it for a long time. It is clear to me that there is no agreement in terms of the debate here. There are substantial parts that we will be discussing in Part 3, and I’ve accepted the closure motion. We shall now move to Part 2.

Part 2 Roles and responsibilities