New Zealand needs to talk about using private data to combat covid-19

Tom Barraclough
17 min readMar 30, 2020

--

Tom Barraclough

People in New Zealand must consider how best to make use of private data for contact tracing purposes, whether by government or non-government actors. I believe civil society must initiate and lead this discussion to ensure that it is conducted in an effective and justifiable way. This post is an attempt to move that civil society discussion forward.

We face an unprecedented problem in a unique context

The pandemic situation we face is incredibly serious. It is so serious that our leaders, while trying to urge calm and maintain order, have not hesitated to inform us that covid-19 could lead to thousands of deaths and an overwhelmed medical system, justifying extraordinary restrictive measures New Zealand-wide.

At the same time, the solution to minimising the risk of transmission, death and overloading the health system is relatively simple: exercise strict hygiene practices and stay away from each other. As the New York Times put it, in relation to America:

If it were possible to wave a magic wand and make all Americans freeze in place for 14 days while sitting six feet apart, epidemiologists say, the whole epidemic would sputter to a halt. The virus would die out on every contaminated surface and, because almost everyone shows symptoms within two weeks, it would be evident who was infected.

Tracking who is infected and who they have have contact with is essential for containing the disease. Containment is essential because of the risk that health systems will be overwhelmed by critically ill patients:

Supporting critically ill patients — those in multisystem organ failure — requires ventilator support, dialysis, and one-to-one or sometimes even two-to-one nursing staff. It takes only a few such cases to stretch an ICU and its staff, together with allied disciplines, such as respiratory therapists, to their limits, or past them.

On 23 March 2020 the Prime Minister announced New Zealand would move into full Level 4 lockdown for the following reason:

If community transmission takes off in New Zealand the number of cases will double every five days. If that happens unchecked, our health system will be inundated, and tens of thousands New Zealanders will die. … Right now we have a window of opportunity to break the chain of community transmission — to contain the virus — to stop it multiplying and to protect New Zealanders from the worst. Our plan is simple. We can stop the spread by staying at home and reducing contact.

Contact tracing capacity in New Zealand

Contact tracing is a practice of identifying and warning people who have come into contact with an infected person so that they can follow self-isolation protocols that limit further spread.

The message from experts is alarmingly clear. Dr Ayesha Verrall, an epidemiologist at the University of Otago, told Radio New Zealand.

“We certainly need to ramp up our staffing on the ground in public health units; our capacity in central call centres; and we also need to urgently look at technology that can make contact tracing almost instantaneous.” … “We are nowhere near where we need to be” ... “We need to be coming out of lockdown with the capability to trace 1000 cases of contact a day.”

When it comes to contact tracing, Dr Verrall’s comments to Radio New Zealand suggest our current capacity allows approximately 50 cases per day to be traced. The technology for conducting this tracing, as I understand it, is a team of qualified humans making telephone calls, and possibly other means). It is clear that the Ministry of Health is investigating how to escalate its contact-tracing capacity:

… Professor Philip Hill at Otago University, last week called for an urgent ramping up of contact tracing. He declined an interview with RNZ saying he is now engaged with the ministry to come up with “novel approaches engaging people from multiple disciplines, to get ahead of the virus using the ‘case contact management weapon’.”

This morning, Professor David Skegg, an epidemiologist, also advised the Epidemic Response Committee that the use of cellphone technology to augment contact tracing must be investigated rapidly.

To me, these expert comments suggests the following propositions:

  1. It is essential to contain covid-19.
  2. Contact tracing is essential for containing covid-19.
  3. Contact tracing as currently practiced is resource intensive, slow, difficult, and potentially not as reliable as other methods of monitoring interactions between individuals.
  4. That other methods of contact tracing must, and will, be explored by Government and other actors.

Private data for contract tracing purposes

This post follows a tweet I posted on 24 March 2020 where I asked the following question:

“why shouldn’t New Zealand compel access to private smartphone and IOT data to enable contact tracing and physical distancing requirements [sic], within appropriate legal constraints? What do you think?”

I posted the tweet in order to test my own emerging view: that New Zealand, as a community of individuals or its Government, should explore the possibility that private data could be used to mitigate the spread of covid-19.

Many of the respondents to my tweet pointed to places where analysis of that possibility already exists. My view now is that further discussion is required to consider how any barriers to enabling the use of private data for contract tracing can be minimised.

The law is not determinative here

My initial interest was in the domestic legal aspect of this question: could the Government legally compel other entities to disclose private information for the purpose of contact tracing and level 4 enforcement? The answer seems to be that it can, although that depends on what data exactly in what format is being sought from which entity.

Anna Prendergrast, Co-Director at AntiStatic, pointed out that the legal aspects of this question pursuant to the Privacy Act 1993 had already been the subject of comment by the Office of the Privacy Commissioner in a piece by the New Zealand Herald. The Herald asked whether “the Ministry of Health could approach a mobile phone company and ask it to handover the movement data of someone infected with Covid-19”, and the OPC said this:

“Under the Privacy Act and Telecommunications Information Privacy Code, telcos are able to disclose telecommunications information where they believe on reasonable grounds that it is necessary to prevent or lessen a serious threat to public health.”

Even apart from pre-covid-19 domestic legislation, since issuing an epidemic notice and entering into a state of civil defence emergency, New Zealand’s legal and political system has been temporarily re-engineered. It has been re-engineered around three key policy goals, in pursuit of the overall efforts to minimise the loss of life in New Zealand as far as possible. It is the third goal that cannot be overlooked:

  1. maintain physical distance to prevent community transmission;
  2. trace the contacts between people who are infected and others; and
  3. maintain New Zealand’s legal, social, political and economic systems while achieving the other two goals.

It can be difficult to understand the extent of this re-engineering. Professor Andrew Geddis, an expert in constitutional law, put it like this:

“The police (and the army, if needed to support the police) are empowered to order any person to stop any activity that contributes to the current emergency — essentially, spreading COVID-19 in the community. Government ministers may set aside virtually any legislative provision that becomes impracticable to apply while the epidemic is in force. … These give the state extraordinary reach into our lives, and transfer extraordinary power to the executive branch. They are a marker of just how severe the threat that this virus poses to us all.”

Essentially, there are few greater legal or constitutional “levers” that can be pulled to enable the government to meet the demands of this crisis. The government is empowered to use force to achieve its objectives, and it is allowed to reconfigure the constraints on those powers if necessary.

People who point to legal protections as a reason for refusing to proceed with private data contact tracing (including domestic legislation and human rights instruments) must make their arguments against this legal background

To assess the justification of this proposal, we must move to specifics

There have been a proliferation of pieces examining the use of private data in covid-19 monitoring. Those pieces effectively amount to the following summary:

There are significant privacy implications of using data to respond to the threats of Covid-19 and there is a tendency for governments to retain emergency powers beyond the duration of the emergency once they have been conferred.

The sentiment of this proposition is not controversial, but now the discussion must progress to a deeper level of analysis. Civil society is obliged to advance the argument to this necessary next stage.

I observe that responses to the question of using private data for containment tend toward the following, which I summarise here for the sake of progress:

  1. This kind of tracking should not be possible ever by anyone for any purpose. All devices and platforms that permit it should not be used, or should be re-designed in ways that make it technically impossible to do.
  2. We should not give government access to this kind of tracking ever.
  3. We should not give government access to this kind of tracking in these circumstances: the current situation does not justify it.
  4. If Government were entrusted with access to this kind of data and these kinds of powers now, it would not relinquish that power once the crisis has passed.
  5. Government should be given access to this kind of data, but only with suitable restrictions, and these restrictions do not already exist in the law.
  6. Government already has the power to compel this kind of data under existing law and should compel it.
  7. There is no point, as any approach based on private data collection will be ineffective.
  8. That the same result can be achieved by less invasive means.

My opinion

My personal view is that, with appropriate technical and other supervision, the undeniable privacy implications of using this data could be managed. Further, any consequent risk is justified given the scale of the crisis we face. I am not qualified to say how this could be achieved at a technical level, but I would like to hear from people who are capable of designing and assessing such solutions.

Based on the seriousness of the situation, it seems unreasonable to flatly object to any use of private information at all for contact tracing. In this case, as in all cases, the right to privacy must be balanced against the rights to health, effective governance and life. Any assertion about the right to privacy must be made more specific by saying, for example:

  1. that privacy of individuals is not being given adequate weight by reference to other relevant factors; or
  2. that changing something about the specific approach being suggested could better give effect to the right to privacy; or
  3. that established frameworks for considering whether any limitation of the right to privacy is justified have not been followed, have been applied incorrectly, or are based on incorrect evidence.

There have been some very good summaries of how other countries are responding. New Zealand must learn from these countries. I have included links to some of these articles at the conclusion of this piece.

Observed difficulties with voluntary arrangements

One notable approach used in Singapore involved the use of open source voluntary alternatives that use minimally invasive technologies like Bluetooth rather than GPS.

There are many reasons why individual consent is important. First, a voluntary scheme may see better uptake and compliance, although I have not seen any evidence to support this suggestion. Second, there is a clear requirement to seek consent and agreement on human rights grounds where possible. The point was eloquently put by @ Nicolalauren_ as follows:

“the idea is, and tis just an idea, you build better societies and better outcomes on agency, trust and respect than with surveillance. The collective right to health in this instance is achieved by everyone complying and communicating voluntarily.”

I agree that we should pursue this idea, but the question is whether present circumstances justify a departure from it. Voluntary compliance is still being backed up by a threat of state coercion. Further, the magnitude of the risks that we are facing make the consequences of non-compliance with a voluntary system drastically different from the kind of factors we expect to be accounted for when it comes to, for example, signing privacy agreements with commercial services providers.

There is another significant difficulty with a voluntary approach to do with problems of collective action. When it comes to covid-19, the consequences of unintentional non-compliance with physical distancing are just as significant as intentional non-compliance. Further, intentionally non-compliant people will not volunteer for a service that monitors their compliance. In addition, unintentionally non-compliant people may be disengaged from what is required of them, and also unlikely to participate in any voluntary scheme. Finally, even intentionally compliant people may not participate in the scheme, for reasons completely independent of privacy considerations.

There is a collective action issue here that needs to be examined, including by reference to evidence about human behaviour and the magnitude of the consequences if contact tracing is not successful. Systems based on trust are desirable and important, but not at the expense of effective outcomes.

Accounting for existing baselines

Any discussion must take account of the way that personal data is already collected, used and disclosed by a range of actors around us all the time. The data already exists. It is used for commercial purposes and can often be purchased in various forms. To me, the question is whether we will use it for this purpose and what restrictions we will place on its use.

Subsequent discussion must also account for the alternatives available if digital first privacy-limiting solutions are deployed. Currently, physical distancing requirements may be enforced by police or other enforcement agencies. It is essential that any discussion explore other minimally invasive ways of achieving the same policy effect.

How do we move forward with this discussion?

The key areas of disagreement on this topic are adequately staked out. The question is unlikely to be what principles to apply, or what considerations to take into account. It is more likely to be a question of how those principles or considerations are incorporated when assessing the specific circumstances at hand in New Zealand. We need a forum for this discussion to take place in an orderly, progressive and respectful way, probably in a digital medium.

Any discussion needs to be localised to New Zealand’s circumstances now, at this point in time in this context, without being abstracted to a generalisable philosophical level. People are not required to concede that allowing for limitations on the right to privacy in this case should lead to limitations in all future cases. It should be clear that few people are seriously suggesting an enduring system of government surveillance without legal constraints: the real question is what constraints we require.

Any future situation where emergency powers are invoked must be justified on the merits of that situation. There is an opportunity now to set expectations (or even conventions) about how government will use such powers in subsequent emergencies.

Points to discuss

New Zealand, and not solely its Government, should be considering how data from a range of devices could be used to enable contact tracing and to monitor compliance with physical distancing requirements.

In order to do so, we will need to address the following questions, among others.

  1. What precisely is the technical description of the process we are discussing? There is no merit to a discussion that proceeds on the basis of comparing apples and oranges. For example, compelling or purchasing data from private providers like Apple, Facebook and Google will be different to the location data that can be provided by mobile phone companies.
  2. What are the permissible data sources? Will there be restrictions on the data that can be accessed because of private legal arrangements, like platform terms of service?
  3. What legal authority is required to obtain data from those sources? Are they based outside of New Zealand’s sovereign borders?
  4. What constraints can be put on the way that data is processed once it has been accessed or collected? What should the legal status of those constraints be? For example, could an Act of Parliament provide time-banded authority that will lapse on a future date?
  5. Can the data be processed in a way that is useful? Is further capability required within the entity responsible for processing the data? Or is effective processing impossible?
  6. Is the data reliable? What does the data show? What are its limitations? What are the biases inherent in it because of the way it is collected, or because of other structural inequalities?
  7. Who should have access to the data? Can we explore innovative custodianship arrangements like the use of non-Government data trusts?
  8. When will the data be deleted and how will we know this has taken place to our satisfaction?

Forum for discussion

The best thing that can occur is for discussion to happen in the open as much as possible. It should be spread beyond the relatively narrow communities present on Twitter and Medium where I have tried to initiate this discussion.

Civil society objectors need to consider that pre-empting government action is likely to be the time when they can have the most influence on government policy. Advocacy should not be left to the point where the train has left the station. But it should be clear to all that the Government does have legal authority to pursue this course of action if it chooses to do so, or can create that authority.

Finally, if digital contact tracing is to take place, then there is little public interest in that contact tracing being ineffective. Even privacy advocates should find it difficult to suggest that, once implemented, the effectiveness of the programme should be sabotaged or avoided. The public should be given confidence in the oversight and control arrangements being put in place.

Ultimately, there is no right answer here, only a best answer, or one that we are prepared to accept in the circumstances.

Conclusion

Contact tracing, resource burdens and level 4 lockdown

The proposal to use private data for covid-19 management cannot be separated from the third policy goal I have set out above; the need to maintain the social and economic systems that sustain our society. New Zealand’s ability to contact trace will be an essential aspect of the wider decision whether to decrease the alert levels for particular geographic areas.

Apart from the number of deaths arising from covid-19, the overwhelming impact of failing to “flatten the curve” is that these systems would be overwhelmed and subsequently undermined. The economic impact of covid-19 will already be unprecedented, and it will have a social impact too. Economic shortages will flow through and inhibit other pro-social policy goals, leading to untold social effects:

Nothing important about us and our success as a species can be understood except by looking at our interdependencies. If many of us could not come to work — because of sickness, because of the need to care for loved ones, or because of mandated social-distancing — then the fabric of our society would begin to tear. … Human beings are ambivalent about their interdependence. To need others is to be vulnerable; when we’re under threat, vulnerability elicits fear.

Our systems must be restored to their normal functioning as soon as reasonably possible. For that, contact tracing is essential.

“Without a vaccine, as soon as the toughest restrictions are relaxed, a second wave of infections would be expected to follow.” … “Contact tracing and intensive testing — as deployed in South Korea — can help extend the effectiveness of measures to stifle infectious spread, as can technology if concerns about civil rights are set aside.”

Personal perspective

Like other New Zealanders, I have friends and family who work in the healthcare system. I also have professional experience in health systems regulation and policy. I am acutely aware of the incredible capacity of that system and those people; I am also conscious that it has its inevitable limitations despite the best efforts of those working within it. My view is that these systems must be supported by the use of digital technologies that can minimise the spread of infection, and better trace infection where it may have spread.

From a personal perspective, I am a high-risk case. I was diagnosed with type 1 diabetes in 2002. Even in a state of health, I am dependent on the public health system as well as the logistical systems that supply it: without insulin, funded by our public health system, I cannot survive for longer than a week. If infected, it is highly likely that I will require critical care.

This personal experience informs my ultimate conclusion on this issue. First, the scale of this crisis is unprecedented in the lifetimes of many alive today. it cuts across every aspect of our society. Second, similar crises have never occurred in the sophisticated digital, economic and political systems that currently exist: these systems are already interconnected. Third, given the impacts of those digital systems on the rest of our lives (consider their impact on an isolated population at the moment) we ought to consider how they can be shaped to our collective benefit, particularly where that will lift the burden on the health systems we are seeking to preserve by going into isolation.

I am not an advocate for generalised surveillance for political or economic purposes, but the question for me is this: if not now, when? What, if anything, can we do to make this capacity available in ways that generate confidence in its use?

References and attributions

The following people (identified by their twitter handles) have participated in discussions with me, or shared research resources with me, that have informed this post (please contact me if you would like your handle removed from this list):

  • @ Apndrgrst
  • @ nicolalauren_
  • @ andrewtychen
  • @ nullary
  • @ Liz_in_Shanghai
  • @ internetrights
  • @ mwilcox
  • @ mackenz_kitchen
  • @ aenertia
  • @ davidt1008
  • @ farfields_0x04
  • @ thomasbeagle
  • @ verbman
  • @ rossdawson
  • @ mengwong
  • @ JCE_PC
  • @ ben_r
  • @ tippy_top
  • @ ChrisKeall
  • @ teh_aimee
  • @ curt_is_online

Relevant links

Links that have been shared with me since my tweet:

--

--

Tom Barraclough

Director and researcher at the Brainbox Institute, a home for discussion about the law, policy and emerging technologies www.brainbox.institute