Ontario is currently engaged in a data strategy consultation process. The stated goals are to create economic opportunities and to improve government services by facilitating greater data sharing and by using more analytics and artificial intelligence. The plan is to do this while maintaining the ‘trust and confidence’ of Ontarians. The consultation process has had an extraordinarily low profile considering what is at stake. That said, it is happening so quickly that it is easy to miss. Even for those paying attention, the consultation is long on boosterism and short on detail. This post outlines some reasons why Ontarians should be concerned.
1. Major transformation without proper debate/consultation
Developing a data strategy is a good idea. Data-driven innovations are dramatically changing our economy and society. There are many ways for government to become more effective and efficient by embracing new technologies. It could also become more transparent and find new ways to engage citizens. To do these things some changes to the law and policy infrastructure will be necessary. Businesses seeking to innovate and grow in the digital and data economy will need better access to quality data and, among other things, new models for data governance and data sharing. Data-driven technologies also bring with them risks of harm and these too may need new legislation or normative frameworks. There is a lot to consider and some of the changes will be transformative, and will rely upon citizen data. These are all good reasons to consult deeply and broadly, both to seek input and to lay the foundations for a transparent public engagement.
The data strategy consultation was announced in February 2019, with a report to be published before the end of the year. The consultation centres around three discussion papers, the first of which was only made available in mid-August 2019 and the last of which has just been released, with comments due by the end of November. The public meetings held as part of these consultations are taking place on very short notice. The process is hurried, obscure and fails to properly engage the full range of stakeholders.
2, Superficiality
Quite apart from the rushed nature of the process, the discussion papers are woefully inadequate. They are full of assertions of the benefits of what is planned, with the occasional nod to the importance of privacy and trust. There is little detail about the nature, scope or timelines for what the government plans to do.
The discussion papers give only brief glimpses of things that merit a much more detailed treatment. For example, on the issue of broad-scale data sharing between different government departments and agencies, we are told that “while ‘connecting the dots’ between datasets can help government provide better services, there are privacy and cybersecurity risks to be managed.” ‘Connecting the dots’ can mean all sorts of things. Perhaps data matching will be used to find ways to improve service delivery. Data analytics might also be used to discover or anticipate certain citizen behaviors. This could include identifying patterns that suggest certain individuals are fraudulently obtaining benefits or cheating on taxes, or identifying children potentially at risk. The data-matching possibilities are endless. And while the goals might be important, there are significant risks of harm. Beyond privacy concerns, there are issues of discrimination and undue surveillance. What processes will be in place to ensure transparency and accountability as these programs are developed and implemented? The consultation documents are so general and superficial that they fail to identify, let alone invite engagement on some of the real challenges posed by the government’s (undisclosed) plans.
An alarming glimpse of what lies beneath the superficial gloss of these documents is found in the second discussion paper which focuses on “Creating Economic Benefits”. The document talks about the value that can be derived by the private sector from data shared by governments. It then casually states “Given that Ontario has a wealth of data in digital health assets, clinical and administrative health data can also be considered as a high-value dataset that may present various opportunities for Ontario.” This suggests that the government is planning to make the personal health data of Ontarians available to the private sector. As the Privacy Commissioner, in his comment on this aspect of the discussion paper, aptly notes, “It is important to distinguish between the high value of health-related data in terms of utilizing it to foster innovation and research, and its high monetary value (that is, health-related data as a commodity to be sold as a source of revenue for the government). The specific scope of what the government may be contemplating is not clear from the discussion paper.”
3. The Gaps
The government is designing a data strategy but its focus is relatively narrow. Ontario’s Privacy Commissioner has pointed out that there are many other data-related reforms that could enhance government transparency including open contracting and open procurement, as well as other reforms to improve access to government information. While the Simpler, Faster, Better Services Act introduced reforms around open data, open data is not necessarily the best route to transparency, especially with a government that has indicated it wants to be more strategic about its release of open data and that sees it primarily as a driver of the economy.
4. Social impacts
This consultation process relies far too much on the increasingly tired trope of “trust and confidence”. The first consultation document, a truly abstract exercise in asking people what they think about plans that have neither been discussed or disclosed, is titled “Promoting Public Trust and Confidence”. Trust and confidence must be earned, not promoted.
Although the first discussion paper identifies a broad range of issues, including bias and discrimination, surveillance, data privacy and security, these are raised largely in the abstract. In the subsequent discussion papers on creating economic benefits and smarter government, the issues are boiled down to individual data privacy and security. There needs to be a detailed, robust and informed discussion on the impacts of proposed technological changes on individuals and communities, as well as on limits, oversight and safeguards.
5. ‘Stakeholders’ and the Rest of Us
Another issue that should concern Ontarians in this consultation is whose voices really matter. The lightning fast consultation hints at some major changes, many of which are driven by industry demands (such as the massive sharing of personal health data with the private sector). Industry clearly has the ear of government and does not need the consultation process in order to be heard.
The data strategy consultation has been poorly publicized. The discussion papers have been published late in the process, contain little detail, and have narrow windows for providing feedback. The paper on trust and confidence was released in mid-August with comments due just after Labour Day. The timing could hardly be worse for ensuring public engagement. The public meetings around the province are scheduled with very short notice. This consultation favours larger organizations with the resources to throw together a quick response or to find someone who can attend a meeting at short notice. It does not favour the general public, nor does it favour civil society groups and academia with limited resources and personnel.
At the same time that this speedy data consultation is taking place, there are closed-door consultations underway with “stakeholders” about reforms to the Personal Health Information Protection Act. While there is no doubt that much could be done to modernize this legislation, the fact that it taking place behind the scenes of the superficial data strategy consultation is deeply troubling. There is also, reportedly, work being done on an ethical AI strategy for the government. Not only is this not part of any public consultation process, it is only hinted at in the third discussion paper.
It is also profoundly disturbing that institutions that serve the public interest such as the Office of the Information and Privacy Commissioner so clearly do not have the ear of government. The Privacy Commissioner’s input has been reduced to letters written in response to the discussion papers. These letters politely invite the government to seek out his expertise on issues that are squarely within his mandate.
Proposed massive technological change, ‘trust us’ assurances about privacy that fall short of the mark, and a disregard for early and inclusive consultations are a recipe for disaster. People are not data cows to be milked by government and industry, and acknowledged with only a pat on the rump and a vague assurance that they will be well looked after. The data strategy must serve all Ontarians and must be built on a foundation of credible and meaningful public engagement. As the Sidewalk Toronto process has demonstrated, people do care, the private sector doesn’t have all the answers, and transformative change needs social legitimacy.