This is the last of three posts covering the digital and data policy changes announced in the 2019 Ontario budget bill. The first post lays out the details of these policy changes, focusing in particular on the proposed Simpler, Faster, Better Services Act, 2019, and the authority assigned to the new Chief Digital and Data Officer (CDDO) position that would be created through this statute. In my second post I argue that at least in principle this top-down, whole of government legal instrument makes sense, given the institutional and cultural inertia that has historically hindered meaningful public management reform (digital and otherwise).
This leaves one last question on the table, the one that matters the most when weighing the implications of these proposed policy changes.
Will Ontario’s new digital and data policies, and the broader public management reforms they’ll initiate, satisfy standards of democratic accountability?
To use David Eaves’ and Ben McGuire’s language, digital government reforms test the social contract between citizens and the state. Where digital government initiatives are implemented without securing democratic license, they often — justifiably — fail. Take three recent Canadian examples.
Following widespread public backlash and most recently a legal challenge launched by the Canadian Civil Liberties Association, the future of the Waterfront Toronto smart city development, at least as currently conceived, is in serious question. Likewise, Statistics Canada’s attempt to compel personal data from financial institutions was ultimately rendered politically unviable given public outrage over the privacy breaches such a data pull could usher in. Last, consider parliamentary scrutiny of federal government digital services, currently underway in the House of Commons access to information, privacy and ethics committee. A core concern occupying committee members is whether the government can aggregate and share personal data to improve service delivery, while also preserving citizens’ privacy and security, and ultimately, their confidence in the state.
More of these challenges are on the horizon as governments move from building capacity for digital transformation at scale — the struggle that has occupied governments the world over for at least the past 20 years — to actually meaningfully reforming the state’s organization, the governance and use of data, and the interactions between government and citizens. However effective it may seem to the enthusiastic public administrator, digital reforms can’t simply be rammed through with little regard for transparent process and the views of citizens.
At this stage, any efforts to evaluate the democratic acceptability of Ontario’s proposed digital and data policies are necessarily speculative.
The budget bill and the digital and data statute give the CDDO a wide berth in which to move, committing only to sometimes vaguely worded principles (which is, to be fair, likely a strength given that legislation on anything digital-related should have room to adapt to a fast-changing environment).
Instead, most of the digital and data law’s impacts will take concrete form in the set of legal and policy changes that the budget bill promises are coming, and that we can presume the CDDO will introduce. Likely, these will include changes to policies and laws on privacy and data sharing (some of which are already included in the budget bill, more on this below), communications, procurement, information management, human resources, program evaluation, and budgeting. The effects of these second-order policy changes on public management and citizens will only become clear as they’re implemented in practice in departments.
With that uncertainty in mind, I’ve highlighted what I see as the three “democratic accountability red flags” that we should be watching for as these policy changes are rolled out in the coming months and years.
Red Flag #1: Who’s the boss?
If you buy my argument that, by virtue of their control over digital and data standards, the CDDO will have significant influence over all government activities, then it also follows that this person (or more technically, their minister in the Ministry of Government and Consumer Services) will be accountable for a broad set of government outcomes that would otherwise be attached to line departments, not one player in a central agency. This is potentially a problem.
To understand why centralized accountability of this sort raises a red flag, it’s helpful to remember that in Westminster governments, accountability is traditionally structured vertically; ministers are democratically accountable to the legislature for the activities that unfold within their respective departments.
So, if a government program fails, goes over budget, suffers a scandal, etc. the minister under whose portfolio that program falls is held to account, by, for example, responding to scrutiny from the opposition in question period or from media and advocacy groups in the public sphere. Ultimately ministers can be asked to resign for their department’s failures.
But what happens when a department’s policy initiative suffers in design or implementation because of dictates issued by the CDDO from within the Ministry of Government and Consumer Services? After all, we’ve seen countless policy and management initiatives fail because of the digital and data components underpinning them (see: Phoenix pay system, Obama’s healthcare initiative).
Of course, the power granted to the CDDO to dictate standards is predicated in part on the idea that departments currently lack sufficient expertise and leadership on digital and data. The CDDO and the standards they set are supposed to improve ministries’ digital capacity and prevent failures. But it would be naïve and frankly unfair to presume that the CDDO won’t make any poor decisions, or that their directives will always be a good fit for particular policy challenges and departmental realities.
What’s more, departments may justifiably argue that they haven’t been given the resources to deliver on the directives mandated by the CDDO. Here I want to underscore an essential point that hasn’t been discussed yet (to my knowledge) in the debates on these new policy changes:
These new digital and data commitments come with no new funding to support departments in implementing them.
Departments are also facing a hiring freeze. It strikes me as deeply unreasonable to make a department accountable for new responsibilities without providing them with resources to do so. This is particularly true when we consider that delivering on new commitments to user research and data analytics requires access to skillsets sorely lacking in the current public service workforce.
With all this in mind, a minister accountable for a policy initiative that fails because of its centrally determined digital and data components, or that lacks the resources to effectively deliver on new legally enforceable digital and data standards, may justifiably kick accountability for these failings back to the CDDO. Here, accountability can become muddy.
I want to be clear that I’m not suggesting centralized standards are new. In some sense, ministers should already be concerned about the policy and management failures they oversee which are in part the product of whole of government policies and standards over which they have little control. I don’t want to glorify the status quo. Nor do I mean to suggest that accountability will necessarily become muddied as a result of the Simpler, Faster, Better Services Act.
Instead, I’m simply arguing that ministers and civil service administrators need to acknowledge that the Act necessitates a model of horizontal or shared accountability that is not standard fare in the Westminster system. By looking at precedents (such as ServiceOntario) and openly discussing the rules of the game that will constitute this model of horizontal accountability, muddiness can be clarified.
Without doing so, I worry that one of the most basic requirements of democratic governance — clear lines of accountability — will be in jeopardy, undermining the capacity of opposition parties, ombudspersons, civil society and individual citizens to oversee the state.
Red Flag #2: The risky rhetoric of “people-centered government”
The budget bill in particular, and the Act to a lesser extent, emphasize that the province’s digital and data policy changes will usher in “people-centered government”. On one level, the use of the phrase “people-centered” simply reflects the Ford government’s broader branding strategy (they are, as we are told, “For the People”).
On another level, “people-centered government” refers to a more concrete commitment to the established practice of “user-centered design”, something the Ontario Digital Service is already championing in the province and that has emerged as a preoccupation of governments around the world in the past few years.
To be clear, where the Act refers to “user research” and says that users of policies and services should be “consulted”, we’re not talking about an ambitious program of participatory democracy that would invite citizens to express their values and preferences on policy decisions. This is not what user-centered design is about.
Rather, and here I’m relying on a very abridged summary of the process, this commitment means that once the government decides which programs and services it will provide, service designers and user researchers will ensure that those programs and services are developed with a robust, almost ethnographic understanding of the experiences that users have when they complete these service transactions in real life (through a paper form, on a website, etc.).
I’m on board with this. User-centered design has proven itself a useful tool for building public services that otherwise are better tailored to the needs and structures of government than they are the citizenry. This tradition is at the root of many large-scale government program failures.
However, there’s an obvious elephant in the room here. It’s particularly cheeky for a government to promise “user/people/citizen-centered” services while also instituting deep cuts to the services that the state offers. You would forgive the parent of an autistic child, the teacher facing ballooning class sizes, and the person denied legal aid for balking at the suggestion that the Simpler, Faster, Better Services Act, with its commitment to people-centered services, makes anything simpler, faster or better for them.
Now, I know there are those who will read this and say, there’s nothing democratically illegitimate about the Ford government’s decisions to cut funding for certain services. They are, after all, supported by a majority in the legislature, and can pursue whatever political agenda they choose. I agree, but that’s exactly the point.
Decisions about public services are inherently and properly political; “people-centered” government means government for some people servicing some needs, but not all. This may be democratically sound from a constitutional perspective, but it doesn’t mean that promising people-centered government while cutting a tonne of services still won’t leave many citizens disillusioned with the state.
Over-promising and under-delivering is an easy way to alienate citizens and undercut their confidence in government. This risk is certainly at play in Ontario right now.
Red Flag #3: Data governance dilemmas on the horizon
Digital transformation at scale rests on important changes to the ways in which data are collected, stored, shared and applied in government. In turn, digital transformation at scale raises a vast set of complex data governance questions for the state. These are coming down the pipeline for Ontario policymakers, potentially now faster and more furious than they are trickling in in other jurisdictions where digital change remains incremental.
The democratic legitimacy of the broader set of digital reforms the CDDO will lead rests in large part on getting these data governance questions right.
As I noted in my previous post, outside its commitments on open data, the budget bill and the Act say very little about data governance. Actually, Teresa Scassa helpfully clarified on Twitter that even on open data, there is a world of ambiguity in the government’s promises; the bill’s use of the language “publicly-available data” versus open data, and references to user licenses, raise questions about the rules that will govern what public sector data is made available, to whom, and under what terms and fees.
A range of other tricky data governance questions also remain unanswered. Will the province’s new approach to data and digital services include provisions enabling citizens to audit how their personal data is collected and used, by whom, and to what ends? To what extent will the province’s data governance model respect the Indigenous data sovereignty movement? How will data be used to “nudge” citizens to achieve certain socially desirable outcomes, and who decides what counts as socially desirable in the first place? Which principles and ethical standards will govern the use of artificial intelligence for policy design and service delivery?
I’m confident that these issues are already on the radar of the civil servants responsible for the province’s digital and data policies, and I’m equally confident that these civil servants appreciate the trade-offs and risks at play in navigating these data governance challenges.
That said, I’m admittedly less confident that these civil servants currently operate in a political environment that will allow them to address these challenges in ways that satisfy standards of democratic accountability. Notably, the policy changes were announced without any references to principles of equity, fairness, justice, or digital rights. Instead of prioritizing these principles, the budget bill and its promotional announcements emphasize that these policy changes are primarily guided by two objectives: saving money and reducing inefficiencies.
To be sure, when describing the digital and data policy changes, the budget bill and statute do cite the importance of protecting privacy and security, but as Teresa Scassa notes, elsewhere in the budget bill, the government appears quick to dispense with these concerns (for example, she notes that changes to provincial privacy laws would expand the ability of public sector bodies to share personal information with law enforcement without consent).
What’s more, Scassa also notes that the policy changes introduced in the budget confirm the government’s position on certain issues that were otherwise supposed to still be up for consideration through its own digital and data consultation, which hasn’t yet concluded. This also threw me for a loop when I first read the budget bill. What was the point of that consultation? And, of course, let’s not ignore a fundamental process issue, that these policy changes are hidden within an omnibus budget bill.
If the government intends to secure democratic buy-in for these policy changes, this isn’t a great start.
The real work begins now.
This is true for the civil servants that will implement these policy changes, and also true for those on the outside scrutinizing these changes. In particular, as new accountability models, promises of user-centered design, and data governance reforms are introduced, its incumbent on outsiders to ask for whom, to what end and at what cost are Ontario public services becoming “simpler”, “faster” and “better”?
The answers to these questions are far more contestable and values-laden than is implied by the budget bill’s technocratic references to “efficiencies” and “innovation”. How the province responds to these questions will be important to watch not only for understanding the changing governance landscape in Ontario, but more broadly, to understand the connections between democratic legitimacy and the possibility, and desirability, of public sector digital transformation.