Sidewalk Labs (a Google affiliate) asserted in a blog post accompanying its October 15, 2018 smart city draft governance proposal that the Sidewalk Toronto project at Quayside “can set a new model for responsible data use in cities – anchored by an independent Civic Data Trust”. (As readers know, I find the use of the term “civic data trust” problematic. Stringing these words together has marketing appeal, but I believe the phrase obscures certain social, political, legal and economic issues. )
In this post, I would like to step back and consider the variety of problems that a public digital stewardship model might address for a smart city. I don’t claim that these are the only problems, and there are many ways that these problems can be articulated. This is just a quick contribution to the discussion.
So, what are the possible problems in a smart city that a public digital stewardship model could be designed to address? I’ll begin with the basic problems involved in creating pools of data and then move to issues of individual and collective control.
1. Preventing incumbent control in order to facilitate data sharing
Even though data is not “owned” in the traditional sense, those with the means to collect it can hoard and control it. This is a basic problem in a smart city if you believe that data collected in public spaces should be used for the public good.
Let’s take a simple example. Theoretically, we can each count the number of cars driving through an intersection. The “number” of vehicles isn’t something we “own” in the way in which we might own one of those vehicles. But we can control who gets to see and use our record of the number of the vehicles. We can keep it confidential and license it to others to use for a fee. Some of us will be more accurate at counting than others. If we are counting with sensors, it will be very likely that not everyone who might want to put a sensor at that intersection will get to do so. If we are lucky to be the incumbents, we can horde the data for ourselves.
One possible solution to the problem of incumbent control is to create a centralized public digital stewardship structure that requires sharing of data as a condition of collecting it.
the main problem with this type of public digital stewardship structure is that its power relies on the voluntary assumption of the obligations by participants. Unless there are parallel coercive structures (e.g. restrictive covenants embedded into the land grants/leases or laws), no one is forced to participate. If I am a telecommunications carrier (as an example) and I operate a tower and I think that the benefits of me keeping the location-based data of individuals within the territory is more valuable than sharing it, then, well, good luck to you. I don’t have to participate unless there is some law, licensing requirement or restrictive covenant in my land rights or other rights that require me to do so. Same with a retail shop. Absent a law, you need some type of enforceable restrictive covenant. Otherwise, no sharesies are required.
2. Creating data lakes that meet privacy requirements
Another problem in the smart city is how to address the privacy interests of individuals in a way that reconciles the differences between public sector and private sector privacy laws.
When data is about an identifiable individual (that is, the data is not de-identified at source), we trigger the right of those individuals to control how their information is collected, used and disclosed. These rights are embedded into private sector and public sector laws. Complicating matters, the rights that individuals have are not the same in the private sector as in the public sector. In the private sector, the fundamental problem is consent. How do you get consent from the person going through the intersection? In the public sector, you might be able to get around consent but there will be problems relating to whether there is the authority to collect the information in the first place and then providing notice of collection to the individual. Restrictions on use all differ between public sector and private sector laws.
So, a centralized public digital stewardship structure can promote standards to ensure that the data is (a) not personal information to which the private sector or public sector laws apply because it is de-identified at source or (b) collected in a manner that would satisfy public sector and private sector laws to permit sharing.
As an example of what I mean, imagine for the sake of argument that the city could establish a lawful basis to collect information on unique vehicles travelling through an intersection without having to obtain consent (but with notice). Say, for example, there was a toll bridge that read licence plates and identified the vehicle owner for the purposes of billing. If the data collected by the city is then sufficiently de-identified it could be shared more broadly with the private sector. So, the data can be used by some in an identifiable form and by others in a de-identified form. The public digital stewardship structure can lay out the framework and data flow permissions to facilitate this sharing in compliance with privacy laws. The private sector entities using the data could be bound by agreements managed by the public digital steward. The information flow could go the other way too from the private sector to the public sector.
A side benefit to a public digital steward handling these information flows is that the public digital steward may also enhance transparency by being a single resource for individuals wanting to understand what many disparate players who are operating sensors in a particular geography are doing and where the information is flowing. It could also provide a single point of contact for Privacy Commissioners and other relevant regulators.
3. Mitigating the effects of size-asymmetric competition for data
One of the problems with “open data” is size-asymmetric competition with respect to its use. In simple terms, a big data company like Google does not compete on a level playing field with the local start-up when it comes to the ability to exploit data. When Google gets data from a smart city, it is potentially adding that to the huge amounts of data it already has. The start-up does not have that same advantage. This perpetuates the incumbency of those who are already bloated with data. When dealing with a public resource, equal opportunity to access that data doesn’t mean equitable opportunities to exploit the data from a competition enhancing view point.
Addressing this problem means looking past just creating big data lakes for everyone to play in. It involves considering how to best allocate the economic opportunities from accessing and using that data for the purpose of serving larger public policy goals.
A public digital stewardship structure could theoretically assist in addressing that problem. A public digital stewardship structure could facilitate the sharing of data (see problems #1 and #2) but do so in a way that enhances competition rather than reinforces the dominance of incumbents. The public digital steward could, for example, provide preferential access to start-ups, researchers and public sector organizations if capturing the benefits of that data in these places was thought to be important for public policy reasons.
Making data partially open in this way can also preserve the economic benefits of the data for the owner or licensee of the digital infrastructure collecting the data. The entity contributing the does not have to worry that the entire economic benefit of the data will be entirely diluted by it being completely “open”. The entity contributing the data might be free to license the data for a fee (or the public digital steward permitted to do so in exchange for a royalty) when the data is being provided to for-profit entities but be required to provide the data freely to city agencies and researchers. Or, certain types of entities, such as start-ups, non-profits, and city agencies might get the data in advance of other entities. These are but two (and not necessarily the best two) options for curbing the problem of size-asymmetrical competition.
These options also have the additional benefit of incentivizing the collection of the data because the raw data will still have some value to the collector through the ability to licence its use for a fee. They also have the additional benefit of incentivizing the collection of the data and participation in the sharing of the data because the raw data will still have some value to the collector through the ability to licence its use for a fee or to have some short-term preferential access to it along with a limited set of other users.
4. Digital layer and data sovereignty and local economic benefit
There are significant problems of sovereignty and local economic benefit as the smart city scales. This isn’t perhaps a big issue if you are talking a few acres in a portion of a city or non-critical infrastructure. However, it is a legitimate question for citizens to ask whether they and their city’s public spaces (funded by them) should be the laboratory in which others experiment without tangible direct economic benefits to the local citizens. As the smart city scales, it is also a legitimate question to ask what risks there are in private sector companies, particularly foreign companies controlling the digital layer and the data of the smart city. I link these questions together because at the level of the practical they are intertwined with one another, although I accept that they can also be examined separately.
A public digital stewardship model cannot, on its own, resolve these issues. However, a public digital steward can be a lever to assist in addressing them. On the economic front, the digital stewardship structure could be a mechanism through which to capture royalties related to intellectual property developed or improved using the city or its data as a laboratory.
On the sovereignty front, a digital stewardship structure could require that data and control of the digital layer remain in Canada. In the extreme, the digital steward could be a trustee holding the critical infrastructure comprising the digital layer in trust for the city if it is not otherwise owned or directly controlled (i.e. more than via contractual rights) by the city or public agencies.
Digital sovereignty remains unfashionable, but as we move into public spaces, citizens will eventually ask why it is the keys to the kingdom were handed away, particularly if the keys were handed to foreign companies. The public digital steward can be one bridge back to democratically elected institutions by ensuring that the rules related to laying down the digital infrastructure and the digital data kept private sector companies rooted in Canada and meaningfully subject to Canadian laws.
5. Collective or individual control mechanisms
The most challenging problem in the smart city is how to curb the panoptic effect of laying down a digital layer that could be combined with data collected by participants not directly involved in the smart city but who interact with individuals living, working and visiting there. For example, the data from the smart city could also be combined with information the same players or their affiliates collect from individuals through smartphones and other consumer devices in the homes or vehicles.
Liberal MP Adam Vaughan dismisses this fear, being quoted as saying “I’ll be dammed if I surrender [the Waterfront Toronto work] just because you don’t like the way Google or Facebook or Twitter handles personal data.” However, the fear of a loss of control is a legitimate response to the reality that devices are capturing more and more data that in discrete instances might not be problematic, but, in combination, tend towards invasive observation.
In its most radical form, the public digital steward could act as a union to bargain collectively with those seeking to collect and use data in the smart city and as the custodian or manager of technology that enables individuals to control their digital identities. In this form, the primary focus of the digital steward is to advance the collective interests of citizens and incidentally facilitate other objectives, such as data sharing. This can be done in a low-tech way, by simply empowering the digital steward to negotiate binding agreements with the collectors and users of the data. It can also be pursued in a high-tech way. One example is the DECODE (Decentralised Citizens Owned Data Ecosystem) project in Barcelona and Amsterdam. The DECODE project’s goals are blockchain-based applications to allow individuals to set permissions, including how their data is shared in a data commons. A side-benefit of the public digital stewardship model may be to enhance community participation.
The major challenge for the public digital steward is, however, the very size-asymmetry that I discussed earlier in this post (see problem #3). Unless the public digital steward has (a) the buy-in of material numbers of individuals to use its technology, (b) the ability to control the means of data collection, or (c) the backing of the authority of the state (through laws and regulatory enforcement), the public digital stewardship model can only nibbles around the edges of the problem and is unlikely to be able to force the key players to the table in a comprehensive way. Just remember what happened with “Do Not Track”.
For all the more complex social and political problems, all roads lead back to governments to use legislative power to enhance governance structures.