Public debate, public governance: what's missing from the Sidewalk Labs project | What's Left Issue 110

The dog days of summer saw a new development in what some are calling a new world-historical event, and not in a good way. It is the creation of Sidewalk Labs on the Toronto waterfront through an *innovative* (one hears this word a lot) and perhaps unholy alliance between the Alphabet subsidiary (Alphabet is both Google's and Sidewalk Labs' parent company) and an odd bird of municipal governance, Waterfront Toronto.

Public debate, public governance: what's missing from the Sidewalk Labs project | What's Left Issue 110
Page content

By: Simon Archer

The dog days of summer saw a world-historical event, and not in a good way. It is the next stage in the developments proposed by Sidewalk Labs on the Toronto waterfront through an “innovative” (one hears this word a lot) and perhaps unholy alliance between the Alphabet subsidiary (Alphabet is both Google’s and Sidewalk Labs' parent company) and an odd bird of municipal governance, Waterfront Toronto.

So, what’s going on? The glib response is that for too long, in Toronto, too little has been going on to maintain or upgrade public infrastructure, including development of the waterfront. The perennial and vexing example is the need for more and improved urban transportation, plans for which have been stalled for over a decade, and have one again been subject to yet another review and revision.

And Toronto’s issues are an example of a wide-spread problem in North America – governments have failed to re-invest in public infrastructure for 50 years, and it is literally falling down.

Into that context comes the Sidewalk Lab. It promises new, world-leading forms of development including affordable housing, efficient transportation, a new mode of sustainability, community, public realms and something called open digital architecture.

Sidewalk Labs and Waterfront Toronto are exploring their ability to innovate new forms of municipal funding, development and governance to lead (read: export to) other cities around the world. They call it variously smart city development, digital infrastructure, people-first cities in a digital age, accelerated urban innovation … you get the gist.

To this end, on June 24, 2019 Waterfront Toronto released a Master Innovation and Development Plan (MIDP) to be available for comment as part of its review process. The Chair of the Board of Waterfront Toronto immediately made clear that it wasn’t Waterfront Toronto’s plan, it was Sidewalk Labs' plan, and Waterfront Toronto was merely considering whether it was in the public interest and “objectives for Quayside that we agreed to with Sidewalk Labs in July 2018 (as reflected in the Plan Development Agreement).”

The whole deal – the partnership, if it is a partnership, with Sidewalk Labs, the process, the MIDP, all have encountered criticism from academics and urban planners, from tech company leaders, venture capitalists, privacy advocates, and the Canadian Civil Liberties association has voiced their critique in a legal proceeding seeking to “reset” the entire scheme – that is, to invalidate the deal with Sidewalk Labs altogether.

But let’s take it on its own terms – against a backdrop of nothing happening at a furious pace, anything new is worth a look. I’m not sure I can recommend reading it cover to cover -- the MIDP is 1,500 pages -- it mainly depends on your tolerance for management-speak and a certain public interest jingoism. But it does make many proposals, some of which sound good and at the same time sound maddeningly vague.

One thing the MDIP does is list examples of what Sidewalk Labs does. For example, Sidewalk Labs is somehow responsible for the Highline project in New York City and its sort-of clone, the Bentway project in Toronto. (The timing seems suspect: perhaps they hired people who did those jobs.)

If you’re in Toronto you know the Bentway, the “park” built under the major highway that cuts along the bottom of the city and over the lakeshore. Now, these are interesting little projects. The Bentway wasn’t created as a public park – it was created through private money, a donation of a wealthy family. That happens a lot with universities and hospital wings, but not a lot with parks, yet. The Bentway isn’t governed through the Toronto municipal parks system, but maintained and programmed through a separate, not-for-profit charitable organization. On the board of that “conservancy”, as it is termed, are the donors, various developers and their service providers, and a couple of municipal counsellors. It is called innovative, of course, hosts events, art shows, and so-on, and is all coordinated through that corporation by its full-time staff. It’s sort of a demi-private urban park development.

So that gives you a clue as to the direction Sidewalk Labs might be headed: a sort of privatized governance over former public realms (to my delight, the MIDP actually uses the word “realm”), all funded through private sources combined with innovative – of course – cost recovery mechanisms and “programmed” in the widest sense of that term. It’s a sort of neo-Victorian philanthropic model meeting neoliberal era urban development.

Proponents (and I’ve spoken to some working on the project itself) make several points in support of the deal and process. Three points stood out to me.

First, little else is happening, they say, and the politics of urban development are frozen and starved of resources from traditional sources. (Those sources have always been mixed, but public sources in particular is what is meant.) There’s something to that diagnosis, although it doesn’t necessarily lead to the proposed cure. We have local and global experience with private funding of (formerly) public goods, services and functions, and in particular infrastructure. Those experiences are very mixed for the users: ask anyone who has taken a train in the UK recently, or paid a water bill in the Thames valley, or wanted to cancel a gas plant. There ought to be a debate about truly public options for funding urban development, but there isn’t much of one these days, and to have one requires a difficult discussion about the power to raise funds and the sources of public revenues – including, even, the taxation of entities like Alphabet.

Well then, it is said, corporations like Alphabet are going to do whatever they do (digital surveillance, commodification of data) anyway, and are doing it already: so why not use an opportunity to engage with them and mediate it? This is an excellent point, and the force behind it is that Alphabet is actually willing to engage, and so must see some of this process as furthering its interests. By contrast, Amazon has wandered around extracting subsidies and rents from target cities, and Zuckerberg wouldn’t even show up when summoned by the British Parliament after breaking a host of laws. So the fact that Alphabet shows up to talk governance turkey is an opportunity. The ultimate question will be: do you think you can govern it? We’ll return to this question through an example.

Doing so – engaging, mediating, seeking to channel the way in which these corporations operate and innovate toward some sense of a public interest – raises questions of legitimation and capacity for regulation. No-one seems to know how to regulate the tech giants into better behaviour. Various options have been proposed, from classifying them as media and bringing them into existing regulatory spaces (doesn’t seem to have done much good in Canada, where we have a functioning oligopoly and some of the highest user fees in the world) to using competition law to break them up, which seems an ill fit to platform-based operations like Google or Facebook.

One “innovation” in this regulatory dialogue is the use of (publicly owned and operated) “data trusts”. There are versions of this idea out there, ranging from “data fiduciaries” to outright public ownership and stewardship of this data. A data fiduciary is held to a (higher) standard of conduct, at least in theory (there has to be someone to hold the fiduciary to that standard of conduct). Outright public ownership (created through the assignment of individuals data righs to a public trust) provides more options than policing fiduciary behaviour: some propose to use or regulate access to data trusts to bid up the costs or otherwise strictly control of use (and abuse) of data such that bad data use behaviours are discouraged and well behaved competitors can enter the market that is currently dominated by a handful of tech giants who have thus far evaded regulation and taxation. Plus, they could generate revenues.

Interesting proposals, and we might assume that the tech giants feel, from breach to breach, some regulatory heavy breathing just behind them, in an amorphous form. An understandable reaction would be to seek to mediate how that regulatory response approaches you: hence, talking governance turkey.

Intriguingly, the MIDP speaks of data trusts – they rebranded a “Civic Data Trust” to an “Urban Data Trust” in response to questions from stakeholders. Those questions were asking who is a trustee, who is a beneficiary: in other words, bothersome questions about who owns data and who they must be accountable to. Fiduciaries, trustees, beneficiaries, standards of care: we’re quickly wading into legal niceties, but there are legal – and non-trivial – distinctions between property owning trustees and other forms of behavioural regulation. These distinctions may, in the end, matter quite a bit.

The MIDP quickly makes clear that by saying “urban trust” they do not mean an actual trust, a legal trust relationship with ownership of property (data) and stewardship of that property for beneficiaries (individually or collectively). What they really mean is a quasi-public regulatory body mandated to monitor compliance with privacy laws through a self-funding not-for-profit corporation staffed by “stakeholders”. Does that governance model sound familiar? It did under the Bentway.

To flesh that out a bit, Sidewalk Labs' proposal is to create a minimum standards monitor (a Chief Data Officer) to ensure compliance with privacy laws, and to develop new industry norms for data use. That right there is, to my mind, the whole point of the Sidewalk Lab. Data trusts sound good, and interesting, but this isn’t really a data trust in the sense of publicly owned and managed property or infrastructure – it is a mode of market regulation, and one that is pretty close to self-regulation.

By contrast, the language the MIDP uses to describe public options (like actual data trusts, or robust public regulation by a government agency) is indicative of its position: public funding is characterized as a disadvantage, and self-funding of regulators as a good (the effect being the regulatory entity is financially reliant on the businesses it regulates to exist). The advantages of public ownership and regulation are said to be scale and applicability in other jurisdictions (scale, yes; but applicability elsewhere isn’t intuitively evident), and in contrast, the disadvantages are that it requires slow-to-respond legislative processes. Obviously bad, those legislative processes – but not always slow: see how many bills got passed by the most recent Ontario government in its first 18 months. Behind all this is an unstated assumption that government should not or cannot perform these functions; the same assumption applied to the funding of public infrastructure mentioned above.

Whatever its assessment of pros and cons, the MIDP does quite correctly note that flexible regulation and self-regulation are forms of regulation that have been around a while – lawyers are infamous exemplars of a self-regulated profession – but others as well, and the MIDP suggests some examples. Universities ethics boards are one, professional self-regulation another, and at one point the Toronto Public Library system is cited as an option to be explored. (Seriously: it proposes that the TPL be used to house data used by tech giants, based on its experience with data management.)

But in invoking self-regulatory mechanisms, Sidewalk Labs is only repeating, or replicating, well-understood modes of governance that have grown to dominance in some sectors and industries over the past 40 years, in particular in response to conditions of globalization and the retreat of the administrative state. The word innovation is used a lot in connection with these modalities. And, in fact, there is a well-developed academic literature on how industry actors use standard-setting mechanisms and these modalities of regulation to define the norms and terms of engagement in an industry and, sometimes more nefariously, preclude more robust public oversight of them, in what one academic called the “housekeeping of capitalism.”

I suppose the real test will be, do we believe that, at this point, in these circumstances, a self-funded 5-person non-profit corporate board staffed by industry professionals and at least one academic and one community member (yup: read the MIDP volume on data management, Chapter 5) and either start-up funded or self-funded through the entities it regulates, is that entity going to be able to stand up to Alphabet or Facebook or Uber or any of the other large tech corporations which have come to dominate the space, and regulate them, sometimes against their interests? Long odds, I’d guess.

It’s the tech world’s version of utopian urban planning. One is left in something of a bind: it is refreshing that new proposals are being made at all – even if some of the key stuff is opaque or recycled. Stepping back for a minute, though, we are reminded that the history of utopian urban planning has a long and often sad history. At its best, it aspires to be an expression of – of all things – of socialism: cheap and available public infrastructure and transportation, salubrious housing, economic opportunity – writ large, the preconditions for social stability and security as well as the environment for economic development. What page will Sidewalk Labs be on?