Alan Harnum

Public Libraries and Public Realm Data: Initial Thoughts on the Toronto Region Board of Trade's BiblioTech Report

A Note on the Origins of this Post

I originally started this post a few days after the BiblioTech report came out (January 16), then got pulled away from finishing it up fully by other matters. Since then, the Quayside story has seen some quite significant shifts and new developments, brought on by the Toronto Star’s story about the full scope of Sidewalk’s plans, based on leaked internal documents.

I’m publishing the post now as a record of my thoughts at the time; my rawer response to the Toronto Star’s story is captured in these tweets from mid-February when the story broke. If anything, the concerns I expressed in the post are even more sharply felt based on the more recent stories.

Disclaimers and Disclosures (short version)

I worked for TPL for ten years, have been a critic some of their technology decisions in the past since leaving (and during my employment there, if we’re being honest), and my current employer (the Inclusive Design Research Centre at OCAD University) did some consulting work funded by Sidewalk that I had minor involvement in. a more fulsome explanation of my personal stakes here are in this section of the post.

The BiblioTech Report

The Toronto Region Board of Trade (TRBOT - incidentally I must give them compliments for owning the bot.com domain) released a report on January 9th. A quote from the press release gives the essential summary:

Today, the Toronto Region Board of Trade (the Board) released a report entitled BiblioTech, recommending that policymakers assign oversight and development of public realm data policy to the Toronto Public Library. The proposal is in response to the debate over public realm data management at Waterfront’s Quayside project. BiblioTech also recommends improved data privacy enforcement, and the creation of a city-led policy for intellectual property commercialization. The Library was selected because it is a longstanding public institution with broad respect in the technology community for its balanced approach to data policy and information management.

The full BiblioTech report is a short read at seventeen pages, so I’d start there before reading my initial thoughts. This is an important issue for Toronto right now because of the Quayside project, but data governance generally is an emerging public policy and public technology concern; I appreciate the issues the report is trying to grapple with, though I don’t agree with its remedies.

To not dance round the point, I don’t support its central remedy of appointing Toronto Public Library to have oversight and development of public realm data policy in Toronto, either in the specific case of the Quayside project, or the more general question of what institutions are suitable to oversee this.

What Is Public Realm Data? What is a Data Trust?

One of the report’s authors, Brian F. Kelcey, has been actively and helpfully engaged on Twitter about the contents of the report. His definition of public realm data, from a longer exchange we had, is the following:

data collected by sensors/cameras in public places in which there is no opportunity for explicit (or in some cases, even implied) consent for collection

Further exploration around this topic are found in the Sidewalk Labs post, An Update on Data Governance for Sidewalk Toronto, which proposes an “independent Civic Data Trust”. This is one potential form for a proposed entity to oversee public realm data (there are others possible).

So: my understanding of the TRBOT report’s main proposal is that Toronto Public Library, funded by Waterfront Toronto, would take on primary responsibility for the policy development and public consultation on the eventual form of the entity responsible for oversight of public realm data. The TRBOT report uses the term data hub here “as a neutral term to describe the final entity, keeping in mind that the Sidewalk-proposed Civic Data Trust is just one option for this project” (p. 8)

Enforcement of the usage of that public realm data would be under the oversight of the Information and Privacy Commissioner of Ontario, with enhanced enforcement powers. (p. 10)

TRBOT specifically identifies two potential options for the final form of the data hub (p. 8):

What Am I Actually Disagreeing With? (for now)

My immediate response was to the question of whether or not Toronto Public Library should have control over data collected from Quayside in the repository / access control / data management sense, which seemed to be the thrust of the press headlines in response to the report:

This is a position I’d argue against at length - I don’t think TPL has the suitable expertise, capacity, or insulation from political pressure to take on this role. To be fair to TPL, I don’t think public libraries have this generally.

But I actually don’t think that’s a position the TRBOT report takes, though it’s presented as one possibility for exploration (what is called the “ring-fenced Data Repository” in the report).

Rather, the immediate recommendation appears to be that TPL should take on the oversight responsibilities for the policy development (and public consultation, etc) that will eventually produce the public entity that will oversee public realm data collection. That might be inside TPL, it might be the independent Civic Data Trust put forward by Sidewalk, or it might be in some other form.

What is actually being proposed now is, I think, a path forward to getting a data hub in place by having TPL own the consultation and policy development process. Making progress and creating clarity in how corporate entities and otherwise will make use of public realm data in “smart city” projects like Quayside seems to be the report’s immediate objective, rather than recommending a specific form.

I don’t want to argue against a position not taken (there’s too much of that on the internet already), so I’ll highlight the specific things I’d be careful with even at this point:

  1. TPL’s suitability to own the consultation and policy development process around this topic (as opposed to being one participant) are concerns for me, for the same reasons of expertise, capacity and insulation from political pressure.
  2. At the level of optics, Google.org (Google’s philanthropic organization) is a sponsor of at least one library initiative, the wi-fi hotspot lending program. I’m not advocating a purity test for involvement in the process; my research group has been funded for projects by both Sidewalk and Google.org as well. I wouldn’t advocate for us leading the process either, for that and other reasons.
  3. The counterpoint to that first point might be: who would be institutionally more suitable? To that I’d put a similar question that many others have been asking: where is this sense of urgency coming from, what is driving it? It’s hard for me to not see some aspects of this as a desire to break a perceived civic deadlock in order to move the project forward by attaching TPL’s reputational halo to it.

“Well, Alan, what would you like instead?”

I don’t know, exactly, but I feel like at this point, the public outcry over what is (I think mostly fairly) perceived as backroom dealing and an attempt to manage public opinion towards a particular outcome is so great that just handing the process over to another institution to “lead” is doomed to fail. Aside from suitability, I worry about damage to TPL’s reputation in the city (very much justifiably high in certain areas) if it were to agree to lead this process.

I suppose in my ideal world this process, if it needs to happen, would be overseen by a custom-created entity (perhaps created through municipal, provincial or federal legislation) that can draw on a multitude of perspectives, have its transparency better assured through specific organizational and technical practices, and be perceived as more independent from political interference.

One of the things the BiblioTech report does well is highlight that while the gravitional pull of this conversation is because of Sidewalk and Quayside, the issue of managing public data is bigger than that; there is a need to grapple with these issues generally, not just in the Sidewalk/Quayside context.

But as Bianca Wylie, Saadia Muzaffar and many others have pointed out far better than I can, for any of this to have legitimacy, the process must both genuinely be democratic and be perceived by those affected to be democratic. Shuffling another institution in to “lead” would most likely simply continue to replicate the existing problems.

Disclaimers and Disclosures

  1. I worked for Toronto Public Library for nearly a decade, until mid-2015, starting out as a public service librarian. From 2009 onwards, I was a member of the digital services team; I had involvement in numerous projects touching on open data, data privacy and related topics as a senior technical staff member.
  2. Post-employment at TPL, I’d describe myself as a friendly critic of the library’s technology choices to the extent they remain visible to me as a member of the public. Specifically, those interested can refer to these two posts to understand more of the background on my opinions:
  3. My current place of employment, the Inclusive Design Research Centre at OCAD University was contracted by Sidewalk Labs to run a number of public co-design sessions related to inclusive cities. I had peripheral involvement with this project.

All of the above to say that I have a perspective on these issues shaped by my particular experiences.