Thank you very much.
I want to thank you for inviting me here. I'm very pleased that Parliament is conducting this study and I am honoured to contribute to it.
I'm here because about six months ago I launched a site called openparliament.ca. I know that some of you have seen it and have said nice things about it, which I appreciate very much. But for those of you who aren't aware of it, it's a site that tries to make it easier to keep tabs on some of the things that are going on in the House and tries to make such things as Hansard a little more engaging and useful. You all have pages on the site that show anything you've said recently on the House floor, along with media coverage, votes, any legislation you've introduced, and so on. It's all searchable. You can sign up for e-mails or updates when a given MP speaks, a bill is discussed, or a particular keyword is mentioned.
I made it as a volunteer, spare time project and I'm hugely pleased that people have found it useful and that it is used by tens of thousands of Canadians each month.
I should say that I've never worked for, in, or even really with government. So if I'm going to talk about open government, the subject of this study, it will be very much from an outsider's perspective.
Open government is a fairly vague term that has meant many things over many years, but the current usage—and you'll also hear “Government 2.0” as a synonym—means to me the idea that recent advances in technology can enable a government that is more engaged, collaborative, cooperative, and better able to spark certain kinds of innovation. This is certainly an appealing notion to me, and I hope to you, but it's also a bunch of fairly vague and happy words that would be fairly difficult to disagree with.
So to talk about something more concrete in an area in which I have at least a little knowledge, I'll focus on one particular idea, that of open data.
Let me quote Australia's Government 2.0 task force, whose absolutely excellent report I'd really recommend you look at: “...public sector information is a national resource and...releasing as much of it as possible on as permissive terms as possible will maximize its economic and social value and reinforce [its contribution] to a healthy democracy.”
Let me turn to data. When many people hear the word “data”, their eyelids start to grow heavier, their shoulders start to slump. I think that's a pity. When I hear data, I get excited. To me, data means possibility, it means opportunity, it means discovery. I really hope I can share with you at least some small part of that sense of excitement.
Let's take care of definitions. When I say data, I mean big piles of information structured so that computers can make sense of it—like Hansard, like pollutant inventories in industrial safety reports, like bus schedules, like satellite imagery, like the list of registered charities and their public filings, like government-funded scientific papers, like digital maps and details in the postal code system, like records of prescribed drugs and of disease occurrence. There are endless examples. If you ask anyone working in technology how great the value of data can be, the answer you'll get is immense.
Increasingly the Internet economy is driven by companies working to figure out how to extract value from data. Ray Ozzie is a computer legend—I'm a computer programmer, I should note—and currently a leader at Microsoft. Let me quote him: “Data is the flint for the next 25 years.”
A corollary to that is that the value of data is often not apparent at first. Less than a decade ago, many people didn't think that web search data, which you type into the little search box in your computers, was all that valuable. Companies offered web search, of course, but often as a sort of loss leader. Then Google came along and realized that in fact this web search data was worth many billions of dollars a year.
Several studies have attempted to measure the value of government data. A European study put the market size for the used public sector information at 27 billion euros, and other reports have come up with similarly staggering numbers. It's tremendously valuable, this data, and to lots of different groups: to those interested in public policy, whether researchers or, as is increasingly possible, just engaged citizens; to businesses in all manner of industries; to civic-minded Canadians, who in some ways have a new way of engaging with government and building things to help each other out; and to government, where data is of course used in planning and programs, which can now have the possibility of benefiting from much of this external innovation.
If we recognize the value of government data, it's the “open” in open data that allows that value to be unlocked. The words means many things, and I think it's worth the time for me to quickly try to unpack the phrase.
The first is that it be--and there is a piece of jargon coming up, but I promise it's important--machine-readable.
Let me use my site to explain what I mean by that. I republish Hansard, which I get from the website of Parliament, but because of the way it's made available, getting the data out so that I can republish it is difficult, and it took quite a bit of time and trickery on my part. The methods I use are fragile, so if Parliament changes the look or format of its site--or your site, I suppose--mine breaks. Because getting the data out is difficult, it's much harder to do all sorts of things, such as making my site fully bilingual--which it isn't--or reporting on committees like this one.
In my case this isn't the end of the world. The hurdles have caused plenty of frustration, but the site nonetheless exists. Often data that aren't machine-readable are simply too difficult to make productive use of. To make data available in a machine-readable way, which is more conducive to exploration, is for the most part not hard from the point of view of technology. The roadblocks here are matters of will and of culture.
“Open”, in the context of open data, also means “free”. In English that's one word, but it means two very important things.
In order to be truly useful, government data should be free of cost and accessible.
They should be free of cost because that's how they'll most efficiently create economic value and support innovation, and because sharing information that already exists over the Internet costs government next to nothing. They should be free, as in speech, by which I mean available under terms that allow repurposing and redistribution, which is exactly where the greatest value lies.
I want to stress that repurposing isn't an addendum or a blue-sky wish list item. Open Parliament is an example, of course, of repurposing, but there are others, even if I just restrict myself to my own life.
Here's one: I lived in New York for a while, where this wonderful website called EveryBlock takes the mundane details of municipal government--building permits, business permits, restaurant inspections, crime reports, municipal hearings--and repurposes them to publish a newspaper for your own block. What's done in aggregate becomes interesting when it's filtered for what's relevant and nearby, and doing that makes participating in local government that much more likely.
Here's another personal example: I studied public health briefly. What one often does in that field is try take disparate data sets--let's say cancer incidence and pollutant release--and try to combine and repurpose them to generate hypotheses to make Canadians healthier.
That's why it's crucial for the default posture of government to be sharing, rather than presenting a closed door, and why the default terms for government information should be an open licence similar to Creative Commons and several others, and not the current innovation-killing restrictions of crown copyright.
Anyone who's tried to work with government data has come up against that default posture of a closed door. Here are some personal examples: under what's called the Speaker's permission, I'm able to republish the Hansards of the House, but that's not true of the Senate. Republishing that is illegal. Your official photographs are under crown copyright, and I was unable to get permission to use them. On the municipal level, in Montreal I've tried to get digital maps of the city's political districts, les arrondissements, and I've tried to get bus schedules, and I've been rebuffed, despite knowing that it would take literally five minutes to send me the latter in computer-readable form.
A friend in Halifax tried a few years back to get the same information there--bus schedules, maps, and political districts--and was also denied, but he went further. An access to information request was denied and a court challenge was denied, with one of the grounds for denial being that the digital map was not a document but a mechanism for producing documents.
Needless to say, I disagree with the decision, but there's something in that phrase that catches my interest: a mechanism for producing information. Yes, in an information economy that's more or less the point: a means of creating information. Just last Saturday thousands of people across the world and hundreds in Canada got together for what was called an open data day, an event spearheaded by David Eaves in Vancouver and a fantastic team from right here in Ottawa.
I worked with a group in Montreal. Some twenty people worked on a dozen or so municipal projects, in particular on a site announcing which municipal rinks are open and have been flooded, and on a system to advise drivers about new road construction they might encounter on their regular route.
Digital maps are a means of creating information. That's borne out by the experience of Natural Resources Canada, NRCan, the only federal department with a thriving open data culture. Its geographic data sets are used by a huge community of researchers, who will tell you just how valuable they are. They're used heavily by industry, including mining and forestry, of course, but also real estate developers and burger joints, and even by me. I used data from NRCan's GeoGratis program for a recent project related to finding polling places in municipal elections.
I really believe this means of creating information and sparking innovation is one of the strongest arguments for an open government policy and release of open data. There are many points of view from which to argue for open data, and from there for the broader concept of open government. I've given you only one. For example, I haven't really mentioned accountability and transparency, which are very useful things in their own right, but I hope that I've been able to communicate at least some of the excitement that I and increasing numbers of Canadians feel about this, and I hope that that Canada will join what is a growing and promising worldwide movement.
Thank you.
