Ontario’s Open Data Policy: The Good, The Bad, The Ugly and the (Missed?) Opportunity

Yesterday the province of Ontario launched its Open Data portal. This is great news and is the culmination of a lot of work by a number of good people. The real work behind getting open data program launched is, by and large, invisible to the public, but it is essential – and so congratulations are in order for those who helped out.

Clearly this open data portal is in its early stages – something the province is upfront about. As a result, I’m less concerned with the number of data sets on the site (which however, needs to, and should, grow over time). Hopefully the good people in the government of Ontario have some surprises for us around interesting data sets.

Nor am I concerned about the layout of the site (which needs to, and should, improve over time – for example, once you start browsing the data you end up on this URL and there is no obvious path back to the open data landing page, it makes navigating the site hard).

In fact, unlike some I find any shortcomings in the website downright encouraging. Hopefully it means that speed, iteration and an attitude to ship early has won out over media obsessive, rigid, risk adverse approach governments all to often take. Time will tell if my optimism is warranted.

What I do want to focus on is the license since this is a core piece of infrastructure to an open data initiative. Indeed, it is the license that determines whether the data is actually open or closed. And I think we should be less forgiving of errors in this regard than in the past. It was one thing if you launched in the early days of open data two or four years ago. But we aren’t in early days anymore. There over 200 government open data portal around the world. We’ve crossed the chasm people. Not getting the license right is not a “beta” mistake any more. It’s just a mistake.

So what can we say about the Ontario Open Data license?

First, the Good

There is lots of good things to be said about it. It clearly keys off the UK’s Open Government License much like BC’s license did as does the proposed Canadian Open Government License. This means that above all, it is written in plain english and is easily understood. In addition, the general format is familiar to many people interested in open data.

The other good thing about the license (pointed out to me by the always sharp Jason Birch) is that it’s attribution clause is softer than the UK, BC or even the proposed Federal Government license. Ontario uses the term “should” whereas the others use the term “must.”

Sadly, this one improvement pales in comparison to some of the problems and, most importantly the potentially lost opportunity I urgently highlight at the bottom of this post.

The Bad

While this license does have many good qualities initiated by the UK, it does suffer from some major flaws. The most notable comes in this line:

Ontario does not guarantee the continued supply of the Datasets, updates or corrections or that they are or will be accurate, useful, complete, current or free and clear of any possible third party copyright, moral right, other intellectual property right or other claim.

Basically this line kills the possibility that any business, non-profit or charity will ever use this data in any real sense. Hobbyests, geeks, academics will of course use it but this provision is deeply flawed.

Why?

Well, let me explain what it means. This says that the government cannot be held accountable to only release data it has the right to release. For example: say the government has software that tracks road repair data and it starts to release it and, happily all sorts of companies and app developers use it to help predict traffic and do other useful things. But then, one day the vendor that provided that road repair tracking software suddenly discovers in the fine print of the contract that they, not the government, own that data! Well! All those companies, non-profits and app developers are suddenly using proprietary data, not (open) government data. And the vendor would be entirely in its rights to go either sue them, or demand a license fee in exchange of letting them continue to use the data.

Now, I understand why the government is doing this. It doesn’t want to be liable if such a mistake is made. But, of course, if they don’t want to absorbe the risk, that risk doesn’t magically disappear, it transfers to the data user. But of course they have no way of managing that risk! Those users don’t know what the contracts say and what the obligations are, the party best positioned to figure that out is the government! Essentially this line transfers a risk to the party (in this case the user) that is least able to manage it. You are left asking yourself, what business, charity or non-profit is going to invest hundreds of thousands of dollars (or more) and people time to build a product, service or analysis around an asset (government data) that it might suddenly discover it doesn’t have the right to use?

The government is the only organization that can clear the rights. If it is unwilling to do so, then I think we need to question whether this is actually open data.

The Ugly

But of course the really ugly part of the license (which caused me to go on a bit of a twitter rant) comes early. Here it is:

If you do any of the above you must ensure that the following conditions are met:

  • your use of the Datasets causes no harm to others.

Wowzers.

This clause is so deeply problematic it is hard to know where to begin.

First, what is the definition of harm? If I use open data from the Ontario government to rate hospitals and the some hospitals are sub-standard am I “harming” the hospital? Its workers? The community? The Ministry of Health?

So then who decides what the definition is? Well, since the Government of Ontario is the licensor of the data… it would seem to suggest that they do. Whatever the standing government of the data wants to decree is a “harm” suddenly becomes legit. Basically this clause could be used to strip many users – particularly those interested in using the data as a tool for accountability – of their right to use the data, simply because it makes the licensor (e.g. the government) uncomfortable.

A brief history lesson here for the lawyers who inserted this clause. Back in in March of 2011 when the Federal Government launched data.gc.ca they had a similar clause in their license. It read as follows:

“You shall not use the data made available through the GC Open Data Portal in any way which, in the opinion of Canada, may bring disrepute to or prejudice the reputation of Canada.”

While the language is a little more blunt, its effect was the same. After the press conference launching the site I sat down with Stockwell Day (who was the Minister responsible at the time) for 45 minutes and walked him through the various problems with their license.

After our conversations, guess how long it took for that clause to be removed from the license? 3 hours.

If this license is going to be taken seriously, that clause is going to have to go, otherwise, it risks being a laughing stock and a case study of what not to do in Open Government workshops around the world.

(An aside: What was particularly nice was the Minister Day personally called my cell phone to let me know that he’d removed that clause a few hours after our conversation. I’ve disagreed with Day on many, many, many things, but was deeply impressed by his knowledge of the open data file and his commitment to its ideals. Certainly his ability to change the license represents one of the fastest changes to policy I’ve ever witnessed.)

The (Missed?) Opportunity

What is ultimately disappointing about the Ontario license however is that it was never needed. Why every jurisdiction feels the need to invent its own license is always beyond me. What, beyond the softening of the attribution clause, has the Ontario license added to the Open Data world. Not much that I can see. And, as I’ve noted above, it many ways it is a step back.

You know data users would really like? A common license. That would make it MUCH easier to user data from the federal government, the government of Ontario and the Toronto City government all at the same time and not worry about compatibility issues and whether you are telling the end user the right thing or not. In this regard the addition of another license is a major step backwards. Yes, let me repeat that for other jurisdictions thinking about doing open data: The addition of another new license is a major step backwards.

Given that the Federal Government has proposed a new Open Government License that is virtually identical to this license but has less problematic language, why not simply use it? It would make the lives of the people who this license is supposed to enable – the policy wonks, the innovators, the app developers, the data geeks – lives so much easier.

That opportunity still exists. The Government of Ontario could still elect to work with the Feds around a common license. Indeed given that the Ontario Open Data portal says they are asking for advice on how to improve this program, I implore, indeed beg, that you consider doing that. It would be wonderful if we could move to a single license in this country, and if a partnership between the Federal Government and Ontario might give such an initiative real momentum and weight. If not, into the balkanized abyss of a thousand licenses we wil stumble.


Original post

Leave a Comment

Leave a comment

Leave a Reply