COMPUTER-ASSISTED REPORTING

Canadian journalists who use computer-assisted reporting techniques to search for new stories are often jealous of our American colleagues.

In the United States, it’s easier and cheaper to get data about all sorts of events – everything from municipal police car chases to federal political party campaign finances is available on the web,and often at no charge.

In Canada,the culture is different at most of the public institutions that collect and publish data about our economy and society.

I wrote a column for The Globe and Mail about this issue, taking a look the some of the data sets published by the country’s biggest data collector and publisher,Statistics Canada. I argued that while lots of Statscan data are made available for free,a great deal of important data,particularly the finely detailed sets of numbers that describe Canadian life at the neighbourhood level,are made available only to those willing to pay a great deal.

Companies in Canada that collect, sell and market this kind of information say that if you want to get detailed figures on household income,dwelling types, education and other variables on a street-bystreet, across-the-country basis,you could pay Statscan more than $10,000 for the privilege.

By contrast, the same kind of data,with similar amounts of detail, for the United States can be purchased from the U.S.Census Bureau for as little as $100.

There is also a great disparity between Canada and the U.S. for the cost of digital versions of street maps.A Canadian set can cost as much as $25,000 while a U.S. set costs $2,000 (U.S.).

For several years now,the U.S.set of digital maps included every urban and rural road.With the current census, Canada is finally catching up in this regard. Until this year,the digital maps of Canada contained only the road networks in built-up urban areas. Digitized versions of maps can be combined with the raw neighbourhood data to help reporters get some powerful insights into their communities.

When you combine geographic location information with list of things like houses and people, you get what is called geospatial data. There is a growing hunger for good geospatial data by all sorts of businesses,non-profits and journalists. Geospatial data users are trying to push Canadian policymakers toward the idea that raw data about our country ought to be made available for as close as possible to free, and that such a policy would have immense benefits to the Canadian economy, as well as to journalists.

In the U.S., data collected by the public’s representatives — the government — about the public are viewed as the public’s good.The job of government, in the U.S. at least, is to get this information into the hands of the public with as little fuss as possible. Some studies say that, for every dollar invested in distributing geospatial,census-based data,users of that data generate $4 in growth,mostly by improved resource allocation.

As I wrote in The Globe and Mail, it feels different in Canada. One gets the sense that data collected by our governments are collected so they can make some money selling the information back to us. That’s wrong. The government ought to give the data to us because they belong to Canadians.

In a response to my column,Martin Podehl of Statscan’s marketing and information services branch,took issue in a letter to the editor with my suggestion that the agency is trying to make money by selling data.

“Statistics Canada,”Podehl wrote,“only recovers the cost of providing data that are supplied for specialized private use. The agency has, over many years,steadily increased the amount of data it provides without charge,most notably since the introduction of the Internet.”

“The agency does charge in certain cases.First, we charge for products we would not produce if their cost were not recovered.Second,we charge for custom services in which a data user requests a specific output to meet some individual need.To do otherwise would be unfair to taxpayers at large.”

For journalists, working in increasingly cost-conscious environments, cost recovery initiatives by our government agencies are effective storykillers. After a presentation by the Statscan folks at one of our computer-assisted reporting workshops at the annual CAJ convention in Ottawa,a freelancer wanted to check out one of Statscan’s data sets. Clicking through to the Statscan site, he found that he would have to pay $45 for the data that interested him.

That meant he would have to file an expense claim with no promise that there would be a story he could sell. And so, the cost of these data pretty much squelched this writer’s investigative urges. To their credit,the Statscan folks say that journalists ought to contact them directly to acquire particular data sets. They say that, in most cases, the agency will waive the charge for a journalist.

Again, that’s all well and good,but it still requires phone calls every month to get that information. Wouldn’t it be simpler just to publish the data on the Web for everyone’s use for free?