Government Information in Canada/Information gouvernementale au Canada, Volume 3, number/numéro 2 (fall/automne 1996)
Comments on "Data Liberation and Academic Freedom" 1, 2

John C. Courtney 3
Department of Political Studies
University of Saskatchewan

The essence of the issue being addressed in this session can be simply stated: it is to ensure full access for teaching and academic purposes to government data sets and information at affordable costs.

For the academic community, prohibitively high costs for information retrieval of government documents and data sets means precisely what the term says: high costs effectively prohibit academic freedom--that is, the freedom to pursue without arbitarary constraint subjects of intrinsic scientific value.

Fortunately for the academic community, the Data Liberation Initiative (DLI) is now nicely underway, albeit with a somewhat uncertain future as a five-year pilot project. Jointly sponsored as it is by a variety of agencies and research libraries, the DLI is a bold and sensible undertaking. It is to be hoped that it works, that its goals are achieved, and that it is able to prove to a sometimes sceptical government apparatus the worth of university-based research. More to the point, it is to be hoped that if it truly does amount to a success story (and all indications point in that direction), its funding and its future are both assured well beyond its current five-year life.

I should note as well that the DLI stands as proof of the effectiveness of academic lobbying. Were it not for the role played by the then Social Science Federation of Canada (SSFC) in pursuing the issue with such vigour, it is an open question whether we would today have the DLI or anything even approximating it. I commend to you the report the Federation produced in support of its lobbying and negotiating efforts. We do not often have occasion to point to success stories in making our case with government, but the establishment of the DLI stands as a happy exception to the all-too-usual pattern of failures.

The strengths of the DLI speak for themselves and need no elaboration. There are, however, some aspects of the initiative that bear directly on the larger matter of ensuring guaranteed access for the university community (and, for that matter, the larger public) to some of the essential tools for scientific research. I wish to comment on these and to use the occasion to raise a further concern I have about a related issue of confidentiality of information available through electronically-based data sets.

The immediate issues that this initiative raises in my mind center around four concerns. They all derive from the fundamental proposition that government has an obligation to inform the public of its activities and the public has a right to be informed of those activities.

I raise these in no particular order of priority:

  1. Having information available in electronic form on computers and networks is fundamentally different from having it available in written form on library shelves. Access to information becomes an issue when, by definition, the publics being served by the different means of delivery of that information are differently defined. The question that we as university-based researchers must ask ourselves is one of adequacy of training of future researchers. Are we training electronically-literate students who can cope with the realities of the data and other material that increasingly are becoming available only in electronic form? In the words of the information highway buffs, "If you’re not part of the steamroller, you’re part of the road." 4 As a political scientist I have this year undertaken for the first time a special series of lectures and graded assignments based on retrieval of information from print and electronically-available Canadian government sources for my Introductory Canadian government class. Without the assistance of Andrew Hubbertz and his staff in the Government Publications Section of the University of Saskatchewan Library, this valuable assignment would not have been possible--a fact that reinforces the critical importance of intra-university cooperation when it comes to introducing new technologies to our students.

  2. Related to my first concern is the question of size of audience reached. Any of us can go into the government publications section of a library and retrieve, for example, StatsCan census data from the shelves. Will that continue to be the case with the shift to exclusively electronically-available material that is contractually to be made available only to university-based researchers? If not, have we further widened the gap between university-based research and the wider interests of the general public? Is this, in turn, likely to create not only the perception but the reality of yet another "knowledge elite"? Contained in that question is an important subtext: what of the climate within which university-based scholars pursue their endeavours? Will this change by creating an added dimension of suspicion, distrust and envy among non-university based users? I understand that the Government of Saskatchewan pays 100% of the billed cost of some federal data sets whereas the University of Saskatchewan obtains the same information at a 90% discount.

  3. What will be the impact on collaborative research between university and non-university based researchers when, as I understand it, data are available only under licence with institutions for non-commercial use? There is a vagueness, at least in my mind, in the arrangements that could be established between university and commercial users when they are involved in collaborative work. In turn, this could effectively establish a barrier to important joint research projects.

  4. What is the precedent that has been set by the DLI arrangement? While understandably welcomed in solving the immediate problem of prohibitively costly information retrieval arrangements, the DLI does accept the principle of user pay, however small that amount may be in the pilot project. That principle is at odds with the much older one that governed the arrangements under the Depository Services Programme (DSP). If the DLI works well beyond its initial five-year term, and one can only hope that it will, it nonetheless becomes a model for access to information in the future. That, in turn, places the future of the DSP in some doubt. One can only hope that there is no willingness to let the success of one new initiative negatively affect the continuation of an older, and no less successful, program.

Let me close with a brief expression of concern about confidentiality of information available on electronically-generated data sets. I understand a proposal currently making its way from Elections Canada to the federal cabinet pertains directly to the possibility of replacing the current enumeration and revision process of compiling voters' lists with a so-called "permanent" list of electors. Quite apart from the issue of whether or not this would be a cost-effective move that, in the end, would produce a more complete list of voters, it strikes me from what I have able to learn so far that one side of the proposal could prove problematic.

That has to do with the lists that themselves would be used to provide names, ages and addresses of Canadian citizens eligible to vote. Some Ottawa sources suggest that the federal voters' lists might be kept up-to-date on a continuous basis (and therefore the mistaken rubric "permanent") by accessing other data sets. These are said to include tax records, health cards and drivers' licences. I would hope not, for people enter those registries in order to comply with tax laws or obtain medical and hospital benefits or drive a car--not to vote.

The information available to one government can easily be transmitted to another electronically; my fear stems from the fact that in this particular instance there is no demonstrable relationship between the information provided to one agency and the needs of the another. Invasion of privacy (a concern the Privacy Commissioner and Revenue Canada would no doubt share when they examine this issue) is a distinct possibility with electronically available information. There is in my mind a certain parallel here of which we must remind ourselves. Just as the researcher (university-based or not) must enjoy the freedom to pursue without arbitrary constraint subjects of intrinsic scientific value, so too must the public enjoy the freedom to know that information provided to one government department is not to be shared heedlessly with another. Both parts of that equation rest, I would hope, on a certain irrefutable logic.


[1] May be cited as/On peut citer comme suit:

Courtney, John C., "Comments on 'Data Liberation and Academic Freedom,'" Government Information in Canada/Information gouvernementale au Canada 3, no. 2 (1996). []
Back to text.

[2] This article is based on a paper presented at "Academic Freedom: The History and Future of a Defining Idea," September 21, 1996, in Saskatoon, Saskatchewan.
Back to text.


John Courtney
Department of Political Studies
University of Saskatchewan
909 Arts Building
9 Campus Drive
Saskatoon, SK
S7N 5A5 Canada
Back to text.

[4] An expression used in Science Bulletin 6 (December 1994): 7.
Back to text.