The case for keeping old census data secret is thin
By Dean Jobb

BACK IN 1857, when a Hants County jury convicted a Catholic railway worker
        of murdering a Protestant, Nova Scotia was as troubled as Northern Ireland.

        So while researching the trial some years ago, I wanted to find out whether - as
        observers of the day suspected - sectarian hatred had prompted jurors to ignore
        clear evidence of self-defence.

        The search led to raw data collected in 1871 for the first national census, which is
        accessible on microfilm through the public archives. Nine of the 12 jurors turned
        up and, sure enough, eight were Protestant.

        A case could be made that religion had entered the jury box - a timely reminder,
        given Canada's growing list of wrongful convictions, that prejudice can taint the
        pursuit of justice.

        After my account of the case was published, not one juror phoned to complain
        about becoming famous at the age of 160 or more. And I've yet to hear from a
        juror's descendant.

        Yet the federal government is determined to stamp out this kind of basic historical
        research to protect the privacy of people long dead.

        Traditionally, census forms have been made public after 92 years. Information
        that Ottawa collected in 1871, 1881, 1891, and 1901 is in the public domain, as
        are pre-Confederation provincial censuses, including Newfoundland's 1945
        count. The "tombstone" information on the forms - name, age, address,
        birthplace, marital status - allows descendants and professional genealogists to
        trace family trees. Scholars use the records to study family structure, population
        movements and social trends. For local historians, they are building blocks to
        reconstruct a community's past.

        Despite the value of these data, Statistics Canada is determined to prevent all but
        a handful of people from using the 1911 and later federal counts, as well as the
        1906 Western Canadian census.

        It has withheld the 1906 records and wants to hold back the 1911 data, due out
        in 2003, on the strength of a strained legal opinion.

        By 1906, to gain the trust of new immigrants who were wary of state officials
        asking questions, census takers were instructed to assure respondents that the
        information would not be used for taxation or other purposes.

        StatsCan sees this as an iron-clad promise of privacy, even though enumerators
        were also told to "write legibly" because the questionnaires were destined for the
        national archives and release, as required by law, after 92 years.

        Ottawa set up a panel of scholars to review the issue. Its report, filed in
        December 2000, found no evidence that census records were to remain secret
        forever. It called for release of 1906 and 1911 census data and legislative
        changes to allow the release of later censuses after 92 years.

        The British release census data after 100 years, the Americans after 72. Census
        records of 620 million Canadians, Americans, and Britons have been made
        public, the panel noted, without a single complaint about invasion of privacy.

        StatsCan, however, is determined to get the response it wants. It has retained the
        Environics polling firm to hold town-hall meetings and focus groups across the
        country.

        The archival roadshow rolled into Halifax this week. While a couple of official
        types spoke of the importance of protecting privacy, applause was reserved for
        those advocating openness.

        Judging by the abundance of silver hair at the afternoon session, those who want
        the data released are a lot closer to seeing their census records become public
        than I am.

        StatsCan is using the sessions to float a proposal for limited access. People could
        apply for access to trace direct ancestors. Historians could take a look if a
        peer-review committee endorses their methodology.

        The restrictions, Environics is being told, are unworkable. Genealogists must
        identify great uncles, cousins and the like to complete a family tree.

        And I doubt an amateur historian who's curious about the composition of an old
        jury can convince a group of tweed-jacketed bureaucrats that the research has
        merit.

        Faced with an unjust verdict in 1857, the Nova Scotia government spared a man
        from the gallows and substituted a two-year prison term.

        StatsCan's so-called compromise, on the other hand, would be a death sentence
        for free access to an invaluable historical resource.

 
 email: djobb@herald.ns.ca