Historical Census Data
Go Digital

December 3, 2003

By Stacey Curry Gunn

Old census records become digital data in U of G's new Historical Data Laboratory in the MacKinnon Building. Members of the census project include, from left, co-ordinator Jean Dalgleish, research assistant Christina Wakefield, Prof. Kevin James, Prof. Douglas McCalla and research assistant Sarah Purton.
Photo by Martin Schwalbe

Cramped handwriting in black ink fills the rows of boxes, capturing the facts for posterity. In 1891, Amelie Carson was a 38-year-old widow of the Catholic faith who lived with her eight-year-old son, Joseph, in a two-storey wood frame house in Hamilton. Born in Ontario to parents from Ireland, she was a grocer who supplemented her income by taking in two lodgers, Helena O'Callaghan, 55, and Freddie Cronan, 13. All members of the household could read and write.

Carson was one of 2,114,000 people across Ontario who were enumerated for the 1891 Census of Canada. And she has a one-in-20 chance of becoming part of a 21st-century research initiative that's now under way at U of G.

In the summer, a team of faculty, students and staff from the Department of History and the Department of Economics launched the University's new Historical Data Laboratory. Their collective mission is to turn five per cent of the handwritten 1891 census information, already available to the public on microfilm, into digital data that can be easily crunched any number of ways to answer wide-ranging social and economic questions about that era.

The cross-college, cross-disciplinary team expects to complete the Ontario portion of the census by 2004 and, if funding continues, to cover the rest of the country by 2008.

The end result, called a public-use microdata sample, will open an important window to Canada's past, says Prof. Kevin James, History, who directs the census project with Prof. Kris Inwood, Economics.

"It's a fun project, even though it's very detailed and we have to be quite pernickety," says James. "We're trying to create a database through the most accurate transcription of what was on the enumerator's page in 1891. Digitizing this information and making it accessible in new ways will greatly encourage its use."

U of G's census project will close a key gap in the continuum of census microdata samples that now stretches from 1851 to the present. Other Canadian universities are working to fill in other gaps, with the result that digitized data sets are complete or under way for all census years except 1861 and 1961. A multi-university team led by the University of Ottawa is currently creating parallel microdata samples for the 1911 to 1951 censuses and has greatly aided the Guelph team's work, Inwood says.

Canada's efforts to build this digital research infrastructure follow the example set by the United States, where the use of census microdata samples for research is far advanced. That country has microdata samples for each of its 19th- and 20th-century censuses, except for 1891 because those records were lost in a fire.

That loss has U.S. researchers anxiously awaiting the results of Guelph's census project, says Inwood. "They look at our 1891 and get really excited. They can use it to fill in what happened in the States."

Guelph stepped forward to take on the 1891 census as part of the mandate of Prof. Douglas McCalla's Canada Research Chair in Rural History. Funded by the Canada Foundation for Innovation, the project advances several research agendas at U of G.

The 1891 census is considered particularly important because of the type of information it contains and its timing as "the last census year before the Canadian population explodes with a major wave of immigrants," says Inwood. It will provide the final set of benchmark data (following the 1871 and 1881 microdata samples that are already available) to show trends for the era before rapid population growth transformed the country.

Historical census data are valued for two key reasons. They provide a complete and accurate picture of the whole of society at a point in time, and the information they contain is no longer deemed confidential (after 100 years). That allows researchers to link the names in the census with other sources of information.

"The census is one of the very best sources for catching everybody in society in a way completely necessary to the work I do, which seeks information on people and processes invisible in many standard kinds of documents," says McCalla. He is writing a history of Canadian settlement from the perspective of farm families, rather than the fur traders who have been at the centre of early Canadian economy history.

"This will offer us enormous insight into the structure of rural Canada in 1891," he says.

The 1891 census is also prized for the detailed way it enumerates occupation, a plus for Inwood, who has a special interest in women's participation in the economy.

"It's clearly interesting in terms of recapturing our past and understanding where we're going, which you can't see unless you have long-term perspective," he says.

James's ongoing research into Irish and Scottish migration patterns will benefit from census information on the birthplace of each individual's parents, recorded for the first time in 1891.

The census is also the first to record the relationships between people living together, something that will help researchers make more accurate assumptions about the patterns in previous census years.

In the Historical Data Lab on the eighth floor of the MacKinnon Building, a handful of history and economics students painstakingly transfer information from microfilm to the database, under the supervision of project co-ordinator Jean Dalgleish.

A software program developed by the University of Minnesota randomly selects the census pages that become part of the sample. After data are entered, they must be verified and any errors fixed.

"We try to establish the accuracy of input by testing it," says James. "The students are looking for everything from typos to questions of interpretation posed by the data where information recorded might be ambiguous or unclear. We found one situation, for example, where a two-year-old child was listed as a farmhand."

The work teaches undergraduate and graduate students "incredibly useful" skills, says McCalla. "They're learning a lot about a source they can use across a wide range of research projects, and in the process, they learn about censuses in general and about working with other kinds of large data sets."

One history undergraduate student, Christina Wakefield, found information about her cousin's great-grandfather, Benjamin Beley, who was one of the first people to settle in Humphrey Township around Lake Rosseau.

"I grew up hearing stories about his life in Canada, so it was really exciting to find him and his family in the census," she says. "It reminds me that the names we enter into our database were real people and not just data."

On completion, the 1891 census microdata sample will be publicly available from the McLaughlin Library's Data Resource Centre.

It will be one more piece in an ever-expanding puzzle as more and more countries around the world turn their own censuses into digital data. Countries in Latin America, Europe and Asia are also engaged in census projects, which will eventually enable international comparisons.

Inwood says the surge in interest and activity is due partly to advances in information technology, which make these projects possible, and partly to the efforts of American census experts "reaching out to the rest of the world to create parallel databases."

Indeed, Guelph's role in opening this new frontier of academic research comes from many layers of collaboration - between disciplines and departments, between universities and countries - a massive joint effort to shake big-picture patterns and perspective from the diverse lives of people like Carson and Beley.

"It couldn't happen for each of us alone," says Inwood. "It's the bringing together of everyone that's been so fruitful. It has become an international movement to recapture this kind of evidence about the ways people lived."