Super-Secret Census Data Assisting Researchers
Some Canadians are voicing concern over Statistics Canada contracting the Canadian division of US company Lockheed Martin to supply scanners to read some forms in the May 16 national census.
But the Canadian government used American technology to help it crunch census data as far back as 1891 says a University of Victoria historian involved in a national research project to make past census data available for future public use.
“In 1891, the government used the hollerith, a punch card tabulating machine, for the first time instead of tabulating data by hand,” says Peter Baskerville of the Canadian Century Research Infrastructure Project, who says concerns over accuracy, and security, remain paramount in the current multi-year, multi-university initiative. “Anyone working on this project must undergo a criminal record check and swear a confidentiality oath to Statistics Canada which regularly checks our facility to ensure it’s secure. Any data we send to Statistics Canada undergoes military-level encryption to ensure the information remains secure. We operate on a closed network.”
Baskerville and his UVic colleagues are converting data from microfilm and microfiche of the 1911, 1921, 1931, 1941, and 1951 censuses to a computer-compatible format. The team is also putting the data in context—finding a consistent way to interpret ever-changing terminology by referencing newspapers, parliamentary debates, even cartoons.
“The 1901 census was the last one to use the term “colour” and the 1911 census was the last one to use “crazy or lunatic” for example,” says Baskerville. “If a user of our public samples wants to understand the meanings behind such usage we provide data from debates of the time in machine readable form. We are also making it possible for users to compare, in a systematic way, information across the early 20th century censuses. As well, we are providing maps for all censuses so that the information can be analyzed from a geographic perspective.”
The UVic team is responsible for entry of census data north and west of Ontario. Once completed, the data is sent to Statistics Canada for “anonymizing,” to strip away any information that could lead to individual identification. Data entry for 1911 and 1921 is complete and will be available to the public in a format it can use on line by 2007.
The $15 million funding for the project (UVic’s share is $2.1 million) is from the Canada Foundation for Innovation, the first time the agency has provided major funding to a non-science project.
“From a historical perspective, nothing can match the information provided by this census data,” says Baskerville. “It’s this kind of data that informs government and influences their policies. For disadvantaged groups, it’s a mechanism to prove they exist and deserve attention. A lack of accurate census data is a tragedy for Canadian society.”
But the Canadian government used American technology to help it crunch census data as far back as 1891 says a University of Victoria historian involved in a national research project to make past census data available for future public use.
“In 1891, the government used the hollerith, a punch card tabulating machine, for the first time instead of tabulating data by hand,” says Peter Baskerville of the Canadian Century Research Infrastructure Project, who says concerns over accuracy, and security, remain paramount in the current multi-year, multi-university initiative. “Anyone working on this project must undergo a criminal record check and swear a confidentiality oath to Statistics Canada which regularly checks our facility to ensure it’s secure. Any data we send to Statistics Canada undergoes military-level encryption to ensure the information remains secure. We operate on a closed network.”
Baskerville and his UVic colleagues are converting data from microfilm and microfiche of the 1911, 1921, 1931, 1941, and 1951 censuses to a computer-compatible format. The team is also putting the data in context—finding a consistent way to interpret ever-changing terminology by referencing newspapers, parliamentary debates, even cartoons.
“The 1901 census was the last one to use the term “colour” and the 1911 census was the last one to use “crazy or lunatic” for example,” says Baskerville. “If a user of our public samples wants to understand the meanings behind such usage we provide data from debates of the time in machine readable form. We are also making it possible for users to compare, in a systematic way, information across the early 20th century censuses. As well, we are providing maps for all censuses so that the information can be analyzed from a geographic perspective.”
The UVic team is responsible for entry of census data north and west of Ontario. Once completed, the data is sent to Statistics Canada for “anonymizing,” to strip away any information that could lead to individual identification. Data entry for 1911 and 1921 is complete and will be available to the public in a format it can use on line by 2007.
The $15 million funding for the project (UVic’s share is $2.1 million) is from the Canada Foundation for Innovation, the first time the agency has provided major funding to a non-science project.
“From a historical perspective, nothing can match the information provided by this census data,” says Baskerville. “It’s this kind of data that informs government and influences their policies. For disadvantaged groups, it’s a mechanism to prove they exist and deserve attention. A lack of accurate census data is a tragedy for Canadian society.”
-- 30 --
Media contacts
Peter Baskerville (History) at (250) 472-4360 or pab@uvic.ca