What I think now...

I had heard that Statistics Canada was discontinuing its E-Stat service. But when I was researching references to the 1831 Census of Lower Canada and realized that it was among the many tables of historical census statistics formerly offered by E-Stat, I thought, “Oh god, what did they do with all those tables?” Luckily, Wilfrid Laurier’s Library website directed me to Queen’s University Library website, where we can find the former E-Stat tables for 1665 to 1871: http://library.queensu.ca/webdoc/ssdc/census-1665-1871 Phew! I found no reference to this transfer within the Statistics Canada website, so I will recommend that they create a weblink re-direct.

Sep 24
Canadian historical census aggregate statistics

…with remarkably little fanfare, I might add, for such an important and detailed census.


To browse the 1921 Census via a geographic index, visit http://www.ancestry.ca/census


Important questions remain:
- Why did this release take so long, given the fact that the images and the geographic index were produced years before the June 1 2013 release date?
- Why so little fanfare—this was a perfect opportunity to get Canadians excited about history and genealogy.
- What is the exact nature of the arrangement LAC made with Ancestry?
- Why did LAC choose Ancestry as a partner?
- Why not involve Canadian academic researchers in such a project and, with some complementary funding from a government funding agency, make it possible to have ALL the questions from the census entered, rather than just the “tombstone” questions (name, age, sex, birthplace, marital status). This would be a great way to leverage government funding to make a “big data set” to be used by genealogists and researchers alike
- What assurances were given for the quality of transcription of French Canadian names, which are often badly transcribed if the data entry operators are not themselves francophone
- Will the 1921 Census images, produced by the CCRI project and paid for by Canadian taxpayer dollars, be made available as a set to other groups interested in indexing?

See also good questions posed by the Ontario Genealogical Society: http://www.ogs.on.ca/ogsblog/?p=4115

Aug 8
1921 Census now available via Ancestry.ca

To sign a petition to release the 1921 Census of Canada, visit: https://www.change.org/en-CA/petitions/the-honourable-james-moore-release-the-1921-census-now

This petition has been signed by 3,198 supporters so far. Now, there have to be far more than 3,198 practicing genealogists in Canada! So get the word out by sharing on your Facebook page, tweeting etc.

Jul 23
Online petition to release the 1921 Census of Canada; just how many genealogists are there in Canada, anyways?

On June 1, 2013, the 1921 Census of Canada entered the public domain; according to Canadian law, this census can now be made freely available. Yet we are now nearing the end of July, and still we await the release of this census by Library and Archives Canada (LAC).


What is really odd about this delay is that rich metadata and image resources pertaining to the 1921 Census of Canada already exist, owing to a previous collaboration of LAC and Statistics Canada (STC) with the Canadian Century Research Infrastructure Project (CCRI), first headed by Dr. Chad Gaffield at the University of Ottawa and now headed by Dr. Peter Baskerville at the University of Alberta. The CCRI project, headquartered at the University of Ottawa from 2002 to 2009 and funded by the Canadian Foundation for Innovation (CFI), created scientific sample databases of the 1911, 1921, 1931, 1941 and 1951 Censuses of Canada for the purpose of statistical analyses at the level of individuals, families and households. The 1921-1951 anonymized databases (with names removed) are currently accessible via STC’s Research Data Centres, while the 1911 database with accompanying documentary metadata is available on a University of Alberta website. The origin of the CCRI project, its goals, methodology, partner collaborations and resulting quantitative and qualitative data are detailed in a 2007 special history of Historical Methods: A Journal of Quantitative and Interdisciplinary History.


One of the first great feats of the CCRI project was to rescue the microfilms of the 1911 to 1951 censuses. At the time, I was a Research Co-ordinator for the Institute of Canadian Studies (ICS), University of Ottawa, and helping to write the CCRI grant application. I remember the day I went to Statistics Canada at Tunney’s Pasture, was duly sworn in as a deemed employee and accompanied to a basement office where these confidential microfilms were held, tucked away from public view. The census manuscripts from which these microfilms were made were long gone, and some of the microfilms were in a fragile state. Historical Methods narrates the CCRI’s subsequent collaboration with both LAC and STC (Gaffield, 2007, 55). One article in Historical Methods describes how the CCRI used digitized images of these microfilms in a sophisticated data entry system based on a geographic index to the census images: “The sample point identification, data-entry and reporting system (SPIDER) … was developed to improve on the established approaches to census microdata creation and to take advantage of the opportunity to use digital images of census schedules rather than the familiar microfilm reels on which surviving Canadian census schedules have been recorded.” (Darroch, Smith and Gaudreault, 2007, 67). Another article in Historical Methods describes how CCRI geographers created geographic information system boundary files based on the census sub-districts for all five census years and digitized a selection of published census tabulation tables (St.-Hilaire, Moldofsky, Richard and Beaudry, 2007, 87).


For many years, in fact, other Canadian scholars have collaborated fruitfully with the public and volunteer sector. The Université de Laval project “Population et histoire sociale de la ville de Québec” collaborated with the Société Généalogique de Québec on data entry of the 1852 Census of Québec City. While at the University of Ottawa, I myself headed a collaboration with the Genealogical Society of Utah (now FamilySearch) to bring their 100% database of the 1881 Census of Canada into the public domain. The CCRI’s partnership with LAC and Statistics Canada then opened the door for my own collaboration with LAC. By 2004, I had taken an academic post at the Université de Montréal where I headed another CFI-funded project to create a sample database of the 1852 Census of Canada. I was able to contact the same LAC mid-level employees I had befriended while working on the CCRI grant application, and create a win-win resource sharing agreement. LAC contributed to my project a full set of images of the 1852 Census of Canada which we used not only for data entry, but also for subsequent checking, cleaning and coding. When the 1852 images arrived, they were a massive lot of sequentially-numbered files, divided only by microfilm number. At the Université de Montréal, three student research assistants toiled for 3 months to go through those numbered files, one by one, dividing up the pages by sub-district and division, setting the “header” and table of contents images aside, and dividing the personal from the agricultural returns. We created meaningful tags for all the images and returned our list of tags to LAC. This tagging is the basis of the geographic index to the 1852 Census available on the LAC website today. My LAC colleagues and I enjoyed a very fruitful collaboration; for a period of time, I was exchanging e-mails on a weekly basis with my counterparts at LAC, trading information about missing sub-districts and labelling issues, as well as identifying local genealogical groups who had started indexing the 1852 census.


Now it is 2013, and those collaborations lie dormant. Yet digitized images of the 1921 Census and the CCRI’s accompanying geographic index still exist. With the geographic index, the 1921 census images could have been speedily mounted on LAC’s website with a basic geographic search engine for June 1. Where are they?


If you visit the “Census” pages on LAC’s website and look carefully at the “Census Search Help” page for each individual census, you can find one sentence at the bottom of each page: “Library and Archives Canada gratefully acknowledges the contribution of Ancestry.ca, without which this project would not have been possible.” Ancestry, a U.S.-based private company which offers “the world’s largest online resource for family history,” provided LAC with a copy of their nominal indices to the 1871, 1881, 1891, 1901, 1906, 1911 and 1916 Canadian censuses which Canadians can search for free on the LAC website. LAC has likely made a similar arrangement with Ancestry for the 1921 Census of Canada. Partnering with Ancestry or a different organization such as FamilySearch brings obvious benefits for genealogists and historians. In the United States, Ancestry has partnered with the Minnesota Population Center at the University of Minnesota to enter all 50 questions on the 1940 census schedule for 100% of the population that year. A press release posted on the University of Minnesota website states “The new database will allow much richer studies of small populations in 1940, such as Dust Bowl migrants to California, Native Americans, and working mothers with young children. Researchers will also be able to link recent economic and health surveys and mortality records to the 1940 database. These linkages will allow researchers to study the impact of early life conditions—including socioeconomic status, parental education, and family structure—on later health and mortality.”


The work of the CCRI as well as the partnership between the Minnesota Population Center and Ancestry indicate that there are rich possibilities for collaboration across sectors, with national institutions like LAC and STC, academic researchers, volunteer genealogical associations and organizations like Ancestry and FamilySearch pooling expertise and resources to produce high-quality population data and metadata. With the 1921 Census of Canada, LAC has let such an opportunity slip away, making its own arrangements with nary a word in advance to the Canadian public, nor to potential Canadian partners. Given LAC’s previous collaborations, there were established pathways for much greater collaboration and transparency on the release of the 1921 Census of Canada. As Historical Methods and another article published in Digital Studies make clear, the CCRI team know intimately the idiosyncrasies of the 1921 census, common enumeration pitfalls, the particularities of each census question, the meaning of myriad Census Bureau tabulation marks overlaid on the manuscript pages, contradictions between the enumerator instructions and enumerator practice, and differences between the geographic units used by enumerators and those used by census tabulators. The image and metadata resources created by the CCRI for the 1921 census were supported by a CFI grant. That’s right, taxpayers have already paid for the digitization, geographic indexing, documentation and georeferencing of the 1921 census, all work which is necessary to launch a nominal indexing effort to capture 100% of the names from the census in a searchable online format. Why was none of this tax-payer funded expertise and technology leveraged for LAC’s current work on the 1921 census? If this expertise had been integrated into a transparent public-private partnership, say, with Ancestry, could LAC have struck a sweeter deal for Canadians, genealogists and academic researchers alike?


What can explain this failure of LAC to leverage all valuable pre-existing partnerships? The previous partnerships between scholars and LAC were negotiated during the period when Dr. Ian Wilson was the chief Librarian and Archivist of Canada. During Dr. Daniel Caron’s tenure as chief Librarian and Archivist, drastic cuts to LAC’s budget coupled with a shift in priorities at the very top led it to privilege partnerships with private industry and to neglect its relationships with the academic community and the volunteer sector. In recent years, LAC has come under fire for numerous dramatic changes, from cancelling the National Archives Development Programme, to eliminating Inter-Library Loans, to slowing down private record collections to focus on government record collection, for eliminating the peers system of merit review and promotion for staff archivists, for removing criteria for LAC managers which require expertise in archival and bibliographic description, and for failing to develop a clear public consultation process. Concomitantly, major staff reductions seem to have clouded institutional memory at LAC of their former collaborations with Canada’s academics.


LAC already possesses microfilm and digital images resources for the 1921 Census of Canada, and perhaps a nominal index as well. By not engaging with a wider variety of partners, LAC has missed an important opportunity to build Canadian capacity in terms of technological development and knowledge transfer across sectors. By not involving representatives from the genealogical and academic community in its project plans, LAC has failed to obtain valuable input on cost estimations, quality control and specific demands which could be written into the agreements. It has also missed an opportunity to leverage additional resources to ensure that all 35 questions from the 1921 census are entered into the database, creating an invaluable social science research file.


While public-private partnerships can pose many advantages given LAC’s tight budget, these partnerships must be undertaken with far greater transparency and with better use of consulting partners who can contribute their own valuable resources to the project. As we await the appointment of a new chief Librarian and Archivist of Canada, LAC has an important window of opportunity to renew its approach to public consultation and collaboration; its next steps with regard to the 1921 census are a chance to send this message.


I call upon newly-appointed Minister of Canadian Heritage Shelly Glover and LAC to
• release the 1921 Census of Canada immediately.
• state in clear detail the deal LAC has made with its indexing partner
• articulate the principles which guided this deal
• create a consultative committee comprised of representatives from the historical, archival and genealogical communities who will be given the opportunity to advise on such deals in the future
• integrate into its future cost-benefit calculations the long-term benefits of building Canadian capacity for population data development and innovation.


Lisa Dillon
Associate professor
Director, Programme de recherche en démographie historique (PRDH)
Département de démographie, Université de Montréal


References :

Statistics Canada’s Research Data Centres: http://www.statcan.gc.ca/rdc-cdr/

G. Darroch, R. Smith and M. Gaudreault, “CCRI Sample Designs and Sample Point Identification, Data Entry, and Reporting (SPIDER) Software,” Historical Methods, Vol. 40, No. 2 (Spring 2007), 65-75.

C. Gaffield, “Conceptualizing and Constructing the Canadian Century Research Infrastructure, Historical Methods, Vol. 40, No. 2 (Spring 2007), 54-64.

M. St.-Hilaire, B. Moldofsky, L. Richard and M. Beaudry, “Geocoding and Mapping Historical Census Data: The Geographical Component of the Canadian Century Research Infrastructure,” Historical Methods, Vol. 40, No. 2 (Spring 2007), 76-91.

To access Historical Methods: A Journal of Quantitative and Interdisciplinary History, see Taylor & Francis Online: http://www.tandfonline.com/toc/vhim20/current#.UcRwzdjNmbQ

See also Anthony di Mascio and Adam Green, “The Canadian Century Research Infrastructure: Enabling Humanities and Social Science Research in the Digital Age,” Digital Studies/le champ numérique, Vol. 1, No. 2 (2009): http://www.digitalstudies.org/ojs/index.php/digital_studies/article/view/169/212#_11.0_CCRI_Delivery

CCRI University of Alberta website: http://ccri.library.ualberta.ca/enindex.html

CCRI University of Ottawa website: http://www.canada.uottawa.ca/ccri/CCRI/Home.html

Population et histoire sociale de la ville de Québec, Université de Laval: http://www.phsvq.cieq.ulaval.ca/annexes/bases/R1851/R1851_resaisie.pdf

The 1852 and 1881 Census of Canada Project: http://www.prdh.umontreal.ca/census/

The 1881 Canadian Census of the FamilySearch website: https://familysearch.org/search/collection/1804541

Lisa Y. Dillon, “International Partners, Local Volunteers and Lots of Data: The 1881 Canadian Census Project,” History and Computing, Vol. 12, No. 2 (2000): 163-176, http://www.euppublishing.com/doi/abs/10.3366/hac.2000.12.2.163

Press release of the Minnesota Population Center’s collaboration with Ancestry:
http://www1.umn.edu/news/news-releases/2012/UR_CONTENT_380428.html

Ancestry: http://www.ancestryeurope.lu/about-ancestry/

FamilySearch: https://familysearch.

Jul 20
The 1921 Census of Canada, a (not so) hidden recent history…