Use of administrative data for the Census of Population: overview of a research project and data quality assessment tools developed by Statistics Canada

Andre Lebel, Statistics Canada

Canada has held censuses every ten-years since 1871, and every five-years since 1956. In industrialized countries, the traditional form of census is being questioned, and some countries, mostly from Northern Europe, have just recently moved away from this model by taking advantage of their long existing population register. The term traditional census refers to the collection of information on individuals through the use of a variety of collection methods (full field enumeration, self-completion paper questionnaire, telephone or internet). Even for countries still holding a traditional census, many have recently been reviewing the potential of using administrative and other alternative data sources to replace the direct enumeration of the population (e.g. United Kingdom, Australia and New Zealand). Statistics Canada has also put in place a research project to assess the extent to which it is possible to complement or supplement the Canadian Census with a “virtual population register” using a variety of administrative databases (tax and economic activity, foreign entries, vital statistics, etc.) already available to the statistical agency. A key element of this project is referred to as the Canadian Statistical Demographic Database (CSDD) which contains basic socio-demographic variables (age, sex, geographic location) which replicate parts of the census universe. The first objective of this paper is to present the CSDD and briefly explain data sources and methods used to build it. In the second part, data quality indicators developed to assess its fitness and comparability with the current Census will be presented. The third part will contain an assessment of the quality of the coverage achieved by the CSDD for different levels of geography (Provinces, Census Metropolitan Areas (CMA) and Census Divisions (CD) and also, by age groups and sex. Finally, this paper will conclude by outlining CSDD’s strengths and weaknesses and propose ways to improve it in 2016.

  See extended abstract

Presented in Poster Session 1