A New Approach for the Development of a Public Use Microdata File for Canada's 2011 National Household Survey
2. Background

IPUMS-International disseminates integrated, confidentialized census microdata samples to researchers world-wide. As of 2013, 238 samples representing 74 countries were made available to more than 7,000 registered users, including several Canadian researchers (McCaa, et al., 2013). In response to requests from that organization, Statistics Canada has agreed to produce a detailed hierarchical PUMF from the 2011 NHS. Unlike current PUMFs produced by the Agency, the protection of data confidentiality would be achieved through data perturbation. The special PUMF would be provided to IPUMS-International to add to their collection after some manipulation on their part. IPUMS-International increases international comparability by adding metadata and content. For example, it creates standardized versions of variables and adds family relationships (Sobek and Kennedy, 2009).

Following discussions with IPUMS-International, Statistics Canada agreed that the special PUMF would include a set of 48 variables for a sample of about 2.7% of the population. Some variables, such as age, occupation and place of birth, would be very detailed, possibly at a cost of higher levels of perturbation.