Prepared by Eric Fong, Department of Sociology, University of Toronto

General Comments

• Participants strongly express the needs for a hierarchical file structure for PUMS to link family members.
• Information to link family members can provide information on many areas, such as immigrant children, economics of family, ethnic identity, and social mobility.
• Categories of the industry variable should be elaborated. Existing PUMS only includes a few industries. The information is too crude for more detailed analysis.
• 2001 PUMS should include filter questions from the post-censal survey. The information may be useful to researchers.
• Some participants suggested that information about the class of immigration and last country of residence is needed.

Section on Language

• Participants strongly suggest inclusion of questions to capture levels of language proficiency.
• Questions should measure various dimensions of language ability. These dimensions should include the levels of ability to write, to speak, and to read English or French. These abilities are not necessarily highly correlated. For example, many immigrant children speak their ethnic language very well, but cannot read it.
• Questions on language at work should be included. This is an important indicator to capture participation in ethnic economy.
• Information on the language used in the interview should be included in the 2001 PUMS.

Section on Socio-Cultural Information

• The question about the year of the immigrant’s first arrival is needed. However, the wording should differentiate between visiting and settling.
• Question.13
• To save some space, the answer should be open-ended instead of multiple choice.
• All participants agree that country of birth is important information which should be kept in the 2001 census.
• Question16
• The word "first" should be deleted. Immigrants can become landed immigrants only once.
• Question 17
• Some participants suggested that the question should be kept because it provides information about changes in the ethnic composition of Canada.
• The term, "ancestors belong" should be replaced by "principally identified." The suggested wording better reflects the situation of a multi-ethnic population.
• There is some concern that the word "ethnic" should be dropped and replaced by "cultural" origin. Ethnicity is essentially a cultural phenomenon. Many Canadians, especially those who are Canadian-born, consider themselves to be Canadians associated with a certain cultural heritage.
• Question 19
• The first phrase should be changed to "Does this person principally identify." The new phrase discourages multiple responses.
• Respondents who check "Latin American," may be white, black, and Asian. This category cannot differentiate different racial groups.

Comments on Religion

• All participants feel strongly that the question should be kept. The information has important policy implications (e.g. school boards need the information to determine the number of Catholic schools in their districts).

Comments on Education

• All participants strongly express the need for information on the country where the highest education was received. This information substantially affects the earning potential of immigrants.

Comments on Post-Censal Survey

• A post-censal survey on immigrants is necessary.
• Some suggested that the survey should oversample (1) self-employed people; (2) immigrants in Toronto, Montreal, Vancouver, and Calgary.
• Target specific groups: Be sure there are enough cases to analyze certain subpopulations (e.g., HALS should oversample immigrant children).
• Samples of subpopulation can be drawn from the "filter" question.