Linkage of the Intergenerational Income Database (1978 to 2018) to the 1996, 2001, 2006 and 2016 census microdata files and the 2011 National Household Survey microdata files

The Intergenerational Income Database (IID) links children and their parents using administrative tax data, allowing researchers and analysts to study intergenerational phenomena such as the transmission of earnings from parents to children and income mobility. Thanks to the longitudinal structure of the database, children can be followed into their mature adulthood.

The IID has been used extensively in the past to analyze intergenerational mobility in Canada. It contains detailed information about various income sources, including earnings, self-employment income, dividends, investment income and government transfers. However, the IID lacks critical information about individuals’ human capital, skills, occupation and work experience.

Hence, the primary goal of the IID–census linkage is to add the rich socioeconomic information from the census, such as the place of birth of respondents and their parents, parental education, and occupation, to the IID files.

Intergenerational Income Database

Data in the IID are organized in two main modules: the family files and the individual files. At the core of the IID are the family files. The information in the family files was retrieved from Statistics Canada’s T1 Family File (T1FF). The T1FF uses the census family as its family concept. Children in the IID are grouped into cohorts according to their birth year. The database consists of six cohorts of children linked to their parents in 1982, 1984 and 1986 (Panel A) and in 1991, 1996 and 2001 (Panel B). These years are known as cohort years. To be included in the IID, the children must be aged 16 to 19 in one of the cohort years, have a valid social insurance number and be residing at home.

The earnings and income information of family members is taken from T1 annual tax returns filed by individuals in Canada. T1 files from 1978 to 2016 were used to create the annual IID T1 files. These files are organized in two panels. Panel A includes T1 files for all members of the original IID cohorts (1982, 1984 and 1986) and their parents. These files cover the period from 1978 to 2016. Panel B includes T1 files for the 1991, 1996 and 2001 cohorts and their parents. These files cover the period from 1981 to 2016. Income and earnings are expressed in nominal dollars. One of the key features of the IID is that it allows for income mobility analysis at the family level.

Since its introduction in the mid-1990s, the IID has been the main source of analytical research on intergenerational income mobility in Canada. The database has been updated several times, most recently in 2016, when three new cohorts (1991, 1996 and 2001) were added to the IID sample. The intention is to add new cohorts in the future to maintain the five-year interval between cohorts.

Census microdata

The Census of Population provides detailed information on the demographic, social and economic characteristics of people living in Canada on a specific date, information about their housing units, and a wide range of other individual and household information. The census uses two primary data collection modes: the Internet and a paper questionnaire. The census microdata files contain information from the long-form questionnaire that was distributed to 25% of all Canadian households in 2016, the National Household Survey distributed to 30% of all households in 2011, and the long-form questionnaire distributed to 20% of all households in 2006 and in prior census years.

Detailed descriptions of census microdata files are available from the Statistics Canada website.

Linkage keys

An essential element of the IID–census linkage is the linkage keys. Because administrative and census data have different unique personal identifiers, the linkage key files, which contain both sets of identifiers, are a necessary component of the IID–census linkage environment.

There is a separate linkage key file for each census year. Currently, available years are 2016, 2011, 2006, 2001 and 1996. Each linkage key file contains unique personal IID and census identifiers. The linkage keys are available to researchers in SAS and Stata formats.

Access

Researchers wishing to use the linked data for analysis should submit a proposal following the standard research data centre (RDC) process. It is recommended that researchers consult the IID User Guide and the relevant Census Dictionary before submitting proposals.

For more information, please contact the local RDC analyst.

Restrictions

While researchers are allowed to conduct analysis in parallel with different single census years linked with the IID, they may not conduct analysis with a subset of a sample that brings together census records of the same respondents from at least two different censuses, as per the current set of vetting and release requirements.