Retirement Savings Data

Detailed information for taxation year 2004

The retirement savings data file provides information on the number of Canadians participating in registered pension plans (RPPs) and contributing to registered retirement savings plans (RRSPs) for the taxation year.

Data release - February 7, 2006

• Description
• Data sources and methodology
• Data accuracy

Description

The retirement savings data file was created in order to provide socio-economic data on the retirement savings of Canadians through registered retirement savings plans (RRSPs), registered pension plans (RPPs) and deferred profit sharing plans (DPSPs). The data provide information on the characteristics of both savers and non-savers, on the contributions to RPPs and RRSPs, and on the extent to which RRSP room is being used by whom.

The data are derived from the information provided annually to the Canada Revenue Agency (CRA) in individual tax returns.

The primary users of this file are the federal and provincial governments. Other main users include private consulting firms, financial institutions and insurance companies.

Reference period: Taxation year, the demographic variables such as the age and sex are at December 31.

Subjects

• Income, pensions and wealth
• Income, pensions, spending and wealth
• Older adults and population aging (formerly Seniors)
• Pension plans and funds and other retirement income programs

Data sources and methodology

Target population

The target population covers individuals who completed a T1 tax return for the reference year by the date the file was sent to Statistics Canada. The universe includes taxfilers who live in the provinces and territories and excludes non-residents, persons who were not alive as of December 31 of the reference year and those who were born after the reference year.

Instrument design

This methodology does not apply.

Sampling

This is a sample survey with a cross-sectional design.

The data were produced from samples that were stratified by province and territory. The sample size for each province was chosen to minimize the coefficient of variation of the estimate of total income.

Data sources

Data are extracted from administrative files.

Data were extracted from T1 tax returns provided to Statistics Canada by the Canada Revenue Agency.

Error detection

Most of the processing of the data was done by the Canada Revenue Agency, when assessing individual tax returns. However, Statistics Canada made some additional adjustments, to ensure data consistency (see imputation).

Imputation

The adjustments included the following:
- dollar values of one ($1) were converted to zero;
- in a few cases, variables that exceeded a prescribed maximum were adjusted;
- some negative values were changed to nil;
- the relationship between certain dollar fields was tested and adjusted if necessary;
- sex, if unknown, was randomly imputed.

Data for the following variables were derived using other information on the file: RRSP normal contributions, maximum room for the year, unused RRSP room and new RRSP room.

Estimation

As part of the estimation process, the retirement savings data are estimated by multiplying information provided annually to the Canada Revenue Agency (CRA) in individual tax returns (T1) by its sampling weight. The sample size is about 5% of all taxfilers and the sample is stratified by province. Estimates provide information on the characteristics of both savers and non-savers, on the contributions to RPPs (Registered Pension Plans) and RRSPs (Registered Retirement Savings Plans), and on the extent to which RRSP room is being used and by whom.

The dollars amounts were converted to 2004 constant dollars by using the Consumer Price Index.

Quality evaluation

The Small Area and Administrative Data Division (SAADD) at Statistics Canada produces annual information on RRSP contributions from the personal tax returns (record number 4106). This information was used to compare both levels and trends. The quality of the estimates was also evaluated using trend analysis.

Disclosure control

Statistics Canada is prohibited by law from releasing any data which would divulge information obtained under the Statistics Act that relates to any identifiable person, business or organization without the prior knowledge or the consent in writing of that person, business or organization. Various confidentiality rules are applied to all data that are released or published to prevent the publication or disclosure of any information deemed confidential. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.

The data are in conformity with the provisions of the Statistics Act and are available only at the aggregate level. Also, data are provided in thousands, for the number of taxfilers, and in millions, for dollars amounts, in order to ensure the confidentiality of the information about individual taxfilers.

Revisions and seasonal adjustment

A conversion with the Consumer Price Index (CPI) is applied to current dollars amounts to derive the amount in constant dollars.

Data accuracy

To obtain an indication of the quality of the data, the proportion of estimates falling into one of the following ranges of coefficients of variation (CV) was determined, see attached table in additional documentation.

The majority of estimates for which an estimate could be produced (i.e., where the domain sample size was greater than zero) fell into one of either the Excellent, Very good, Good, or Acceptable categories. For the tables which were computed for all of Canada using more detailed sex, age, and income breakdowns (Tables 1A to 9A), 85.3% of the estimates are at least "Acceptable", with 41% in the "Excellent" category. For the tables which were computed for each province using less detailed sex, age, and income breakdowns, (Tables 1B to 9B), 93.5% of the estimates are at least "Acceptable", with 48.3% in the "Excellent" category.

Tables that contain CVs are available upon request. In general, the greater the sample size within a domain, the more precise the estimate will be. For the estimates where there were a relatively large number of domains for which the coefficient of variation fell into the "Use with caution" category, the domains tended to be a rare characteristic in the population.

• Additional documentation