Tuesday, August 30, 2005

CRA and GST/HST Statistics Compendium


I have a user who is looking for the amount of GST revenue earned/generated by NAICS or SIC industry classification, in 2001, 1996, and 2001. I have checked the STC web site, CANSIM (E-stat), Financial performance indicators, and Industrial monitor.

I'm now stumped - what am I missing?


Try the Compendium of GST/HST Statistics on the CRA website.


They include tables on GST/HST collected, which they define as "This amount is based on all GST/HST charged, at 7% GST or 15% HST, on paid and unpaid invoices during the reporting period. The total of GST collected in a reporting period is based on
all GST/HST collected and collectable reported on line 105 of Form GST34, and includes adjustments from all forms filed by
GST/HST registrants."

They use SIC to classify industries.

Table 7g is the distribution of GS/HST collected by major industry group.

The time series runs from the 1994-95 fiscal year to 2003-04.

Tuesday, August 23, 2005

Canadian mortality microdata


Yet another researcher is looking for microdata from Canadian registration of deaths records, or whatever they are called now. All years from 1984 to 2004 would be appreciated.

The variables needed (and that were available at least in the 1980s) are: year and month of death; province, county, locality of residence; marital status; sex; age; year of birth; place of birth; province, county, locality of death; ethnic origin; occupation; industry; place of death (hospital, sanatorium, etc.); autopsy; cause of death, nature of injury (International classification of disease, ICD-9); birthplace of mother, father.

I am aware that not all these variables are consistently filled in by all provincial vital statistics agencies, but for those that are, the info would be great.

The alternative, I suppose, is to use again US mortality records, which are fairly readily available.


Statistics Canada does not disseminate vital statistics information as a PUMF. Aggregated statistics at the Canadian level is all that is available for detailed tables.

Vital statistics are administrative data (meaning: they are collected by another government department and then remitted to Statistics Canada for publishing). The owners of the data (provincial ministries of vital statistics) still own the "rights" to the data and do not allow any release of the information by Statistics Canada to other users.

STC's Health Division advised that every single provincial ministry would need to be contacted in order to release the data. The rumour is that requests have been pending for over two years and the outcome is not looking favourable.

Most unfortunately, your user may need to do as you describe - use American data.

Thursday, August 11, 2005

Pension Plans


A student working on a research projet here at Concordia is looking for the following data:

Number of pension plans (or members of those plans) by industry and by type of plans (Defined benefit and Defined contribution pension plans) at the national level. This data seems to be available from the "Pension plans in Canada" database but I could find no currently available product that would answer the question. I know custom tables could be produced at a cost, but would the information be available in any other form through DLI or DSP?

The other aspect of the research has to do with level of unionization by industry also at the national level. Those numbers are available from CANSIM (Labour force survey estimates (LFS), employees by union coverage, table 282-0077). Now, would there be any survey that would cross-index this variable (unionization by industry) with the other one (type of pension-plan per industry)? In other words, would it be possible to associate, say the "defined benefit segment" of industry X with its level of unionization?


I researched what I could for you, but came up with very similar results:

1) Pension Plans in Canada : Key Tables (74-508-XWE) has some great tables, but nothing that crosses industry with union participation. It does cross public vs. private firms, but that is not detailed enough for your patron. I think that one of the only options to you then is to explore a custom extraction for a fee.

2) The Labour Force Survey does not ask pension questions, so it would never be crossed with union participation.