Thursday, January 31, 2008

Crime Victim Data Question


A student here is working with some crime data and would like to know whether, for the UCR, for detailed offences (we're just looking at sexual assault), if it is possible to distinguish between male and female respondents. That is, we only want to know how many women were sexually assaulted, specifically in 1993 and 2004. It doesn't appear that it is possible to determine this, but we wanted to be sure. We are aware that there is the VAWS and the GSS Cycles where we can tell (and in fact, we're using this data as well).


You are correct. CCJS has confirmed that it won't be possible to use the Uniform Crime Reporting Survey (UCR) to determine how many victims of sexual assaults were women between 1993 and 2004. There is some data on victims of crime in the UCR2 Incident Based Survey (a version of the UCR that was introduced in 1988 and that operates simultaneously with the UCR Aggregate version of the survey) but there weren't enough police forces responding to the UCR2 between 1993 and 2004 for it to provide the data your student needs. The data it contains for this period is from a limited number of police forces.

(On a more general note, there is an interesting description of the UCRs design in the IMDB under

Friday, January 25, 2008

Cost of living comparisons between U.S. and Canadian cities


Can anyone point me towards a source that would provide information on the cost of living between specific cities in the two countries?


Statistics Canada produces "Canadian Foreign Post Indexes" (62-013-XIE). "These indexes are calculated to establish and adjust the Post Living Allowance (PLA) paid to Canadian government employees serving outside of Canada. They are comparative measurements that numerically express the difference between the retail prices of a representative basket of goods and services at a foreign location with prices for a similar basket of goods and services in Ottawa." They cover prices in a number of American cities (but always in comparison to Ottawa). There are a few different reports, each of which applies to a specific group of employees, so have a careful look at the notes and descriptions that accompany them to verify that they are appropriate for your user's needs if you consider using them.

The Economist Intelligence Unit
and Mercer Consulting
also have international cost of living surveys and related fee based products.

Wednesday, January 23, 2008

Correctional Services and Race


A student is looking for statistics around race and prisoners/prison population. I know the Adult Correctional Services Survey does tell us how many aboriginal inmates there are, but I'm wondering if that's as far as we can get with respect to race/ethnicity of the Canadian prison population. Am I missing anything?


Looking at the Canadian Centre for Justice Statistics, the only information available on race/ethnicity for the prison population is an Aboriginal identifier.

They suggested that you contact the Correctional Service of Canada (CSC) ( to see if they collect data on the race or ethnicity of individuals in their facilities. They are not sure however if they would collect data for provincial/territorial institutions.

You could have luck with CSC. I browsed the Publications section of the CSC website and one of their publications includes a figure that shows the % of the offender population (as of April 9, 2006) which is Caucasian, Hispanic, Aboriginal, Asian, Black and Other. The publication also includes a table that shows a more detailed description of the offender population as well as counts and % changes for 2001-2002 and 2005-2006. (Source: Corrections and Conditional Release Statistical Overview 2006 - -- see pages 51 and 52. The notes below the figure and table define "offenders".)

MIZ Areas


I have a researcher who wants to look at 2001 census data by Census Metropolitan Area and Census agglomeration Influenced zones (MIZ). I found the following table:

Cat #95F0495XCB2001012 - Cumulative Electronic Components of Profiles, 2001 Census

This table does break down the census data by the 4 MIZ categories (strong, moderate, weak, no MIZ) within a province. However, this researcher would like to find a table which shows the census data for the specific MIZ subdivisions for each corresponding CMA and CA. Is there such a table or would this require a custom tabulation?


The percentages of the employed labour force living in a particular CSD outside CMAs and CAs and working in the urban core of any CMA/CA are combined to determine the degree of influence that one or more CMAs/CAs have on that CSD. The data is not a one-to-one correlation between one CSD and one CMA. For background information, consult the Statistics Canada Web site ( to download a free copy of the Geography Working Paper entitled Census Metropolitan Area and Census Agglomeration Influenced Zones (MIZ): A Description of the Methodology, Catalogue No. 92F0138MIE.

They also confirmed that the data the client is looking for is not available in a standard product. If your client would like to describe further what he is trying to achieve, our contacts offered to continue to look into this in order to see if they could produce a custom product for him. They mentioned that this type of custom request would involve geocoding which could be very expensive (they didn't specify how expensive).

Tuesday, January 22, 2008

Labour Force Survey 2007 Question


Looking at the 2007 LFS files - what sps file corresponds to these? All the files I see go as far as 2005. Is this file still valid for the 2007 data files?


The author division has confirmed that the record layout and the micro-data file haven't changed since 2005 and that you can use the 2005 LFS SPSS files with the 2007 LFS data files

Monday, January 21, 2008

Cultural Diversity in Housing


I have a researcher who is examining cultural diversity in housing. Specifically he is looking dwelling type by place of birth for the CMA level. I am having quite a time finding anything, can someone put me on the right track?


Consider using the Census public use microdata to pursue your patron's request. The individual level file has the following dwelling variables as well as CMA and place of birth.

6 1 10 N HHCLASSP Household classification
7 2 11-12 N HTYPEP Household type
8 1 13 N UNITSP Household size
9 2 14-15 N ROOMP Number of rooms
10 1 16 N CONDWELP Condition of dwelling
11 6 17-22 N VALUEP Value of dwelling
12 1 23 N TENURP Tenure
13 1 24 N RCONDP Tenure -- condominium
14 4 25-28 N OMPP Owner's major payments (monthly)
15 4 29-32 N GROSRTP Monthly gross rent

Correctional Facilities Population


I understand from the Census Dictionary that residents of correctional facilities are enumerated using the institution's administrative records. Is it possible to find out exactly the nature of the information recorded? And would anyone know the method by which the admin record is translated into a census record?

Is the correctional facilities population included in aggregate records for the Census at any geographic level?


It is correct that correctional facilities are enumerated using the institution's administrative records. Census enumerators complete the 3A Census questionnaire for usual residents in correctional facilities using their administrative records. The 3A form collects basic demographic data: marital status, sex, date of birth(age), and first language.

Usual Residents of correctional facilities are considered as part of the institutional population. Any Census product (regardless of geographic level) where the universe includes institutional population will include residents of correctional facilities unless otherwise stated. For example, the 2006 Census population counts released on March 13, 2007 and the Age & Sex data released on July 17, 2007 include the institutional population and as a result include the residents of correctional facilities.

Wednesday, January 16, 2008

Annual data for Automobile/Auto Parts Exports and Imports


A student here is hoping to get data back to 1961 for Canada - United States trade in automobiles and auto parts (North American Industrial
Classification System codes 3361 through 3363). He would like annual data for these.

We can go back to 1980 with the World Trade Database - it uses the Standard International Trade Classification codes: the student is going to try to
identify the codes comparable to the NAICS.

Any ideas on earlier data? If someone had happened to digitize the back issues of Canadian international merchandise trade, (and they all have what the earliest electronic copy does) I think the student's life would be made much easier.


You may want to have a look at the trade data available through the UNs Comtrade database ( Comtrade has commodity based trade data for a number of reporting countries, including Canada, and some of their data goes back to 1962. The trade data won't be industry based however, it will be commodity based. Some helpful Comtrade user guides are available on the right side of Comtrade's home page ( and on the Basic Query Screen (

After checking with the International Trade Division, they stated that: "On a cost-recovery basis, we could produce automobile data by SEG (Summary Export Goods) (1966 to present) but it's not necessarily limited to manufacturing and it could become extremely costly. For your information, here are the SEG codes that cover the automobile industry:
51110 Passenger automobiles and chassis
51120 Trucks, truck tractors and chassis
51131 Other motor vehicles
51132 Motor vehicle engines and parts
51139 Motor vehicle parts, except engines".

They also told me that we started using NAICS in 1992 and that we had SIC (Standard Industrial Classification) prior to that. They mentioned that Industry Canada's Trade Data Online ( would give you NAICS based data back to 1992 but that it would be difficult for the student to find NAICS data going back further than that.

Rounding Census 2006 Numbers


If I recall properly, the census numbers used to be perturbed during rounding - is this the case with the 2006 Census?


Below is the description of how random rounding is used and how it applies to data for the 2006 Census. The information was provided by our Census Consultants and comes from pages 6 and 7 of "Data Quality and Confidentiality Standards and Guidelines" for the 2006 Census

From pages 6-7

2.2. Random rounding
All counts in census tabulations are subjected to a process called random rounding. Random rounding transforms all raw counts to random rounded counts. This reduces the possibility of identifying individuals within the tabulations. For 2A (100%) data, all counts are rounded to a base
of 5. This means that all 2A counts will end in either 0 or 5. The random rounding algorithm employed controls the results and rounds the unit value
of the count according to a pre-determined frequency.

2B (20%) data require a slightly different random rounding algorithm. All counts greater than 10 are rounded to base 5, as is done for 2A data. Counts less than 10 are rounded to base 10. This means that any 2B counts less than 10 will always be changed to 0 or 10. The table below shows the effect of rounding on 2B counts with a value less than 10.

The random rounding algorithm uses a random seed value to initiate the rounding pattern for tables. In these routines, the method used to seed the
pattern can result in the same count in the same table being rounded up in one execution and rounded down in the next.

FED Land Area & Population


A researcher at McMaster is looking for population density data of Canadian Federal Electoral Districts from 1972 to present. So the researcher needs land areas and populations for each FED. Does anyone know where this data might be available?


The Geography Division recommends that your client use the 1991 - 2006 GeoSuite products to get the population data for FEDs. Beyond 1991, the Geography Division suggests that you use the Geographic Attribute File (GAF) files, which go back to 1971, to get the population data. The GAF, their user guides and record layouts are available in the Geography folder on the DLI FTP site. (Please note that there is no user guide for the 1996 GAF, only a record layout.)

Geosuite 2006 and 2001 will also provide you with land area for FEDs. GeoSuite1996 and GeoSuite1991 do not have land area for FEDs however, nor do the GAF files. You may need to refer to print publications or other sources to find FED land areas for 1996 and previous years.

Monday, January 14, 2008

Publication of Beyond 20/20 Tables in Textbook


I know part of the answer to this question, but I was hoping for a bit more clarification. A professor here is writing a textbook and wants to generate tables in Beyond 20/20 (I think using homicide/victim/crime data) and wants to know if she can use them in her book. Now, I'm pretty sure that if she is getting the B2020 tables from DLI, the answer is NO and that she would need to seek permission directly. However, what if she is getting them from E-Stat? I suspect the answer is still no because E-Stat is for educational use? AND, what if she generates them on the Stat-Can website and then pays for the time series that way? Any help/clarification you can provide is much appreciated.


Publishing Statistics Canada products in textbooks falls outside the DLI license. The Regional Office should help your professor get the approval she needs.

The DLI Licensing database on the DLI website says, "Faculty are referred to the Marketing Division of Statistics Canada for discussions on textbook licensing."