Thursday, December 12, 2019

Labour Force and Commuting for Work Question


I’m helping a researcher sort out some travel to work data for the CSD of Houston, BC. From this table: looking at Houston (DM) we get a labour force of 1065 living in Houston, but the census profile of Houston (DM) gives a labour force size of 1615, and it seems to both of us the difference between the two seems much too large to be due to suppression and random rounding. I’m wondering if there is documentation I’m just missing that would account for this difference?


On the first table, your client has given the “universe” or population captured, is people who are employed in the labour force aged 15 years and over who have a usual place of residence. Whereas, in the second table your client is looking at a universe that we would title persons in the labour force aged 15 years and over. We can find this information from looking at either the table title or the profile section title.

Here is a little explanation on why this population is so different from the total labour force that you see on the second table.

Seeing as this table is a commuting flow chart, respondents must have a usual place of work which limits the number of respondents that you see. This helps with the accuracy of the table as those with a flexible work environment are not always commuting to the same place. This table is also only showing those who are employed in the labour force. From reading this definition of both labour force and labour force status we see that those counted in the labour force may have been unemployed during the reference period which would also mean that they would not have an eligible destination to use in the commuting flow chart.

After seeing these differences I would recommend that your client look at journey to work sections of the profile for comparability instead of the labour force sections. You can find this by selecting “Journey to work” using the “Topic” toggle bar at the top of the profile.

One last consideration with this flow chart is that suppression rules do apply to protect the privacy of Canadians. Random rounding has of course been used on all of these tables however as well, there may be some respondents living in Houston ,BC who have a very unique commuting destination. The six possible commuting destinations listed on the first table may therefore not be exhaustive, which would account for the sum of these 6 answers not adding up to exactly the same total as the journey to work universe.

Wednesday, December 11, 2019

Canadian Active Living Environments - Geographic Variable in CCHS 2015-16


I am helping a researcher with the CCHS 2015-16 dataset and we were wondering about the “Canadian Active Living Environments” geographic variable. I am struggling to find metadata for this variable in the accompanying documentation. The Data Dictionary points me toward the “Derived Variable (DV) Specifications”, but there is nothing in that document listed for “Canadian Active Living Environments”.

Can someone direct me towards a webpage or pdf file with information about how this variable is derived?


Does this StatCan article help?

or the Canadian Active Living Environments Database (Can-ALE) User Manual & Technical Document:

Monday, December 9, 2019

Prostitution Offences


I was helping a researcher who's looking for "the breakdown of the prostitution offences by each offence (e.i. incidents for s.213, s.286.1 etc)".  

From this table: 
35-10-0177-01 - Incident-based crime statistics, by detailed violations, Canada, provinces, territories and Census Metropolitan Areas

We can get the breakdown of the prostitution offences (classified under Total Commodification of sexual activity violations, and Total Prostitution) by types of offences. 

Next, in order to map those types of offences to criminal codes (such as s.213, s.286.1 etc), we consulted the questionnaire and the reporting guide linked from this page:

Basically, we figured out that on the questionnaire(PDF), there is the code for "Prostitution Total" (046) as well as the more specific codes below it; and from the reporting guide (p. 86), it shows how that code is linked to criminal code (section 213 etc).  Although the codes in the stats don't align with the codes on the questionnaire, for example, bawdy house 3110 vs. 047, we can just match them up by the same description (i.e. bawdy house). 

However, there are still some critical problems -- the questions in the questionnaire don't entirely reflect the crime hierarchy of stats, for example, there's no "Total Commodification of sexual activity violations" and its breakdown on the questionnaire, and we suspect it's because the questionnaire/reporting guide on the website needs to be updated. Furthermore, since the researcher is interested in the impact of the 2014 legislation, it would be useful to get access to the questionnaires/reporting guide of different times, which are unavailable from the website.  

Sorry for this rambling email. I'd appreciate it if you can shed some light on this. 


I’ve received the following response from subject matter:

“If your researcher hasn’t seen the tables of concordance contained in the 2019 UCR manual (attached), it could be very useful. It is a comprehensive lookup table that cross references all Criminal Code sections to their respective UCR2 violation codes.

The 3-digit UCR codes mentioned, such as ‘Prostitution Total (046)’, come for the old UCR1 Aggregate survey, which is no longer in use and has not been for a long time.

All of our current online CODR (new CANSIM) tables are based on the 4-digit UCR2 codes.”


"... [data tables] are readily available via the website. These are customizable and downloadable. Below is the CAN/PROV/CMA table, there are also similar police service level tables that can be found on the site as well.

Incident-based crime statistics, by detailed violations, Canada, provinces, territories and Census Metropolitan Areas


"[The manual] is not  available on the website, we distribute it to our partners, but also to anyone who requests it”

Friday, December 6, 2019

Cannabis Cost Data



I have some researchers looking for cost data related to cannabis at the provincial level. Specifically they are looking for data on consumer spending on legal vs. illegal cannabis (I already sent them to StatsCannabis, but they are hoping for market size data and not just cost/gram), as well as wholesale costs of legal retail cannabis shops.

They’ve already been to our RDC and there doesn’t seem to be anything to suit their needs there.


I’ve received the following response from subject matter:

“Thanks for your inquiry, we just released provincial level detail in the core Provincial and Territorial Economic Accounts. We do estimate Household Final Consumption Expenditure by province for legal medical, legal recreational and illegal recreational cannabis consumption. This data is available by province to 2018. Nationally, the same data is available up to 2019Q3 and for more timely data, data on legal recreational only is available from the Monthly Retail Trade Survey (see the three links attached). 

As a note, the Monthly Retail Trade estimates only include legal cannabis sales made by retail stores, it would not include any retail sales made by provincial wholesalers or own use consumption, which are included in the household final consumption expenditure estimate.

Wholesale costs for legal cannabis retail shops are not currently available.!recreate.action?pid=3610022501&selectedNodeIds=2D1,3D128,3D129,3D130&checkedLevels=0D1,0D2&refPeriods=20140101,20180101&dimensionLayouts=layout2,layout2,layout3,layout2&vectorDisplay=false (Provincial Household Final Consumption Expenditure)!recreate.action?pid=3610012401&selectedNodeIds=2D1,3D1,4D130,4D131,4D132&checkedLevels=0D1&refPeriods=20180701,20190701&dimensionLayouts=layout2,layout2,layout2,layout3,layout2&vectorDisplay=false (National Household Final Consumption Expenditure)!recreate.action?pid=2010000801&selectedNodeIds=2D30,3D1&checkedLevels=0D1,0D2&refPeriods=20190501,20190901&dimensionLayouts=layout3,layout2,layout2,layout2&vectorDisplay=false (Monthly Retail Trade Survey)”