3.11 Labour section

3.11.1 Journey to work

During the 2011 Census consultation, there was strong support for enhancing the commuting data available from the census.  Key enhancements were to:

• modify the mode of transportation question to add detail on the type of public transit used
• allow multiple responses so as to better understand the methods people use to access public transit facilities, such as park-and-ride and subway stations
• for those using 'car, truck or van,' collect information on the number of people sharing the ride to work to support initiatives such as high occupancy vehicle lanes
• collect information about commuting times (i.e., the times when people leave home and how long their commute is in minutes).

Mode of transportation

Two versions of the mode of transportation question were tested.  For both versions, the 2006 response category 'Public transit' was replaced with four more specific transit categories.

The 2B.1 control questionnaire used the same wording as the 2006 Census question and asked for a single response.

For the alternate question, the instruction was replaced with 'If this person used more than one method of travel to work, mark all methods.'  Skip instructions were added for a new follow-up question on carpooling.

Figure 3.32 Mode of transportation to work, single response question

Source: Statistics Canada, 2008 Census Test 2B.1 control questionnaire.

Figure 3.33 Mode of transportation to work, multiple response question

Source: Statistics Canada, 2008 Census Test, 2B.2 alternate questionnaire.

The alternate questionnaire allowed multiple responses while the control questionnaire instructed the respondents to mark one mode of transportation that was used for most of the travel distance.  For respondents who gave a single response in the census test, the mode reported was similar to the 2006 Census results.

Reporting of the transit modes was assessed geographically.  It was found that those reporting subway, light rail, and commuter rail mostly lived in census metropolitan areas (CMAs) where the service exists.  When that was not the case, the 'place of work' responses were checked and the respondent worked in a CMA with the service.

The 2008 Census Test results showed commuter rail being reported in Calgary, however Calgary does not have commuter rail service.  Confusion due to local branding of transit services, such as CTrain for their light rail line may have led to confusion with the commuter train category.  This situation has also been noticed in U.S. census data.  In qualitative testing, Vancouver residents had trouble deciding where to report SkyTrain among the choices of subway, light rail, elevated and commuter rail.  To address the confusion, it was recommended to add specific examples to the question for the 2009 Census Test.

To assess the proportion of multiple responses, comparisons were made with the 2006 Households and the Environment Survey (HES) and the Australian 2006 Census, as these survey questions invited multiple response.  The multiple response rate observed in the test was 1.5 to 3 times higher than in the compared surveys.

In the census test, more than a quarter of the most common multiple response combinations observed were for modes that were unlikely to be used in the same trip such as driver and passenger.  Qualitative tests for the 2008 Census Test showed that some respondents are reporting different modes that they used on different days.

Another issue is confusion due to conflicting 'go to' instructions for respondents who report 'car, truck or van' and one of the modes from the second block of responses.

For the 2009 test, the recommendation was to keep the additional transit categories but to revert to the single response instruction that was used in the control version of the question.  The revised question is shown below.

Figure 3.34 Mode of transportation to work

Source: Statistics Canada, 2009 Census Test, 2B questionnaire.

The 2009 test included a large sample in the Montréal area including many in the suburbs.  These suburbs, which are serviced by commuter train, also show some reporting of 'Subway, light rail or elevated rail.'  This could be misreporting between 'light rail' and 'commuter train.'

For Calgary and Edmonton combined, transit was primarily reported as 'Bus or streetcar.'  Both these cities have light rail systems, so it appears that there is confusion between the terms streetcar and light rail.

Changes to the first three categories relating to public transit were proposed for the long-form census questionnaire, now the 2011 National Household Survey:

• Bus
• Subway or elevated rail
• Light rail, streetcar or commuter train

More comprehensive examples for the two latter categories were also proposed:

• Vancouver SkyTrain
• Toronto Subway/RT
• Montréal Metro

Mark 'Light rail, streetcar or commuter train' for:

• Vancouver West Coast Express
• Calgary CTrain
• Edmonton LRT
• Toronto streetcars
• Toronto GO Train
• Ottawa O-Train
• Montréal commuter trains.

The word 'usually' in Question 47(a) appears in bold font.  The use of bold font to emphasize this important word in all the journey to work questions was proposed for the long-form census questionnaire, now the 2011 National Household Survey.

Top of Page

3.11.2 Vehicle occupancy

Figure 3.35 Number of people sharing ride to work

Source: Statistics Canada, 2008 Census Test, 2B.2 questionnaire.

The vehicle occupancy results were compared with the 2006 Households and the Environment Survey (HES) and the 2000 U.S. census results.  The census test results were very similar to the HES results for the proportion reporting 'drove alone.'  The 2000 U.S. census asked a very similar question but with more response categories.  Compared to the U.S., there is less car use and fewer people driving alone but more public transit use in Canada.  Thus to make the data comparable, the results were only compared for carpoolers.  Even though the results are slightly different, both the U.S. census and the 2008 Census Test show that the majority, 77.4% (U.S. census) and 82% (2008 Census Test) report two persons in a vehicle as carpoolers.  In both countries, the proportion reporting each response category decreases as the number of people increases.  Thus, it was recommended to use this new question for the 2009 Census Test but to condense the last four categories as '3 or more people' due to limited space on the questionnaire.

This additional question was proposed for the long-form census questionnaire, now the 2011 National Household Survey.

Top of Page

3.11.3 Time leaving for work

The 'time leaving for work' question, included in the 2B.2 alternate questionnaire, was one of the new commuting questions to enhance the commuting data in response to suggestions during the 2011 Census consultation.

Figure 3.36 Time usually leaving for work

Source: Statistics Canada, 2008 Census Test, 2B.2 questionnaire.

The results for Question 49(a) were compared with three other sources: the 2005 General Social Survey (GSS) on Time Use (Cycle 19), the 2000 U.S. census, and the 2006 Transportation Tomorrow Survey (extended Greater Toronto Area).  All the three surveys used different methodologies.  The GSS respondents were asked to report on their activities for a 24-hour period using a diary format.  The 2006 Transportation Tomorrow Survey was conducted by telephone interview and asked respondents to report on all trips made during a 24-hour period on a specific weekday.  The 2000 U.S. census question was similar to the tested question.

The results for this new question were similar to the results obtained by the GSS and U.S. census.  Most of the minute responses were on the hour, half hour or the quarter hours, and almost all the other responses were a multiple of 5 minutes.  Similar patterns were observed in the other surveys despite the different methodologies.

The removal of the title above the boxes for entering 'time' and the retention of this question were proposed for the long-form census questionnaire, now the 2011 National Household Survey.

Top of Page

3.11.4 Travel time to work

Question 49(b), which asked about the commuting time to work, was also new to the 2B.2 alternate questionnaire.

Figure 3.37 Usual travel time from home to work

Source: Statistics Canada, 2008 Census Test, 2B.2 questionnaire.

The 2008 Census Test results were compared with the 2006 HES and GSS on Time Use, and the 2000 U.S. census.

The 2006 HES asked respondents if their commute was under 30 minutes or over.  Analysis of the 2008 test showed fewer respondents with a short commute and more with a long commute compared to HES.  In general, the response distributions of the census test results were similar to those from the GSS and the U.S. census.

It was proposed that this question be retained for the long-form census questionnaire, now the 2011 National Household Survey.

Top of Page

3.11.5 Child care and support payments

Figure 3.38 Payments for child care in order to work

Source: Statistics Canada, 2008 Census Test, 2B.2 questionnaire.

Figure 3.39 Support payments paid to former spouse or partner

Source: Statistics Canada, 2008 Census Test, 2B.2 questionnaire.

A question about child care costs paid in the previous year and a question about support payments made in the previous year for children or prior spouses were added to the 2008 Census Test, 2B.2 questionnaire.  The intent is to employ these and other measures of spending or expenses to derive a concept of disposable income.  This disposable income may be used to implement measurement on the census of the market-basket measure of poverty designed by Human Resources and Skills Development Canada (HRSDC).

The qualitative testing showed that these new questions were not unduly sensitive.  Some concerns were raised as to the clarity of the child care expenses question and more qualitative testing was required.  After these were carried out, it was suggested to replace 'enter the amount ' by 'enter the total amount for 2010.'

Compared to the data from the Survey of Labour and Income Dynamics (SLID), the response patterns to both questions were considered quantitatively similar.  There was no measurable impact on acceptance to the tax permission question that followed on the questionnaire.  It was recommended that the two new questions be retained for the 2009 Census Test and for the proposed long-form census questionnaire, now the 2011 National Household Survey.