Communiqué

May 2012

Summary

  • Changes to Canada's domestic travel survey, Travel Survey of Residents of Canada (TSRC), between 2011 and previous years are of sufficient magnitude that they will likely result in a break in the historical series.
  • Based on preliminary findings, survey partners and other users can anticipate changes in the volume/value estimates for 2011 relative to previous years that are beyond the ones expected because of economic or demographic changes.
  • The direction and size of changes in volume/value estimates for 2011 is still unknown and may not be the same across all regions or levels of geography.

Overview of Changes

  • Like the 2010 TSRC, the 2011 study is based on a single rotation of the Labour Force Survey.Footnote 1 Unlike the 2010 version, the new design collects information on overnight trips taken by adult Canadian residents over a two-month rather than one-month period.Footnote 2
  • Between 2010 and 2011, the number of trips for which a respondent was asked to report spending, lodging and activities in a detailed manner changed. Other changes included main purpose categories and the boundaries of in-scope versus out-of-scope trips.Footnote 3
  • For a full description of changes between TSRC 2010 and TSRC 2011, readers are advised to consult Differences Between the 2011 Redesigned TSRC and the 2010 TSRC, available on Statistics Canada's website.Footnote 4

More Specifically

  • In the 2011 survey, limited information is collected for all in-scope reported trips via a trip roster. Depending on the number of trips reported, full details are collected for up to three trips. Details include trip spending, nights spent in lodging types at specific locations, and activities. Statistics Canada's selection system to identify trips to be reported in detail favours out-of-province, more recent and overnight trips.
  • Statistics Canada developed a complex imputation procedure to assign characteristics of fully reported trips to those for which only limited information was captured in the trip roster. Approximately 13% of overnight trips and 25% of same-day trips that were rostered but not explored in detail will have characteristics of other people's trips assigned to them via this imputation procedure. This process could result in some anomalous findings, particularly with respect to activities, lodging types and locations of overnight stays for trips with more than one overnight location.
  • Because the methodologies are different, pooling of the 2010 and 2011 TSRC files is not feasible. Thus, the number of respondent records available for geographic or sector analysis will be smaller than the numbers available in pooled data for 2009 and/or 2010 reference years.Footnote 5

    As a consequence of the two-month recall period for overnight trips, each respondent has the opportunity to report more overnight trips than in the one-month recall surveys of 2009 and 2010. Despite the increase in trip records in the 2011 file (including those with full details reported by the respondent and those with imputed trip characteristics), the total available for analysis falls short of the number of trip records contained in the 2009-2010 pooled file. Thus, data for some sub-provincial locations may be less reliable for 2011 than corresponding estimates from the 2009-2010 pooled files.

  • The manner in which main purpose of trip is asked has changed. As of 2011, respondents are asked whether a trip was for (1) personal or (2) business or work-related reasons. Subsequently, respondents are asked to provide a more specific reason for the trip such as to visit friends or relatives; for holidays, leisure, or recreation; to go to a conference, convention or trade show (business), and the like. Additionally, routine other business trips that were considered out-of-scope prior to 2011 are now considered in-scope.

    Early indications suggest that these conceptual and wording changes may have altered overall volume estimates and the relative shares of trips by main purpose.

What to Expect

  • The TSRC 2011 data file is scheduled for external audit in November 2012 and for release by Statistics Canada in December 2012.Footnote 6
  • Following the TSRC 2011 release, the TSRC Working Group will explore the feasibility of creating adjustment factors at the national and provincial levels to permit comparisons between reference years 2010 and 2011 (bridging).
  • The TSRC Working Group will also examine options for pooling 2011 and 2012 TSRC files.