Household travel surveys (HTS) are traditionally conducted every decade or so. While providing fundamental inputs for transportation and land use planning and modeling, the infrequency of these comprehensive and large sample surveys pose challenges to North American transportation agencies. A growing obstacle is successfully committing a large budget for a “once in a decade” effort, and anticipating data requirements for the next decade to keep up with modeling needs that increasingly demand fresh and more detailed data.
To overcome these challenges, some agencies have started moving towards a more frequent survey approach used in Europe and Australia – continuous surveys. This presentation focuses on two case studies where agencies are implementing a continuous HTS approach. The first case study is the four-county Seattle region, where the Puget Sound Regional Council (PSRC) collected 6,000 households in spring 2014 and will collect about 2,200 households in spring 2015 such that the spring 2015 data collection will be a mix of new cross-sectional data and 2014 panel households. PSRC’s intent is to evaluate how best to conduct a HTS every two or three years, while also separately conducting targeted supplemental survey efforts such as PSRC’s planned 2015 sub-sample smartphone GPS data collection effort and a special generator university travel diary. The second case study is the fast-growing greater Calgary region which is conducting continuous data collection from March 2015 through December 2016 for 24 months of data collection for 3,000 households and a sub-sample smartphone GPS data collection effort in order to assess the how best to conduct ongoing data collection in 2017 and beyond.
Beyond comparison and discussion of the approach of these two case studies, we will also discuss the implications for project planning and administration, data deliverables, and processes for implementing updates and edits over the course of these projects. The value of richer, longitudinal data from continuous surveys comes with the need to automate and implement efficiencies so that costs per sampled household are ultimately not significantly different from traditional cross-sectional HTS approaches and thereby support the extra value the data provides agencies.