| Survey Year | First Travel Date | Last Travel Date | Unweighted Counts | ||||
|---|---|---|---|---|---|---|---|
| Households | Persons | Travel Days | Trips | ||||
| 2019 | 2018-10-05 | 2019-09-29 | 7,516 | 15,434 | 79,556 | 335,284 | |
| 2021 | 2021-06-22 | 2022-02-05 | 7,907 | 15,031 | 49,567 | 180,242 | |
| 2023 | 2023-01-12 | 2024-01-16 | 3,749 | 7,280 | 28,838 | 108,413 | |
| TOTAL | — | — | — | 19,172 | 37,745 | 157,961 | 623,939 |
Travel Behavior Inventory 2018-2024 Report
Executive Summary
Note
You can download a PDF version of this report here.
The Travel Behavior Inventory is a study of household demographics, daily travel activities, and typical transportation patterns throughout the greater Twin Cities region.
From 1949 to 2010, the study was conducted every ten years. In 2018, the Metropolitan Council (Met Council) transitioned to a recurrent, every-other-year study program. The recurrent program collected data for approximately 12 months of data collection every other year (Table 1). This recurrent study introduced efficiencies and operational improvements enabling the Metropolitan Council to collect more current, detailed data for use in their travel models.
This report synthesizes data from the first six years of the recurrent program (2018-2024), encompassing three waves of survey data collection.
More than 37,000 people, comprising over 19,000 households, participated in the 2018-2024 TBI (Table 1).
Study Goals
Key objectives of the Travel Behavior Inventory include the following:
- Collecting core information on household demographics, typical travel behavior, and regional transportation patterns to support the Met Council’s transportation modeling and planning needs.
- Capturing information about new transportation modes and behaviors to keep pace with rapid changes in the transportation industry.
- Employing a thorough, multi-pronged approach to reach a representative sample of the population in the final dataset.
- Leveraging new technologies and methods to reach a more complete, detailed, high-quality dataset.
Methodological Highlights
This section describes three methods employed by the Study to collect accurate and representative travel behavior data: smartphone-based data collection, online travel diaries, and calling into the survey call center. All three of these methods represented methodological improvements over previous iterations of the Travel Behavior Inventory.
Smartphone-based travel diaries improved data quality and quantity.
Households with smartphones were required (2019) and/or encouraged with higher survey incentives (2021, 2023) to complete their travel diaries using the rMove™ smartphone app for up to seven consecutive days. The number and share of household who completed their travel diaries using rMove™ is shown in Table 1.2. Households without smartphones or households who were not willing to participate using rMove™ participated by completing their travel diary online (rMove™ for Web) or by calling into the survey call center. These households reported travel for one day (Tuesday, Wednesday, or Thursday).
Compared to diaries completed online or by telephone interview that rely on recall, smartphone-based travel diary collection offered significant benefits for data quality and quantity (e.g., detailed trip paths, and lower degrees of under-reporting).
| Survey Year | Households with rMove Travel Diaries | |
|---|---|---|
| Number | Percent of Total | |
| 2019 | 5,026 | 66.9% |
| 2021 | 3,242 | 41.0% |
| 2023 | 1,902 | 50.7% |
Multiple days of data collection capture atypical travel behaviors.
While multi-day data collection increases the volume of data collected per household, it also helps capture a wider variety of behaviors.
Figure 1.1 and Figure 1.2 below show the variation of trip mode and purpose for rMove™ respondents by the number of reported days of travel. Increasing the number of days completed increases the number of non-walk, non-vehicle trip modes included in the dataset, (Figure 1.1) and greatly increases the number of trip purposes other than home, school, or work included in the dataset (Figure 1.2).
Address-based sample representativeness improved over time.
Most of the sample recruitment was accomplished through address-based sampling (ABS), a type of probability sampling, with a focus on reaching segment-level recruitment goals. The sample plan was reviewed and refined on an iterative basis, both within and between survey waves, to improve response from typically hard-to-survey groups. (For an example, see the 2021 Survey Management Plan).
Major efforts were made in the initial pilot (2018) and into the 2019 wave of data collection to recruit a large and representative sample. The team broadly disseminated information about the survey online, tested invitation materials and incentive amounts, and even tried door-to-door followup with survey non-respondents (See 2019 Survey Methodology Memorandum). Although the 2019 survey exceeded its sample size targets, the overall demographic composition of the survey was more biased than desired, especially related to race: only 2.5% of the people surveyed in 2019 identified as Black or African-American, compared to 8.1% in the 2015-2019 American Community Survey (ACS; see Table 6).
In the 2021 wave of data collection, the study team made a number of improvements to survey design, incentives, outreach, and supplemental sampling to recruit a more representative sample - with a special focus on race, ethnicity, and income (See 2021 Survey Methodology Memorandum). The percentage of the sample identifying as White decreased from 2019 to 2021, bringing the total sample composition more in line with Census estimates (Table 6), but the team felt there was more room for improvement.
Differential incentives and outreach efforts in 2023, however, did not yield further improvements to sample representativeness over those afforded in 2021: the percentage of respondents identifying as Black or African American remained at approximately half of Census estimates (Table 6).
Taken as a whole, these results point to a need for future waves of the Travel Behavior Inventory to implement further modifications to boost sample representativeness. Alternative methods of recruitment beyond mail, invitation redesign, panel sampling, and strategic oversampling are all possible avenues of exploration.
| Race | 2019 | 2021 | 2023 | 2015-2019 ACS | |
|---|---|---|---|---|---|
| American Indian or Alaska Native | 0.4% | 0.7% | 0.7% | 0.5% | |
| Asian | 3.3% | 4.9% | 5.1% | 6.2% | |
| Black or African American | 2.5% | 5.2% | 4.2% | 8.1% | |
| Native Hawaiian or other Pacific Islander | <0.1% | <0.1% | 0.2% | <0.1% | |
| Two or more races | 2.7% | 3.1% | 3.1% | 3.2% | |
| White | 89.9% | 84.8% | 84.9% | 77.7% | |
| Other race, ethnicity, or origin | 1.2% | 1.2% | 1.8% | 4.2% | |
| Total | — | 100.0% | 100.0% | 100.0% | 100.0% |
Key survey findings
Number of trips
In 2023, the average resident made 3.48 trips on a typical weekday. This trip rate was below that of 2019 (4.17 trips per day), but represented a rebound from the 2.85 trips per day in 2021 (Table 1.4).
| Survey Year | Unweighted | Weighted | ||||
|---|---|---|---|---|---|---|
| Trips | Days | Trip Rate | Trips | Days | Trip Rate | |
| 2019 | 329,021 | 79,556 | 4.18 ± 0.02 | 15,348,964 | 3,682,918 | 4.17 ± 0.04 |
| 2021 | 176,093 | 49,567 | 3.48 ± 0.02 | 10,670,349 | 3,748,414 | 2.85 ± 0.04 |
| 2023 | 104,611 | 28,838 | 3.68 ± 0.03 | 13,038,925 | 3,750,006 | 3.48 ± 0.05 |
| Error margins are +/- 1 standard error of the mean. | ||||||
Trip purposes
Despite the extraordinary effects of the COVID-19 pandemic on daily life and work culture, the share of trips made by purpose stayed mostly consistent from 2019 to 2023 (Figure 1.3). Across all three years, the most common trip purpose was work or work-related, followed by social/recreational trips and shopping trips.
From 2019 to 2021, the share of trips made for work and work-related purposes declined from 26% of trips to 23% of trips, then declined again slightly to 22% of trips in 2023. Conversely, the share of trips made for shopping and errands increased from 2019 to 2021, then declined from 2021 to 2023.
Comparisons of the share of trip purpose by year must be made carefully, because overall trip rates declined during the COVID-19 pandemic and did not fully recover (Table 1.4). An alternative measure, the participation rate, can be used instead: this measure represents the share of people who make any given type of trip on a typical day (Figure 1.4), and as such accounts for both changes in the types of trips people make and the changes in the amount of travel people make.
The trends in Figure 1.4 suggest a number of underlying processes at work.
The effects of telework on travel behavior are easy to spot: the share of people making work and work-related trips declined from 2019 to 2021, then slightly rebounded from 2021 to 2023 (Figure 1.4). Work-related trips (to meetings, deliveries, worksites) took a greater hit than trips to and from primary workplaces. From 2019 to 2023, the share of people making trips to work-related locations fell from 16% to 8%. Teleconferencing could be replacing trips that used to be made in person.
The fingerprint of e-commerce is also visible in the data. For example, the percent of people making a routine shopping trip on a typical day fell from 14% in 2019 to 10% in 2021, and then to 9% in 2023 (Figure 1.4).
Finally, the lingering effects of the pandemic on social life are also apparent. The share of people making a social visit trip more than halved from 2019 to 2021, then rebounded only slightly in 2023 (Figure 1.4) . Similarly, the share of people making trips to exercise or walk outside declined from 2019 to 2021.
Travel choices
Driving remains the predominant mode of travel in the region, representing 83% of trips in 2023. Apart from a decrease in transit mode share from 2019, the share of trips made by each mode was stable throughout the entire six-year study period.
Explore the Data
To facilitate data exploration and analysis, the study team:
- Created a cross-wave dataset and codebook, available for download by the public (Met Council: link when dataset is published).
- Developed an interactive app for data exploration and crosstabs (link).
- Developed an open-source R package to support analysis of household travel survey data (link).
These resources are open-source and freely available. For support in accessing or using the data, or any elements of this report, please contact Met Council Public Information.