How employer benefits actually move commutes.
A peer-reviewed study of 3,668 downtown Denver employees shows which programs shift behavior, and which quietly push people back into their cars. Run by UC Denver and Downtown Denver Partnership, with data and modeling by the Commutrics research team.
Six numbers every program should know.
Each figure below is a logistic-regression odds ratio, how much more (or less) likely an employee is to choose a mode when a specific benefit is in play.
Secure bike parking lifts biking
Employees with access to secure bike parking are nearly nine times more likely to bike to work, by far the strongest program effect in the study.
Free parking nearly triples driving
Employees with free or subsidized parking are 195% more likely to drive alone, and meaningfully less likely to bike, ride transit, or telework.
Free transit passes double transit use
A free transit pass (EcoPass) makes employees 106% more likely to use transit. A subsidized pass still lifts ridership by 56%.
Intermediate stops drive driving
Employees who run an errand on the way to work are 363% more likely to drive alone, the single biggest behavioral predictor of solo driving in the dataset.
Remote work policy drives telework
A formal remote-work policy makes employees nearly three times more likely to telecommute. Flexible schedules add another +100%.
Gender shapes mode choice
Men are 217% more likely to bike and 31% more likely to use transit. Women are 47% more likely to telecommute, programs should reflect this.
3,668 commuters. One honest look at what works.
In the fall of 2022, Downtown Denver Partnership surveyed downtown employees on 47 questions covering home and work location, demographics, income, employer benefits, and a week of actual trips across every mode. After cleaning, 3,133 responses were used in modeling, an 85% completion rate.
Our team applied frequency analysis, Spearman correlation, and multinomial logistic regression to isolate which programs change behavior, in which direction, and by how much, controlling for age, gender, income, and commute distance. The full study is published in the ASCE International Conference on Transportation and Development 2025 proceedings.
Frequency distributions
Modal share by industry, income, age, to surface the raw patterns.
Spearman correlation
Non-parametric tests link each program to each mode, with p-value filtering.
Logistic regression
Odds ratios quantify how much each program shifts the likelihood of a mode.
Social factor analysis
Age, gender, income tested as moderators of every relationship.
What lifts each mode, and what blocks it.
Odds-ratio results for the four most-studied modes. Green = program increases the likelihood of this mode. Red = program decreases it.
Biking to work
Secure bike parking dominates as the lift. Free car parking dominates as the drag.
Riding transit
Free passes are the headline. Free parking and a lack of program awareness are quiet killers.
Driving alone
Intermediate stops and free parking are the two largest pushes toward solo driving. Remote work and flexible schedules push back.
Telecommuting
Policy and flexibility do the work. Free parking actively discourages it.
Five things to do on Monday.
The same evidence that drives the Commutrics platform. Use it to sharpen any TDM program, yours or ours.
Stop subsidizing parking by default.
Free or subsidized parking is the single most consistent drag on every sustainable mode, it lifts drive-alone by 195%, cuts biking by 63%, transit by 60%, and telework by 31%. If you keep it, charge for it or means-test it. The same dollars redirected to transit passes or bike infrastructure will move far more behavior.
Invest in secure bike parking before you invest in bike events.
No other single program in the study moves a mode as much as secure bike parking moves biking (+789%). It signals seriousness to existing cyclists and removes the top barrier for would-be ones: theft and weather. Pair it with showers and a guaranteed ride home for an even bigger lift.
An EcoPass beats a partial subsidy, and you can measure it.
A fully free transit pass roughly doubles transit use (+106%), while a partial subsidy lifts it by 56%. The gap matters: many partial-subsidy programs pay for the same employees who would have ridden anyway. Track per-pass utilization to see whether you're funding new riders or existing ones.
Treat intermediate stops as a program design input.
Errand-running is the strongest predictor of solo driving in the entire dataset (+363%). Programs that ignore mid-day logistics, such as childcare pickup, groceries, or a stop at the gym, will lose to the convenience of a personal car. Guaranteed ride home, on-site amenities, and last-mile partnerships are the counter-moves.
Design differently for different commuters.
Men, women, younger, older, lower- and higher-income employees do not respond to the same programs the same way. A one-size benefits menu disadvantages the groups your program needs to reach most. Segment your incentives, and report results by segment, to surface and close those gaps.
Read the full study, or cite it.
The complete paper is published in the proceedings of the ASCE International Conference on Transportation and Development 2025, with full methodology, regression tables, and statistical detail.
The same data powers your program.
Every benefit, social factor, and odds ratio in this study sits inside the Commutrics platform, applied automatically to design, optimize, and report on real TDM programs.
The same benefit lands differently by who you are.
Gender, age, and income each move mode choice independently of any employer program. Equitable program design starts here.
Gender
Men and women in the dataset choose modes very differently, and program response is gendered.
Age
Younger commuters lean active. Older commuters lean car-dependent, by a lot.
Income
Income shapes both opportunity and constraint. The relationship with transit and telework runs in opposite directions.