WELCOME 2004 Florida Commuter Choice Summit What’s New in TDM Research Philip L. Winters TDM Program Director Center for Urban Transportation Research University of South Florida Overview Recently completed research (partial list) Commuter Choice Program Case Study Development Worksite Trip Reduction Model and Manual Clearinghouse Price Elasticity of Rideshare: A Case Study for Vanpools Analyzing the Effectiveness of Commuter Benefit Programs: A Descriptive Analysis Approach Research in progress Research about to begin Commuter Choice Program Case Study Development and Analysis Sara J. Hendricks, AICP Study Results Will Help You: Target Most Receptive Work Sites Provide Tips to Employers to Increase Work Site TRP Effectiveness Provide Tips to ETCs Study Questions What makes work site trip reduction programs successful? What explains the other 82 percent of variance in effectiveness? Hypothesis: Work site trip reduction program effectiveness influenced by work site organizational culture. Research Results The null hypothesis is sometimes true. Supportive management and an effective ETC is necessary if there is poor access to high quality transportation alternatives. The most effective ETC usually cannot overcome lack of management support. The worst ETC usually cannot undermine a work site TRP that has management support. What We Learned External factors usually trump effects of internal organizational culture. ETCs shoulder great responsibility, but are powerless, unsupported. Where ETCs can make a difference, their work style influences their success. Relative Importance of Factors Contributing to TRP Success Work site has access to high quality transit Large staff for whom cost of transportation is more important than time savings and convenience Top management support Effective ETC For More Information Final Report Available from the National Center for Transit Research at the University of South Florida in pdf and HTML versions Streaming on-demand presentation http://www.nctr.usf.edu Worksite Trip Reduction Model and Manual Philip L. Winters Rafael Perez, PhD. Ajay Joshi Jen Perone Data Collection Compile 6,000+ worksite trip reduction plans from employers with 100 or more workers that have been developed and tracked for several years Southern California State of Washington Pima County (Tucson) Data Summary Over 40% of worksite trip reduction plans showed modest reductions (up to 7 vehicle trips reduced per 100 employees) over approximately one-year period About 13% of worksite trip reduction plans had substantial reductions (reduced more than 7 vehicle trips per 100 employees) in vehicle trip rates Variables Results 45.00% Bin Classification Accuracy on bins a2 to a5 40.00% 35.00% 30.00% 25.00% 20.00% 15.00% 10.00% 5.00% 0.00% 58-26-1(NN) Linear model 86-13-1(NN) 77-59-1(NN) Linear model No variable selection force enter regression No variable selection No variable selection force enter regression Equally combined data Equally combined data LA Grouped incentives with records with 'no incentives' removed Washington full sample Tucson full sample ungrouped incentives data ungrouped incentives data Models on Equally Sampled Combined Data & best independant models Accuracy on LA Validation Accuracy on Wash validation Accuracy on Tucson Validation Accuracy on Training set Worksite Trip Reduction Model www.nctr.usf.edu/worksite Clearinghouse TRANS-TDM listserv has 770 subscribers Online Help Desk (over 330 Q&A) Netconferences Paying for Performance: Cash for Commuters (November 4, 2004) Talk the Talk: Communicating TDM in Business Terms (June 3, 2004) Transit-Oriented Development: Possibilities for TDM Professionals (January 27, 2004) Using TDM to Manage Traffic at Special Events (October 15, 2003) Access Management: Expanding the Congestion Management Toolkit (August 20, 2003) Bus Rapid Transit: A New Commuter Choice for your Community. (June 12, 2003) Getting to Yes!: Lessons Learned for Increasing the Effectiveness of Commuter Benefit Programs (December 11, 2002) Making Telework Happen: Tips for an Effective Regional Telework Program NetConference (August 14, 2002) Other resources: Carpool/Vanpool Road Signage report Price Elasticity of Rideshare: A Case Study for Vanpools Francis Wambalaba, PhD, ACIP, Sisinno Concas Marlo Chavarria Elasticity Defined The degree of responsiveness to quantity consumed with respect to price Elastic: Quantity changes easily when price changes Inelastic: Quantity doesn't change easily with changes in price Elasticity = (% Change in Quantity)/(% Change in Price) If elasticity is greater than one (elastic), then a 10% change in price results in a more than 10% change in quantity consumed. If elasticity is less than 1 (inelastic), then a 10% change in price results in less than 10% change in quantity consumed. And if elasticity is equal to 1 (unit elastic), then a change in price by 10% results in exactly the same 10% change in quantity consumed. Direct Point Elasticity Analysis Vanpool Elasticity = ▲Ridership/▲ Cost * Mean Cost/Mean Ridership VOTRAN (Daytona) Elasticity= -1.69 Fare increase from $28 to $30 per person in 2000 10% increase in fares leads to a decrease in vanpool ridership by 16.9% Vanpool Elasticity (Puget Sound)= -0.61 A 10% increase in fares leads to a decrease in vanpool ridership by 6%. Application of More Technical Research Methods Used over 260,000 employee records from State of Washington for 1997 and 1999 Applied logistic regression modeling technique Addresses short-comings of early models Model is based on mode choice, accounting for competing modes Model includes socio-economic predictors such as employee job descriptions Model assesses the impact of subsidy Results Vanpool Cost Odds ratio value of -2.6 $1 increased in vanpool price is associated with 2.6% decrease in the predict odds of choosing vanpool with respect to drive alone Vanpool Subsidy Odds ratio of 1.089 Odds of choosing vanpool with respect to drive alone increase by 8.9% Results Work Status Odds of choosing vanpool increase 50% for administrative employees 23% for technical field employees Fare Elasticity (-0.61) For 10% increase in vanpool price, there is a 6% decrease in vanpool choice with respect to auto Conclusions and Recommendations Elasticity rates for vanpooling vary widely (limited datasets) More likely to be very elastic relative to transit Vanpool industry faces volatile conditions and rapid growth complicating elasticity – influence of fares and subsidy on ridership – and making it difficult to generalize Conclusions and Recommendations Vanpools have a “Tipping Point” where loss of one rider may collapse vanpool group Agencies should have “Vanpool Save” program to sustain short-term fluctuations in ridership to avoid loss of groups Sharp decreases in fares (e.g., employerprovided commute benefits) could increase vanpool ridership but data not available Conclusions and Recommendations Vanpool industry should develop guidelines for comparable data collection More cooperation needed Future models should recognize the multiplicity of factors influencing mode choice More research needed with respect to the effect of on-going subsidies versus temporary discounts Analyzing the Effectiveness of Commuter Benefit Programs: A Descriptive Analysis Approach Philip L. Winters Chris Hagelin Ajay Joshi Under subcontract to ICF Consulting Methodology Benefits (frequency) Isolate worksite records where an individual worksite either introduced or eliminated a benefit Focus on records where other programs did not change (control) Examine changes in vehicle trip rate as well as transit share after either introducing or removing a benefit Compressed work week (37) Direct non-financial benefits (119) Facilities & Amenities (73) Financial benefits other than transit and vanpool (104) Flextime Guaranteed Ride Home (48) Marketing (88) Onsite (145) Parking management (13) Rideshare matching (73) Telecommute (14) Transit Benefits (57) Transit Benefits (no control) (943) Vanpool (51) Data Age of records Quality control issues Representativeness Mandatory program Confounding factors Ignores $ level Weaknesses No self-selection bias Number of examples Strengths Impact of Introducing Transit Benefit Not controlling for changes to other benefits 58% reduced Vehicle Trip Rate following the introduction of transit benefits 400 350 No. Companies 23% decreased VTR by an average of 9 trips per 100 and transit share increased from 1.8% to 2.9% Modest 300 Decrease 250 200 (0 to -5) Modest 150 Large Increase Decrease (0 to 5) 100 (<-5) Large Increase 50 (5+) 0 No Control Impact of Introducing Transit Benefit Controlling for changes to other benefits Worksites that introduced transit benefits were more than twice as likely to have the vehicle trip rate increase as decrease No. Companies 28% reduced VTR following the introduction of transit benefits 25 Modest 20 Increase (0 to 5) 15 10 5 Large Increase (5+) Modest Large Decrease Decrease (0 to -5) (<-5) 0 Intro Transit Benefits - Control for Other Strategies Impact of Removing Transit Benefit Controlling for changes to other benefits 47% reduced Vehicle Trip Rate following the elimination of transit benefits 29% decreased VTR by an average of 7 trips per 100 and transit share remained steady at 0.6% No. Companies 20 Modest 15 Increase Large 10 (<-5) 5 (0 to 5) Decrease Modest Decrease (0 to -5) Large Increase 0 Control - Remove Benefit (5+) Impact of Vanpool Benefit at Southern California worksites Introducing Vanpool Benefit No. Companies 20 Modest 15 Modest Increase Decrease (0 to 5) 10 (0 to -5) Large Large 5 Increase Decrease (5+) (<-5) 0 Introducing Vanpool Benefit 47% experienced a reduction in VTR Vanpool may be found in the most comprehensive (8 other incentives) programs Worksites with the largest reductions in VTR saw their transit share fall by more than 1% point Removing Vanpool Benefit No. Companies 20 Modest 15 Increase Large 10 Decrease (<-5) 5 (0 to 5) Modest Decrease (0 to -5) Large Increase (5+) 0 Removing Vanpool Benefit 28% saw their transit share increase by over 1% point 46% had an average decrease in VTR of 5.9 trips per 100 employees Findings and Conclusions More than likely the introduction of transit benefits may result in a reduction in VTR, but it is not guaranteed Conversely, the elimination of transit benefits does not mean a loss of transit share Findings and Conclusions Transit benefits are most effective when there are fewer other incentives programs to compete for the commuter’s attention Within commuter choice programs, more choices often means more competition between benefits Employers must understand that some benefits complement each other and others compete with one another Partial List of Research in Progress at CUTR Traveling Smart: Increasing Transit Ridership By Automatic Collection (TRAC) of Individual Travel Behavior Data and Personalized Feedback Return on Investment Analysis of Bikes on Bus programs South Florida Commuter Services Evaluation Incorporating TDM into the Land Development Process Teenage Attitudes and Perceptions Regarding Transit Use Impacts of Development on Public Transit Ridership Enhancing the Rider Experience: The Impacts of Real-time Information on Transit Ridership TDM Evaluation and Measurement for Atlanta’s Framework Partners Research About to Begin at CUTR National Smart Transportation Archive Researcher (NSTAR) (case studies) Impact of Employer-based Programs on Transit System Ridership and Transportation System Performance Wireless Video for Instant Access (Wi-Via) Security System National Research - Completed TCRP Report 63: Enhancing the Visibility and Image of Transit in the United States and Canada TCRO Report 102: Transit-Oriented Development: State of the Practice, and Future Benefits TCRP Report 87: Strategies for Increasing the Effectiveness of Commuter Choice Programs New Publications - FHWA Mitigating Traffic Congestion: The Role of Demand Side Strategies Traffic Congestion and Reliability: Linking Solutions to Problems Commuter Choice Primer: An Employer's Guide to Implementing Effective Commuter Choice Programs National Research in Progress Analyzing the Effectiveness of Commuter Benefits Programs TCRP H-25A: Completion Date: December 31, 2004 Update the "Traveler Response to Transportation System Changes" Handbook TCRP B-12A. Completion Date: December 31, 2004 Carsharing: Where and How It Succeeds TCRP B-26. Completion Date: April 9, 2005 Guidelines for Evaluating, Selecting, and Implementing Suburban Transit Services TCRP B-25. Completion Date: April 22, 2005 Understanding How Individuals Make Travel and Location Decisions: Implications for Public Transportation TCRP H-31. Completion Date: August 16, 2005 National Research - Pending Determining the Elements Needed to Create High-Ridership Transit Systems Ensuring Full Potential Ridership from Transit-Oriented Development For More Information Philip L. Winters TDM Program Director Center for Urban Transportation Research University of South Florida winters@cutr.usf.edu (813) 974-9811 Ramifications for Evaluating Work Site TRP Success (what TDM professionals can do) Set realistic trip reduction targets for organizations based upon benchmarking Figure 1: Change in Vehicle Trips Reduced for Participating Work Sites 100 80 60 40 20 0 1995 1997 1999 2001 2003 "A" "B" "C" "D" "E" "F" "G" "H" "I" "J" "K" "L" "M" What TDM Professionals Can Do Encourage employers to locate where there are high quality transportation alternatives Target more receptive organizations Target Receptive Organizations Work site access to good quality transit Large staff for whom transportation cost savings is more important than time savings and convenience Employees remain in an office setting Employees work routine predictable hours Target Receptive Organizations Organizations that: Deal with environmental hazards Want to cultivate a “green” image Have employee recruitment/retention problems Feel a responsibility to take a leadership role ETCs Shoulder Great Responsibility… Most ETCs did not volunteer for job. ETCs required to do duties on own time. ETC duties not recognized in job description. Many ETCs could not identify a supervisor. Performance of ETC duties not part of job evaluation. Administering Commuter Survey onerous. Policy Considerations for Designing TROs (What TDM Professionals can do) Designation of an ETC may not be necessary. For commuter surveys, require a random sample that is representative of the employee population than an across-the-board high response rate. What Employers Can Do to Help Their ETCs Ask for a volunteer ETC Incorporate job duties of ETC into job description Arrange for ETC to report directly to top management, preferably to same supervisor as for other duties Carefully select volunteer with work style that matches demands of the job ETCs Can Make a Difference Profile of ETCs with More Successful TRPs High “Influencing” work style (DiSC™) High Expressed Affection (FIRO-B) Low need for control (FIRO-B) Values Relations over Work (CVAT) Values Flexibility and Political Savvy (CVAT) DiSC™ Instrument Premise 1: No work style is better than another. Every work style makes a valuable contribution. Each person has strengths and weaknesses under varying work conditions. Premise 2: People are capable of adapting their behaviors to fit the needs of a situation. Scenarios for ETC Effectiveness Where Top Management is Supportive Effective ETC Work Style (DiSC™) i Program of incentives does Yes not require active administration Program of incentives requires active administration Hands-off management style Program of incentives needs refining C D S Yes Yes Yes Traveling Smart: Increasing Transit Ridership By Automatic Collection (TRAC) of Individual Travel Behavior Data and Personalized Feedback Department of Computer Science & Engineering, Center for Urban Transportation Research (CUTR), and the National Center for Transit Research (NCTR) TRAC-IT Personal Digital Travel Diary Complete System Wireless Data Connection through Cellular Provider Global Positioning System Satellites TRACIT Server Communication Tower Internet WLAN 802.11b Personal Digital Assistant w/ Global Positioning System and Wireless Connectivity Card Wireless Router Other Sources of Real Time Information Testing Automatically Captures: User Enters: GPS Points Recorded Using Two Different Algorithms – Continuous Update vs. Selective Update by Walking Time Distance Speed Route Trip Purpose Occupancy Mode User Uploads Data to database Next Steps – Deploy Expert System GPS Satellite Alternate Locations Transit Data Server Internet Expert System Database Remote PC