Guidebook for Measuring, Assessing, and Improving Performance of Demand-Response Transportation (2008)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

DRT and its performance are creating considerable interest within the public transit industry. But what exactly is DRT? Why is its performance important, particularly now? If performance is important, how is it measured and assessed for DRT? And once performance measures are calculated, then what? This chapter provides a framework for the Guidebook, with a brief background on DRT and its performance, establishing a starting point for the chapters that follow. 2.1 Start at the Beginning: What DRT? A Family of Definitions Other DRT-related terms have been used, sometimes interchangeably to refer to various forms of public transit that are not traditional fixed-route, fixed-schedule transit services, but rather respond in some manner or form, to individualized requests or demands for trans- portation service. DRT is actually a type of paratransit service; the latter encompassing a wide range of transit services defined, essentially, as services that are not conventionally scheduled, fixed-route tran- sit services, but rather are those types of public transportation that fall in-between the private automobile and conventional fixed-route transit (see Figure 2-1). From a terminology standpoint, what may be confusing is that, since enactment of the Americans with Disabilities Act (ADA) in 1990, much of the focus of those of us concerned with DRT has turned to ADA complementary paratransit service. ADA paratransit is actually a type of DRT that fixed-route transit systems must provide to individuals with disabilities who cannot use fixed-route service because of their disability. Paratransit, however, is the broadest term. DRT is a subset of paratransit, and ADA comple- mentary paratransit is a subset of DRT. DRT—Definitions Building on the definitions from the earlier literature, DRT is characterized by its flexible rout- ing, shared rides, and activation at the initiation of the rider by prior arrangement (as opposed to, for example, street hail for taxis). The Federal Transit Administration (FTA) has defined DRT in the following way for purposes of NTD reporting: 3 C H A P T E R 2 DRT Performance—The Basics

DRT is also known by other terms, particularly in references and publications from earlier years: dial-a-ride, demand-activated transportation, demand-responsive transportation, dial-a- bus, shared-ride paratransit, flexible-route service, and flexible-transport services. Given the range of the DRT definition, it follows that not all systems providing DRT are the same. One common way to distinguish DRT services is by the rider group or groups that are served. For example, DRT systems may serve the general public, or they may serve only specifi- cally defined groups of riders, such as seniors and persons with disabilities, or “transportation disadvantaged persons,” defined within a community to meet local objectives. And, since the early 1990s, there are many DRT systems that provide only ADA paratransit service. This notion of different types of DRT systems has been one focus of the research for this Guidebook, since differences in the type of DRT influence performance. 2.2 Why Measure DRT Performance? In recent years, there has been growing attention on DRT, and systems around the country are under increasing pressure to improve performance because of escalating demand for service and financial constraints. Pressure to improve performance comes from transit boards; city and county councils; state departments of transportation; and transit managers, among others. Community 4 Guidebook for Measuring, Assessing, and Improving Performance of Demand-Response Transportation Adapted from: “Paratransit—Neglected Options for Urban Mobility,” Ronald F. Kirby, et al., The Urban Institute, 1974. Pr iva te C ar Paratransit Services Ta xi Jit ne y Ca rp oo l, V an po ol Su bs cr ip tio n Bu s Sp ec ial ize d/ Li m ite d El ig ib ilit y D RT Ge ne ra l P ub lic D RT De m an d- Re sp on se , “ Di al -A -R id e” Tr ad iti on al Fi xe d- Ro ut e, Fi xe d- Sc he du le Tr an sit AD A Pa ra tra ns it Figure 2-1. Paratransit: service between the private car and conventional fixed-route transit. Demand-response is a transit mode comprised of passenger cars, vans or small buses operating in response to calls from passengers or their agents to the transit operator, who then dispatches a vehicle to pick up the passengers and transport them to their destinations. A demand-response (DR) operation is characterized by the following: a. The vehicles do not operate over a fixed route or on a fixed schedule except, perhaps, on a temporary basis to satisfy a special need, and b. Typically, the vehicle may be dispatched to pick up several passengers at different pick-up points before taking them to their respective destinations and may even be interrupted en route to these destinations to pick up other passengers.

members and riders may also create pressures for improved performance, should day-to-day service not always meet established objectives or riders’ expectations. While pressures to improve performance and address financial constraints are certainly not limited to DRT, the pressures may be more severe for DRT systems than for their fixed-route counterparts, given the significantly higher per-passenger trip costs for DRT: national data show that a passenger trip on DRT costs more than eight times that of a passenger trip on a bus (see Figure 2-2). Moreover, efforts to improve DRT performance may increase demand, which then may require additional service and resources, in turn increasing rather than containing or decreas- ing operating costs. Unlike a fixed-route system that can absorb increased ridership (improved performance) on its routes until the buses are full and no more standees can fit (with a very low marginal cost per additional passenger trip), a DRT system may have to add resources to serve additional trips. Each new trip responds to a rider’s individualized request, with a new origin and destination. The marginal cost of each additional trip may be as much as the full cost of a passenger trip. Responding to such pressure requires that, first, DRT systems measure their performance, establishing a baseline. DRT systems must then assess that performance, and depending upon the assessment, take additional steps, as appropriate, to improve performance. Performance Measurement Much has been written for the transit industry on performance measurement. TCRP Report 88: A Guidebook for Developing a Transit Performance-Measurement System, provides useful information to transit agencies and managers for developing a performance-measurement sys- tem, using traditional as well as less-traditional measures. Essentially, the performance measurement process starts with establishing goals and objec- tives for service; identifying selected measures or indicators that capture key aspects of DRT ser- vice and operations; measuring performance through the collection and tabulation of the data needed to calculate the measures; assessing the resulting measures; and developing actions to address any deficiencies. The process should also involve a feedback loop so that the goals and objectives are reviewed periodically after performance has been measured and assessed. It may be that, over time, a par- ticular goal or objective needs to be adjusted based on service operations or changes to the overall environment within which the DRT system operates (see Figure 2-3). DRT Performance—The Basics 5 $13.76 $14.54 $15.06 $16.09 $16.74 $18.38 $20.76 $21.74 $22.92 $23.92 $2.00 $2.05 $2.04 $2.07 $2.19 $2.27 $2.39 $2.59 $2.71 $2.81 $0.00 $5.00 $10.00 $15.00 $20.00 $25.00 $30.00 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Demand Response Fixed Route Bus Source: National Transit Database. Figure 2-2. Operating cost per passenger trip, national data 1996–2005.

For purposes of this Guidebook, another step in the performance assessment process for DRT is developing clear definitions for the performance data and for performance measures. DRT systems do not always use the same definitions for data elements, nor do they calculate performance meas- ures in the same way. For example, some systems calculate their productivity using total vehicle hours as the denominator and others use revenue hours. Comparisons between DRT systems that use these different calculations for determining productivity are less meaningful since the measures are not measuring the same thing. Some of the differences in both terminology and calculation of measures have arisen over time, and some have come through common software applications and new calculations now possible with the powerful data capabilities of computer-assisted scheduling and dispatch systems that many DRT systems commonly use. It is perhaps even more important to now establish com- monality of terms and definitions, as the use of software proliferates and opportunities created for yet new terms and calculation practices. 2.3 Measuring DRT Performance—Issues and Complications While the chart depicted in Figure 2-3 seems relatively straightforward, the process of meas- uring DRT performance can be complicated for a variety of reasons. First, there is the issue of inconsistent data and data collection: • Data definitions vary. What one DRT system uses to compute a performance measure, miles, for example, may differ from another; the first system may use a figure based on odometer readings, while another calculates miles from pull-out in the morning to pull-in later in the day. Which definition is the right one to use? 6 Guidebook for Measuring, Assessing, and Improving Performance of Demand-Response Transportation Establish Goals and Objectives for DRT Select Performance Measures Collect Data and Calculate Measures Assess Resulting Performance: Trend Analysis Peer Comparison, Other As Needed, Consider Improvement Actions & Strategies Monitor Performance on On- Going Basis Review Goals and Objectives Periodically, Adjust As Needed Figure 2-3. DRT performance measurement process.

• Data collection is not always rigorous. DRT systems that contract some service to taxi com- panies count the passenger trips provided by taxis, but not necessarily the revenue time, since taxi companies typically charge by mileage rather than time and may not report the time. When the DRT system computes its productivity, the figure will be somewhat over- stated, as the taxi passenger trips are included, but the revenue time is not. Or a DRT system may count only its ADA riders when totaling up its passenger trips. If it does not also include the companions and Personal Care Attendants (PCAs), its productivity will be somewhat understated. Second, DRT performance is affected by the type of riders it serves, its operating environment and other factors, some of which cannot be controlled. How can performance measurement account for such variables? • The different types of DRT systems have differing characteristics that affect their operations and resulting performance. A DRT system serving the general public in a small community has a different operating environment than does an ADA paratransit system serving a major metropolitan region, and the performance of the two systems will differ. • Various factors impact DRT performance, and some of these are outside the control of the DRT system. A DRT system that serves a large geographic area with a wide dispersion of ori- gins and destinations will have longer trip lengths than a system that serves a more compact service area with shorter trip lengths. Systems with longer trip lengths will not be able to serve as many passenger trips in a given amount of time, thus their productivity (as measured by passenger trips per hour) will be lower. Third, goals and objectives for a DRT system may be contradictory, establishing a service dynamic where efforts to achieve one objective adversely affect achievement of another. In a sim- ilar way, policies and procedures may also influence performance in contradictory ways. Examples of contradictory objectives can be found at many DRT systems: • Maximizing Productivity versus Achieving High On-Time Performance (OTP) Operationally, if the DRT system schedules passenger trips tightly in an effort to maximize the number of passenger trips served within a given number of service hours, OTP is likely to suffer as actual service on the street may not meet the tight schedule: riders may take longer to get out to the vehicle for their trip than was planned; they may take longer to board than planned; the operator may take extra time getting to a pick-up location because he or she took a wrong turn; traffic may be worse that day; and so on. All of these will serve to undermine a tightly developed schedule and adversely impact on-time performance. • Minimizing Telephone Response Time for Trip Reservations versus Maximizing Customer Service DRT systems often establish an objective that a certain percentage of calls for reservations should be answered within, for example, 2 or 3 min. However, depending on call volume and the number of call-takers, this may force call-takers to abbreviate some calls. To the extent that riders need more time on the phone to explain their trip needs or if they have special needs, the response time objective may work at cross purposes toward an objective of maximizing customer service. Finally, DRT systems may also have policies, practices, or procedures that are contradictory. Examples include: • Having operators wait up to 10 min at pick-up locations for riders versus ensuring travel times are not lengthy. If a DRT system routinely has its vehicles wait for longer periods of time at pick-up loca- tions, either because of the established wait time or because riders are slow in appearing for their trips, the travel times of other riders already on the vehicle will be adversely affected. DRT Performance—The Basics 7

• Defining a 30-min on-time window for timely vehicle arrival versus having call-takers inform riders once their trips are booked: “the vehicle will come 15 min before to 15 min after the scheduled time.” Many DRT systems do not adequately educate and inform their passengers regarding when to be ready for their pick-ups and how the “on-time window” works in practice. Once a rider sched- ules her 9:00 a.m. pick-up time and is told, “You need to be ready 15 min before to 15 min after the scheduled time for your vehicle to arrive,” what she likely remembers is her scheduled time— 9:00 a.m.—not 8:45 a.m. and 9:15 a.m. Experience has shown that many DRT riders believe that the vehicle is early if it arrives 15 min before their scheduled time (after all, it’s before the sched- uled time) and in some cases, the rider will not be ready to depart. When this happens, the rider may be deemed a no-show, which then may require the DRT system to schedule another vehicle to go back and pick up the rider later. Or the operator may be instructed to wait for the rider, typ- ically longer than the established wait time. In both cases, performance is negatively impacted. A DRT system may not be able to establish completely balanced and complementary objectives, policies, and procedures, given the nature of public transportation in general and DRT in specific. But clarity in establishing those objectives, policies, and procedures is important for effective performance measurement. If a DRT system strives to achieve a very high OTP percentage, for example, it should be clear that this will likely have a negative effect on productivity, other things being equal. Beyond clarity in establishing objectives for the DRT system, it is also important to under- stand the characteristics and factors that influence an individual DRT system, only some of which can be influenced or “controlled” by a DRT system. Understanding Characteristics and Factors Influencing DRT Performance Understanding the different service characteristics and factors that influence DRT performance may help the DRT manager and DRT governing body put their system’s actual performance into a more realistic perspective. Understanding the sometimes-conflicting nature of objectives, policies and procedures may help the DRT system focus on those most important areas or at least recognize that achievement in one performance area may hamper achievement in another. This Guidebook can serve to help DRT managers and policy makers with that understanding. Chapter 7 of the Guide- book includes information on the specific factors that influence DRT performance, identifying those that DRT systems can influence to improve performance. And, as any DRT manager knows, there are some factors that cannot be influenced, but rather should be recognized for their impact on ser- vice and performance. Beyond helping with this understanding, a primary goal of the Guidebook is to provide a how-to resource for DRT performance measurement. What are the key data that should be used? How are the data elements defined and what data are used for what measures? And once the performance data are calculated, what’s next? The Guidebook provides sample data from a wide range of DRT systems from around the country, providing reference points for DRT systems assessing their own performance. Addi- tionally, there is information gleaned from these sampled systems about actions and strategies to improve performance. Again, these data can serve as reference points for other DRT systems interested in similar types of actions that can be considered for improving performance. Remember the Big Picture Since the topic of DRT performance measurement necessarily involves such details as how to measure revenue hours, the mechanics of the process may sometimes overshadow the big picture. 8 Guidebook for Measuring, Assessing, and Improving Performance of Demand-Response Transportation

For DRT systems, that big picture involves providing transportation to community members who, for different reasons, need the more personalized transportation provided by a demand-response system or, in communities with fixed-route service, may not always be able to use this more tradi- tional transit service. But providing transportation can involve more than operating DRT service day-to-day. It may also involve working with other community organizations to support other specialized transportation services in the area. If the DRT system is part of a larger transit organi- zation, it means that DRT managers work together with their fixed-route counterparts to ensure that the transit system as a whole is available and accessible to all community members. Moreover, the big picture also means that the DRT system must operate within its policy, financial, and environmental framework. Measuring performance is an important way for the DRT system to chart its course within that framework. DRT Performance—The Basics 9