Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance (2009)

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.

46 This chapter presents performance data from the representative rural DRT systems that par- ticipated in the research project. These data provide benchmarks, allowing comparisons among similar types of DRT systems. 6.1 Rural Systems Participating as Representative Systems To assist DRT systems in comparing their performance against other systems, the Guidebook provides data on key performance measures for a representative number of rural DRT systems, categorized by the typology defined in Chapter 5. Finding Representative DRT Systems To find representative DRT systems within the categories of the typology, various DRT sys- tems of different types and in different parts of the country were contacted for participation in this research project. The objective was to provide valid reference points for each category, not to provide a statistically balanced sample of systems. There was an attempt made to provide geo- graphic diversity of representative DRT systems, but no attempt to find high performers or low performers to frame the data. Collecting Data from Representative DRT Systems Once representative DRT systems agreed to participate, the researchers asked for the key per- formance data, with Fiscal Year 2007 as the target year (although the specific months defining a fiscal year varied), using both on-site visits and telephone interviews. Information about the sys- tem, its service, and its operating environment was also collected to develop an understanding of the factors and circumstances affecting each system’s performance. It was agreed with the par- ticipating systems that the research report would not relate specific performance data to individ- ual systems. The 24 participating systems are listed in Table 6-1, and their locations are shown in Figure 6-1. The data elements requested from the participating DRT systems include passenger trips; vehicle-miles; vehicle-hours; operating costs and safety incidents, as defined by Rural NTD; and on-time performance. Data for passenger trips, miles, operating costs, and usually hours were readily available from the participating systems; however, there was very limited data on NTD- defined safety incidents. This is not to say that all the systems had perfect safety records, but not all the participating systems collect safety data conforming to Rural NTD definitions and for those that do, there was little data to collect. C H A P T E R 6 Performance Data from Representative Systems

Regarding on-time performance, again, there was little data to collect from the participat- ing systems. The majority of the DRT systems report that pick-up and drop-off times are recorded, usually with vehicle operators writing arrival times at their stops on their manifests. However, these data are not summarized from the manifests for any operating reports on a routine basis. Typically, the DRT manager stated that the data were available should there be a question or complaint about vehicle timeliness, and they check the applicable operator manifest for the data should an issue arise. Only one of the participating systems had formal on-time perfor- mance percentages to report. Rural DRT systems, however, should be encouraged to collect and report on-time performance statistics for service monitoring and performance assessment. Timeliness at both the pick-up end and drop-off end are important and critical for evaluating DRT service reliability. Performance Data from Representative Systems 47 DRT System Location Provider and Service Area Albert Lea Transit Albert Lea, MN Private non-profit agency serving developm entally disabled adults, serving prim arily a single city ALTRAN Munising, MI County, serving primarily a single county Atomic City Transit Dial-A-Ride Los Alamos, NM City, serving primarily a single city Bay Transit Urbanna,VA Private non-profit senior agency, serving 10 counties in eastern Virgin ia B.C. Country Rural Dial-A-Ride Vestal, NY County, serving primarily a single county Central Florida Regional Planning Council (RPC) Transportation Disadvantaged Program Bartow, FL Regional planning organization, with a contracted pr iv ate service operator, serving 3 counties in central Florida City of Cleburne Transportation Cleburne, TX City, serving prim arily a single city Columbia Area Transit (CAT) Hood River, OR Transit district, ser vi ng prim ar ily a single county Dial-A-BATS Bullhead City, AZ City, serving primarily a single city Fresno County Rural Transit Agency (FCRTA) Fresno, CA Joint powers transit agency created by agreement between participating cities and county, serving the county Garrett Transit Service Oakland, MD Privat e non-profit comm unity action agency, serving primarily a single county Hancock Area Transportation Services (HATS) Findlay, OH Private non-profit community action agency, se rv in g primarily a single county Hill Country Transit District San Saba, TX Transit district, affiliated with non-profit community action agency, serving 8 counties in central Texas Inter-County Public Transportation Authority Elizabeth City, NC Private non-profit health services agency, serving 5 counties in northeastern North Carolina Intracity Transit – Paratransit System Hot Sp rings, AR City, serving primarily a single city Johnson County Transportation Cleburne, TX Provided through City of Cleburne, serving prim arily a single county McIntosh Trail Transit System Griffin, GA Regional planning organization, with service operated by a non-profit contractor, serving 5 counties in central Georgia Monroe County Shared Ride Scotrun, PA Transit district, ser vi ng prim arily a single county Moscow Valley Dial-A-Ride and Paratransit Moscow, ID Private non-profit transp ortation agency, serving prim arily a single city Paul Bunyan Transit Bemidji, MN Private non-profit transit organization created by joint agreem ent between participating city and count y , serving primarily a single county Pulaski Area Transit Pulaski, VA Private non-profit senior agency, serving primarily a single city Regional Coordinated Area Transportation System Asheboro, NC Private non-profit senior agency, serving 2 counties in central North Caro lina River Cities Public Transit (RCPT) Pierre, SD Private non-profit transportation agency, serving 10 counties in central South Dakota VTA Paratransit Edgartown, MA Transit authority, with service operated by a private c ontractor, ser ving primarily a single county Table 6-1. Rural DRT systems participating in the research project.

Figure 6-1. Representative rural DRT systems participating in research project. Oakland Hood River Moscow Los AlamosBullhead City Fresno Cleburne San Saba Hot Springs Griffin Bartow Asheboro Pulaski Findlay Albert Lea Bemidji Pierre Munising Vestal Edgartown Scotrun Urbanna Elizabeth City TX CA MT AZ ID NM NV CO OR UT IL WY KS IA SD NE MN ND OK FL WI MO WA AL GA AR LA MI IN NC NY PA MS TN KY VA OH SC ME WV VT NH MD NJ MA CT DE RI DC I I I I I I I LEGEND Serving Primarily a Single City Serving Primarily a Single County Serving Multi-Counties

In analyzing the collected data, the researchers found that data for other service modes were sometimes commingled with that for DRT and that the reported performance data did not always match the definitions being used for the project. The most common data issue found was that, where rural systems operate demand-response and route deviation service, the system reported combined operating data, so the data for the two modes were commingled. In such cases, additional information was requested from that system so that the route deviation data could be separated out, with only demand-response data remaining. The other issue related to the reporting of vehicle-hours. As was found during the urban phase of the research project, some systems substitute operator pay-hours or some other hours data for vehicle-hours. Again, additional information was requested so that the data could be adjusted to conform more closely to the data definitions being used for the Guidebook. Where adjust- ments were made, they were reviewed with the specific DRT system to ensure agreement, and all adjusted data are noted where appropriate in this chapter. Finally, despite efforts to adjust the performance data, the statistics for one of the participat- ing 24 DRT systems are not included in the peer data presented in this chapter. 6.2 Comparing Your Performance Against Other Systems—Performance Data of Representative Rural DRT Systems This section provides the performance data from the representative systems for four of the key measures—passenger trips, vehicle-hours, vehicle-miles, and operating costs—using “stock” graphs. For each of these measures—passenger trips per vehicle-hour, operating cost per vehicle- hour, operating cost per vehicle-mile, and operating cost per passenger trip—the performance data are shown within the three categories of rural DRT, as defined by the typology developed through this project (and described in Chapter 5). In addition to the data from the representative systems, data from the Rural NTD Report Year 2007 are shown on the graphs, providing data from a much larger sample of rural DRT systems. With access to the 2007 Rural NTD dataset, data for rural systems operating DRT were extracted for analysis. In order to assess cost performance measures, rural systems operating only DRT were selected because rural reporters provide total operating cost without any mode-specific cost data. Without mode-specific costs, it is not possible to assess DRT only for those rural systems that operate DRT and other modes. For each of the graphs presenting data from the representative rural DRT systems and for each of the three service-area types, the range and average of the rural DRT-only NTD reporters are shown. These same data are shown in table form in Appendix A. It should be noted that the 2007 Rural NTD data were reviewed prior to inclusion in this project, and DRT systems with incom- plete data or with datapoints far outside what would be expected were deleted, generally data more than two standard deviations from the mean. While the Rural NTD data did not receive the same level of scrutiny that was possible through this research project for the participating DRT systems, inclusion of the NTD data alongside that of the systems from the research project provides a larger framework for rural DRT performance data. Passenger Trips Per Vehicle-Hour—Productivity Productivity may be the most important single performance measure for a DRT system. Data from the representative systems, as well as data from the Rural NTD, show that productivities are generally somewhat higher for rural DRT systems serving smaller service areas (see Figure 6-2). Performance Data from Representative Systems 49

The productivities of those systems serving predominately a single municipal area tend to be higher than those of systems serving predominately a single county, which in turn trend higher than those systems serving multi-county areas. This would be expected, as systems serving smaller service areas will have shorter trip lengths, so more trips can be served in a given vehicle-hour. DRT systems with smaller service areas would also be expected to generally have less deadhead time, and this will benefit productivity since the productivity measure for rural DRT uses vehicle- hours, as opposed to revenue-hours, in the denominator. A few of the representative systems deserve note for their high or low datapoints compared with the other representative systems. The municipal DRT system with productivity of 7.05 oper- ates in a small and compactly developed geographic area, and some of the service appears to operate less like demand-response and more like route deviation. This is an informal arrange- ment that seems to be a carry-over from when the service operated more as fixed route. The municipal system with the lowest productivity—2.38—is an ADA paratransit service. The county system with productivity of 6.23, which is the high point for primarily single- county systems, operates as immediate response and, significantly, is actually a composite of a number of smaller, community-based DRT systems that are linked together in a very large ser- vice area with intercity routes. While the productivities of the different community DRT systems within the county vary, together they average over six passengers per vehicle-hour. From a per- formance reporting perspective on this measure, this county system should more appropriately be seen as a number of primarily municipal systems. 50 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance 2.38 2.86* 1.91 2.96* 3.18 2.14* 7.05 2.53 4.34* 3.19* 5.75* 3.15 4.01 2.57* 2.85 3.53 6.23* 3.56 4.26 1.57 2.43 2.06* 2.61 NTD AVG (4.44) NTD AVG (3.13) NTD AVG (2.91) NTD MIN RANGE (1.49) NTD MIN RANGE (0.49) NTD MIN RANGE (0.61) NTD MAX RANGE (9.93) NTD MAX RANGE (8.98) NTD MAX RANGE (8.31) 0 1 2 3 4 5 6 7 8 9 10 Municipality County Multi County PA SS EN G ER T RI PS P ER V EH IC LE H O UR TYPE OF SERVICE AREA B31A Representative Data, 2007 Rural NTD data for DR only systems, 2007 Report Year * Adjusted Data Figure 6-2. Rural DRT systems: passenger trips per vehicle-hour.

The county system with a productivity of 5.75, also considered high for a county-based service, operates predominately as immediate response; with the vehicles equipped with AVL and MDTs, the dispatchers proactively dispatch service in real-time, sending out trips to the operators about 30 min in advance. With the technology and skilled dispatchers, “the drivers are pushed most of the time,” according to the system manager. For the multi-county rural systems, both the high and low productivity points can be noted. The high productivity of 4.34 passenger trips per vehicle-hour is found at a system where, while the service area is a large 10-county region, the majority of the trips are provided within a 5-mile radius of the primary community. Also, the system serves a significant number of school-aged riders, providing service to and from daycare and other before/after school destinations. This rural system has actively sought out such trips, which, since they are “many-to-one” and “few-to-one,” will improve productivity. The low productivity datapoint within the multi-county service-area category at 1.57 passen- ger trips per vehicle-hour is a result of several factors for that particular rural system. Chief among them is the fact that almost one-fourth of its trips are for Medicaid purposes, and many of these go to destinations beyond the primary three-county service area. Additionally, the system uses a taxi company as one of three primary service contractors, and the taxi trips tend to be single-ride. Operating Cost Per Vehicle-Hour Operating cost per vehicle-hour is a key cost-efficiency measure. Data from the representative systems are shown in Figure 6-3. It can be seen that systems in the predominately municipal and predominately single-county categories show similarities on this performance measure, ranging from $32 to $35 per hour up to $74 to $78, with clustering roughly between $40 to $49 per vehicle-hour. The ranges shown for the Rural NTD data are also similar for the two service- area-type categories. Data for the multi-county rural systems, however, show somewhat lower cost per vehicle-hour figures. This may be explained, at least in part, by the fact that most of the multi-county systems are non-profit organizations, where, according to recent research, compensation is less than at governmental units (e.g., municipalities or counties) and transit districts (15). This impacts their cost structure since labor costs are the major component of operating costs. Operating cost differences may also result, to some extent, from cost allocation procedures; systems that are part of a larger organization and particularly a multi-purpose human service agency may not fully allocate costs to their transit service. A transit district, on the other hand, will have the full complement of functions needed for transit within one organization and full cost accounting is more typical. The high datapoints in Figure 6-3 merit note. For the predominately municipal service-area category, the high point of $74.04 per vehicle-hour is influenced by the fact that this is an ADA paratransit system, operated by a small city where there is labor market competition for vehicle operator positions and where there is no wage or training distinction between fixed-route and paratransit operators. Additionally, all operators have Commercial Driver Licenses (CDL). There were also recent scheduling improvements that increased shared riding with less one-on-one taxi-type service, and this reduced paratransit vehicle-hours and subsequently increased passen- ger trips per vehicle-hour. Without reductions in related operating and overhead costs, such as for scheduling/dispatch and ADA eligibility certification, the factors worked together to increase the cost per vehicle-hour. For the predominately single-county service-area category, the two high datapoints—$77.90 and $78.05 per vehicle-hour—come from rural systems that are transit authorities, which have Performance Data from Representative Systems 51

a higher cost structure compared with other transit organizations (e.g., cities, counties, and non- profits). Additionally, both of the systems, while rural, are located near major metropolitan areas in higher-wage regions of the country. Operating Cost Per Vehicle-Mile Operating cost per vehicle-mile, similarly to operating cost per vehicle-hour, is a cost-efficiency measure. Data from the representative rural systems (see Figure 6-4) show that generally the operating cost per vehicle-mile is somewhat higher for the predominately municipal service-area type, with costs for systems in the predominately single-county service area trending lower, and, with the multi-county service-area systems, lower still. The Rural NTD data show somewhat higher costs for the predominately municipal category, while costs per vehicle-mile are more similar for the other two categories. Based on data from the representative systems, it seems the high datapoints for both the predominately municipal service-area systems ($5.84 and $4.65 per vehicle-mile) and the predominately single-county systems ($5.75, $5.60, and $4.47 per vehicle-mile) stem prima- rily from relatively high costs per vehicle-hour combined with slower average speeds. These systems operate in smaller service areas compared with other systems in their categories, with no or limited out-of-primary-service-area trips. This results in minimal or no higher-speed highway driving among these systems. 52 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance $74.04 $43.41* $29.31 $36.66* $40.09* $30.26* $35.23 $67.09* $34.66* $47.47* $45.16* $26.08 $48.82* $78.05* $33.70 $40.09* $32.72* $42.27 $32.47 $32.69 $57.65 $40.50 NTD AVG ($34.33) NTD AVG ($34.86) NTD AVG ($32.84) NTD MIN RANGE ($15.05) NTDMIN RANGE ($14.49) NTD MIN RANGE ($11.85) NTD MAX RANGE ($79.92) NTD MAX RANGE ($61.75) $0.00 $10.00 $20.00 $30.00 $40.00 $50.00 $60.00 $70.00 $80.00 O PE RA TI NG C O ST P ER V EH IC LE H O UR $77.90 NTD MAX RANGE ($77.27) Municipality County Multi County TYPE OF SERVICE AREA B31A Representative Data, 2007 Rural NTD data for DR only systems, 2007 Report Year * Adjusted Data Figure 6-3. Rural DRT systems: operating cost per vehicle-hour.

These data from the predominately municipal and predominately county rural systems can be contrasted, for example, to the low datapoint for the multi-county service-area systems at $1.16 per vehicle-mile. That particular rural system, operated by a non-profit agency in a lower- wage region part of the country, has a cost per vehicle-hour of less than $30 and, with trips throughout its five-county service area with many including highway travel, an average speed of 22 mph. This is significantly faster than the average speed of the rural systems with the high costs per vehicle-mile in the other two categories, which show average speeds of 11 to 13 mph. Operating Cost Per Passenger Trip This measure is considered a cost-effectiveness measure, combining elements of operating cost per vehicle-hour and passengers trips per vehicle-hour. The representative rural DRT system data, shown in Figure 6-5, show that those systems with low productivity combined with relatively high costs per hour have high costs per passenger trip, as would be expected. In the primarily municipal service-area category, for example, the system with an operating cost per passenger trip of $31.17 operates ADA paratransit service with a relatively low productivity and also has the highest cost per vehicle-hour of systems in that category. In the primarily single-county service area, the two high datapoints—$30.76 and $30.38 per passenger trip—come from systems with lower productivities compared with other systems in that category (and one of these systems is predominately an ADA paratransit service) and both are transit districts, with a higher cost structure compared with the other organization types. Performance Data from Representative Systems 53 $5.84 $2.73* $1.64 $2.67* $4.47 $1.40* $2.96 $5.75 $2.67*$2.57* $2.92* $1.16 $4.65* $2.83 $1.59 $3.02* $5.60* $2.19 $1.49 $1.48 $2.41* $1.73* $3.05NTD AVG ($3.12) NTD AVG ($2.09) NTD AVG ($2.13) NTD MIN RANGE ($1.31) NTD MIN RANGE ($0.75) NTD MIN RANGE ($0.79) NTD MAX RANGE ($5.93) NTD MAX RANGE ($6.97) $0.00 $1.00 $2.00 $3.00 $4.00 $5.00 $6.00 $7.00 $8.00 O PE RA TI NG C O ST P ER V EH IC LE M IL E NTD MAX RANGE ($7.64) Municipality County Multi-County TYPE OF SERVICE AREA B31A Representative Data, 2007 Rural NTD data for DR only systems, 2007 Report Year * Adjusted Data Figure 6-4. Rural DRT systems: operating cost per vehicle-mile.

While systems with low productivity and high costs per hour show high cost per passenger trip, the opposite is also shown: those DRT systems with high productivity and relatively lower costs per hour have the lower costs per passenger trip. In the primarily municipal service-area category, for example, the system with the $5.00 per passenger trip has both the highest produc- tivity and lowest cost per hour in the category. Interestingly, the operating cost per passenger trip for rural DRT systems in all three of the categories—those operating in primarily a single municipality, primarily in a single county, and in multi-county areas—appears to cluster in similar ranges, from about $11 to $15, although the primarily single-county systems show a somewhat broader grouping, from $11 to $18 per passenger trip. Performance Measures for Safety and Service Timeliness As noted above, there was very limited data from the representative rural DRT systems on the safety measure—Rural NTD safety incidents per 100,000 vehicle-miles—or for on-time per- formance. The majority of the participating DRT systems does not formally measure on-time performance and, thus, had no on-time data to provide for the project, but all monitor their safety performance, as would be expected. Only a few had the NTD safety incident data. When questioned about their safety experience, most of the managers indicated that they had no acci- dents and certainly none that reached the NTD reporting thresholds. 54 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance $31.17 $15.16* $15.37$12.37* $18.12 $14.13* $5.00 $30.76 $7.99* $14.89* $7.85* $8.27 $12.18* $30.38* $11.82$11.34* $10.77 $11.88 $7.63 $20.76 $16.51*$15.87$15.54 NTD AVG ($9.07) NTD AVG ($13.36) NTD AVG ($13.77) NTD MIN RANGE ($2.54) NTD MIN RANGE ($3.76) NTD MIN RANGE ($4.66) NTD MAX RANGE ($38.42) NTD MAX RANGE ($68.14) NTD MAX RANGE ($54.46) Municipality County Multi-County TYPE OF SERVICE AREA B31A Representative Data, 2007 Rural NTD data for DR only systems, 2007 Report Year * Adjusted Data $0.00 $10.00 $20.00 $30.00 $40.00 $50.00 $60.00 $70.00 O PE RA TI NG C O ST P ER P AS SE NG ER T RI P Figure 6-5. Rural DRT systems: operating cost per passenger trip.

The 2007 Rural NTD data can be examined to assess safety incidents for rural transit systems, providing a larger database than that of this project. Based on the DRT systems in the Rural NTD database that were assessed for performance statistics in this chapter (N = 479 systems) and based on the reported safety data, the average incidents per 100,000 miles is 0.21. This is based on a total of 387 reportable major incidents that were reported to NTD that year. The safety data can also be reviewed by the typology of rural DRT systems, with the average reportable safety incidents per 100,000 miles as shown in Table 6-2. When looking at actual reported numbers of incidents, the NTD data show that the vast majority of rural DRT systems have no reportable safety incidents. The average number of reportable major incidents per rural system in the municipal and single-county service-area categories is less than one; for systems in the multi-county category, there is an average of two incidents per system. While rural systems in the multi-county category had on average more reportable safety incidents, its average per 100,000 miles is less than that in the other two categories because of the significantly more miles traveled: multi-county DRT systems traveled more than 800,000 annual miles on average, compared with 115,000 for primarily municipal systems and 243,000 for primarily single-county systems. 6.3 Summary Rural DRT Performance Data The performance data from the representative rural DRT systems are summarized in Table 6-3, within the three categories of the typology. The important factors influencing the reported per- formance are also shown, as gleaned through the on-site visits and interviews with the DRT sys- tem managers. The various influencing factors, discussed in more detail in Chapter 5, include those that can be controlled or at least partially controlled by the transit system, while others are uncontrollable but must be acknowledged for their impact on performance. In addition, sum- mary performance data and selected operating characteristics are shown by system for the rep- resentative rural DRT systems in Appendix B. Passenger Trips Per Vehicle-Hour This measure of system productivity showed a large range: a high of 7.05 passenger trips per vehicle-hour to a low of 1.57. As was found in the urban phase of the research, those DRT sys- tems with smaller service areas tend to have higher productivities. Based on the research on rural DRT, the main controllable and partially controllable factors that influence higher productivity include the following: • A focus on routine group trips that are many-to-one and few-to-one. In particular, several of the higher productivity systems provide before- and after-school trips, with high group loads. • The ability to group trips for unaffiliated riders, particularly for longer-distance trips. Performance Data from Representative Systems 55 Source: 2007 Rural NTD database, rural transit systems operating only DRT, with outliers removed. Rural DRT Category Average Reportable Major Incidents Per 100,000 Miles Primarily municipal service area (102 systems) 0.14 Primarily single-county service area (262 systems) 0.28 Multi-county service area (115 systems) 0.09 Table 6-2. Rural NTD safety incident data, FY2007.

Representative Rural DRT System Passenger Trips per Vehicle-Hour Effectiveness Operating Cost per Vehicle-Hour Cost-Efficiency Operating Cost per Vehicle-Mile Cost-Efficiency Operating Cost per Passenger Trip Cost-Effectiveness Primarily-Single-Municipality Systems (5 systems) 2.38–7.05 $35.23–$74.04 $2.57–$5.84 $5.00–$31.17 Primarily-Single-County Systems (10 systems) 2.06–6.23 $32.47–$78.05 $1.49 -$5.75 $7.63 -$30.76 Multi-County Systems (7 systems) 1.57–4.34 $26.08–$42.27 $1.16–$2.67 $7.99–$20.76 Factors Influencing Performance Controllable/ Partially Controllable Group trips for agency clients Ability to group trips for unaffiliated riders, particularly for longer-distance trips Use of AVL Use of immediate response vs. advance reservation service Extent of long-distance, out- of-primary-service-area trips Characteristics of contracted service, in particular Medicaid Measures to reduce deadhead No-shows/late cancellations Administrative/overhead costs Costs for operator labor Administrative/overhead costs Costs for operator labor Administrative/overhead costs Costs for operator labor Group trips for human service agency clients and ability to group trips for unaffiliated riders Use of AVL Use of immediate response vs. advance reservation service Extent of long-distance, out- of-primary-service-area trips Measures to reduce deadhead No-shows/late cancellations Uncontrollable Size of service area Geographic constraints of service area Requirements for long- distance, out-of-service-area trips Type of ridership, i.e., ADA paratransit vs. non-ADA Type of organization, i.e., transit district, city/county, private non-profit Location in higher/lower labor-wage region of country Type of ridership, i.e., ADA paratransit vs. non-ADA Type of organization, i.e., transit district, city/county, private non-profit Size of service area and its influence on miles traveled Types of roadways traveled and operating speeds on those roadways Weather conditions that impact operating speeds Type of organization, i.e., transit district, city/county, private non-profit Size of service area and geographic constraints Requirements for long- distance, out-of-service-area trips Type of ridership, i.e., ADA paratransit vs. non-ADA Table 6-3. Summary performance data from representative rural DRT systems and influencing factors.

• Use of AVL in particular and of MDTs, which facilitate immediate response/same-day ser- vice. The AVL allows the dispatchers to “see” where the vehicles are, and they can send or re-direct the closest vehicle for an immediate pick-up. This ability to know vehicle loca- tions also gives the dispatchers improved control over the operators, allowing them to ensure that the operators are effectively deployed. • Use of immediate response or same-day scheduling, where dispatchers schedule trips in real-time, often providing very short response times, used in smaller service areas. • Efforts to reduce deadhead time and miles. • Low rate of no-shows and late cancellations. • Characteristics of contracted service, in particular Medicaid transportation. One of the rural systems that is a Medicaid transportation provider for a brokerage has discretion as to which long-distance Medicaid trips it accepts, according to the agreement—a factor that enables the system to ensure that the Medicaid trips it does provide can be grouped with other trips to maximize productivity. Uncontrollable factors include the following: • The size of the service area. • Geographic constraints of the service area, including mountains, rivers, and other bodies of water that impact the roadway network, resulting in longer trips to travel around these constraints. • Requirements for long-distance and out-of-service-area trips. • The type of ridership, with the few rural systems in the sample that serve predominately ADA paratransit trips showing lower productivities. Operating Cost Per Vehicle-Hour Performance on operating cost per vehicle-hour also showed a large range across the repre- sentative rural DRT systems contributing to the research: a low of $26 to a high of $78 per vehicle- hour. The systems with the lowest cost per vehicle-hour are non-profit organizations serving multi-county areas, which may not be fully allocating costs to the DRT system (e.g., all admin- istrative and overhead costs attributed to the DRT program). Those with the highest costs per hour are transit districts or governmental units with more extensive administrative structures operating in primarily single-county or municipal service areas. According to the research, the controllable and partially controllable factors include the following: • Costs for DRT administration, including the need for such functions as contract monitor- ing, certification, and Medicaid transportation administration. • Costs for operator labor. This is generally controllable for a transit system, but less control- lable if the DRT system must pay according to a pre-determined pay scale. Uncontrollable factors include the following: • The type of transit organization—that is, whether the DRT service is provided by a transit dis- trict; a governmental unit (city, county, or joint powers arrangement); or a private non-profit. • The location of the system in a higher- or lower-wage region of the country. Recent research on transit employee compensation in rural and small urban areas found that the region of the country is among the key variables correlated with wages (15). • The type of riders that are served. While there were only a few systems in the study sample that are predominately ADA paratransit, these had higher costs per hour, resulting from various factors including greater administrative needs related to eligibility/certification and longer service spans to match fixed-route service. Performance Data from Representative Systems 57

Operating Cost Per Vehicle-Mile The measure operating cost per vehicle-mile showed a low of $1.16 per vehicle-mile for a multi-county system operated by a non-profit in a lower-wage region of the country. The high was more than $5.00 per mile as seen at several systems that share few characteristics except for the fact that they have relatively high costs per hour compared with the other participating rural systems and traveled at slow average speeds compared with others in the sample. The slower speeds were impacted predominately by smaller service areas, with limited travel beyond the pri- mary service-area boundaries. Interestingly, one of the participating transit systems that appears as a high datapoint on cost per vehicle-hour for the county service-area category is a mid-point on the measure cost per vehicle-mile. This is because the system has significant travel outside its county borders, with those miles traveled on highways at highway speeds, with a systemwide average speed of 27 mph. Controllable and partially controllable factors influencing this measure include those same fac- tors related to operating costs, including costs for administration and operator labor. Operating costs may also be impacted if the rural system has significant long-distance trips to destinations beyond the primary service area, which will increase operating costs related to vehicle factors such as fuel and maintenance. These costs can be controlled to some extent by scheduling practices although these efforts may be limited if the trips are required for Medicaid purposes. There are a number of uncontrollable factors including the following: • The type of provider organization—that is, whether the system is a transit district, city/county, or private non-profit. • The size of service area, which influences miles traveled. • The type of roadways traveled (e.g., dirt roads, two-lane roads, or Interstate highways) and operating speeds on those roadways. • Weather conditions, which can impact operating speeds on roadways. Operating Cost Per Passenger Trip This measure is a composite of both operating cost per vehicle-hour and passenger trips per vehicle-hour. For the rural DRT systems included in the project, the cost per passenger trip ranges from $5, considered low for a DRT system, to a high shown by several systems of more than $30 per passenger trip—costs more similar to those at larger urban DRT systems. The data show a clustering within the range of $11 to $18 per passenger trip across the three service-area types. The factors influencing this measure are the same as those that impact cost per hour and passen- ger trips per hour. Important to recognize is the critical role that productivity plays with this mea- sure. One of the rural county-based systems with a relatively high cost per hour—more than $65 per vehicle-hour—has a cost per passenger trip of less than $11, due to the system’s high productivity. 58 Guidebook for Rural Demand-Response Transportation: Measuring, Assessing, and Improving Performance