5. Public Transit Trends and Performance

Mobility options available to North Dakotans, ranging from the private vehicle to local and regional public transportation, human service transportation, and commercial intercity air and surface transportation, were described in the previous chapter. The description of each type of service includes significant data on the quantity of service provided, the extent to which it is used, and some financial information on the providers. In addition, the appendices to this study include listings of services and providers known to exist in North Dakota at the time of this study. Of particular interest to this study is the data presented in Appendices F and G that includes profiles of the state and federally funded public transit providers and operating and financial data for 2004. This inventory of existing mobility options represents the best and most up-to-date information on transportation in North Dakota.

This chapter analyzes trends and performance information for a portion of the providers for which data is available - the fixed-route and demand-responsive transit systems supported in part by state and federal transit funds. Data on these providers is the most complete of all providers since the NDDOT and USDOT require regular reporting by grantees that can be used for the needed analysis. By analyzing these trends and making some observations about the performance of individual systems we can make judgments as to the adequacy of existing services and identify untapped capacity that can be used to provide additional mobility. More will be said later about the availability and accuracy of data. Recommendations for future data collection are included in Chapter 8.

5.1 Public Transit Trends

Public transit service has grown significantly in North Dakota over the past six years. Tables 5.1 - 5.3 summarize key operating and financial statistics for all public rural and urban transit systems for fiscal years 1999 through 2004. As can be seen, the total number of vehicle miles of service has grown 29.1 percent over these six years. Most of the growth occurred in the four urban systems (49 percent). Rural systems grew to a lesser extent (12.1 percent). Likewise, ridership has grown statewide but not by as much as vehicle miles. Overall, ridership increased by about 11 percent. Ridership in urban areas went up by about 23 percent while ridership in rural areas declined by 10 percent. The growth in vehicle miles of service without a comparable growth in ridership resulted in a decline in system efficiency for both urban and rural services as measured by passenger trips per vehicle mile.

These divergent trends between urban and rural areas in terms of ridership and service highlight several mobility challenges facing North Dakota and other states. First, in urban areas ridership did not grow as fast as the number of miles of service offered. The likely explanation of this phenomenon is that transit systems expanded their services into lower density, less productive areas to meet more mobility needs. However, the new services did not result in the same level of ridership intensity. Likewise, in rural areas the transit systems offered more service, but actually lost ridership over the six-year period. These statistics are consistent with the observations of the steering committee and the demographic data presented earlier in this report. In many rural areas, population is declining and the already sparsely populated areas are becoming even more difficult to serve efficiently with public transit. Further, to meet the mobility needs of those individuals with no other mobility options, transit providers are traveling more miles with fewer passengers to connect residents to health care and other high priority services.

Over the six-year period, significantly more resources were provided to public transit systems as measured by operating expenses and number of vehicles. The number of vehicles grew by nearly 22 percent with the greatest growth in rural areas. Operating expenses also increased. Some interpretation is needed to judge the growth in expenditures, especially for the urban areas. The 1999 expense data is skewed by a one-time increase in operating expenses reported by one of the urban providers (Bis-Man Transit).

A comparison of 2000 to 2004 data shows expenses increasing by 11 percent, about the same rate as the number of vehicle miles of service. Likewise, the trend in operating expense per vehicle mile must be adjusted to exclude 1999 urban data. Doing so suggests that operating expense per mile for urban systems remains about the same in 2004 as in 2000. This situation can perhaps be explained by economies of scale in operations in that systems can add services without greatly increasing fixed expenses so that the cost per unit of service remains relatively constant though some costs such as fuel and wages increase. Rural systems experienced an 8.6 percent increase in cost per mile, a very modest increase over a six-year period.

The last two columns of Tables 5.2 and 5.3 show the percentage of each financial and operating statistic accounted for by urban and rural systems in 1999 and 2004 and reveals the relative shifts in service accounted for by the two types of systems. Over the six years, a slightly higher percentage of total public transit system expenditures have occurred in rural areas, but urban areas account for a greater percentage of ridership and vehicle miles of service.

Several key conclusions can be reached from the data in these tables. First, public transit services in North Dakota have increased over the past six years in both urban and rural areas. Second, increased service, as measured by vehicle miles, has resulted in more ridership, but not in proportion to the increase in service or expenditures. Finally, the costs of providing service in North Dakota have increased, but only at a modest rate suggesting that transit systems are controlling costs and that service expansions require marginally less expenditure per mile of service than the average for a system.

5.2 Transit System Performance Evaluation

Performance evaluation is a key activity of both transit system management and policymaking and needs to be a part of a mobility plan such as the one presented here for North Dakota. Tracking key indicators allows both managers and policymakers to determine how well existing policies meet stated goals and provides insights into the likely success of future policies and service changes. Therefore, this section of the report summarizes key performance data for North Dakota's public transit providers.

Data availability and accuracy is a major hurdle to be overcome when appraising service performance. This is certainly true in the case of North Dakota's transportation services. As indicated in the previous section of this chapter, reliable data is not available for all transportation services, even those receiving public funding, because funding agencies have different reporting requirements and different definitions of terms and data elements These agencies have also placed relatively little emphasis on auditing and training to improve data accuracy. Therefore, as was the case in the first section of this chapter, the performance evaluation is limited to the approximately 50 providers that receive state and federal transit funds and report data to NDDOT.

Transit services can be evaluated on a number of attributes generally falling in the categories of efficiency and effectiveness. Efficiency indicators measure how well a transit system uses the inputs it receives (operating funds, vehicles, hours-of-service) and is measured with indicators such as expense per mile or hour of service. Effectiveness, on the other hand, is a measure of how well the transit system meets its goals in terms of quality of service, overall ridership, or ability to meet the needs of specific population groups. Another group of indicators especially relevant for this study measures the payoff from public funding and the equity of funding between providers and users as measured by subsidies per passenger trip and the share of the costs borne by the passenger.

Once calculated, performance indicators can be evaluated using one or more of three approaches. The first is time-series analysis where a performance indicator for a system for one time period is compared to the indicator for the same system over different time periods. This approach allows conclusions to be made in regard to the direction of change in performance, but says not about performance as compared to other systems or any industry standards that may exist. The other two approaches are peer comparisons, where an indicator for one system is compared to similar systems, and a comparison of a system's performance to industry standards or regulatory guidelines.

Time-series comparisons are the easiest to perform because there is no need to determine which systems are peers, and concern about comparability of data from year to year within a system should be less than comparability to other systems that may define data elements differently or collect the data in a different way. Time-series comparisons are best used by individual systems to manage and evaluate their own progress. Peer comparisons or comparisons to standards, while more difficult, are probably the most useful approaches in the context of this mobility study.

Unfortunately, data is not available to calculate many of the indicators important to public transportation. Two key missing data elements are vehicle hours of service and information on system reliability that might include on-time performance or breakdowns. In addition, there is some evidence that even given the best efforts of NDDOT staff to obtain accurate information from grantees, some of the data is incorrect. The most likely reason for this inaccuracy is a lack of clear definition of data elements and specific training on how to collected and report the information. Nevertheless, even given these limitations, useful performance information can be obtained for the data reported to the state.

Table 5.4 presents 2004 data on seven performance indicators reflecting three major attributes of the transit services and reports these indicators for each of the systems receiving public funding. A complete listing of the 2004 data is included as Appendix E. The indicators are divided into three categories including: efficiency, service fit, and subsidy policy. As indicated above, data availability greatly limited the indicators that could be presented and the ones listed are only a small sample of ones that would be desired for a complete analysis, yet they provide information that will be useful in formulating the North Dakota mobility policy.

5.2.1 Efficiency Indicators

Three efficiency measures were calculated using 2004 data. The first, expense per mile, measures the cost-efficiency of the system in producing a mile of public transit service. The average value for rural systems was $1.79 with a range in values of from 40 cents to more than $6 per mile. Urban systems ranged from $1.78 to $3.15/mile and are generally more costly to operate than rural systems because of the operating environment where congestion and frequent stops results in average speeds slower than in rural areas and the higher expenses associated with operating larger vehicles.

The range in expense per mile for rural systems points out the data accuracy problem mentioned previously. Properly determined operating expenses per vehicle mile for rural systems are most likely in the $1.25 - $2 per mile range, not 40 cents or $6.03 – the extremes reported in Table 5.4.

The most likely cause of the low cost per mile (Sitting Bull College) is an over-reporting of miles because this system also reports an average of 39,400 miles per vehicle, nearly 2.5 times the rural average. Likewise, the highest cost per mile (Kenmare Wheels) is likely due to an under reporting of miles because its miles per vehicle statistic is just over 3,000 miles, one-fifth the state average. NDDOT has made efforts to verify the accuracy of the data reported and the data appears to improve in accuracy over the past five years. However, more needs to be done to assure the usefulness of the information for grant administration purposes and for performance evaluations. Specific recommendations to improve the usefulness of the data are included in Chapter 8.

Operating expense per one-way passenger trip is the second efficiency statistic reported. Again, there are concerns about data accuracy because the range of values is significant – from just over $2 per trip to more than $50 per trip. It is difficult to determine if individual system values are reasonable simply by comparing them to averages because the type of service provided by the system will greatly influence a statistic such as the expense per passenger trip. For example, if a system provides many short local trips then the expense per passenger trip number will be significantly lower than a system where long trips to distant medical facilities for one or two persons is a major part of their service offering. However, the statewide averages will give some indication of performance.

The final efficiency indicator, vehicle miles per vehicle, indicates how much a vehicle is used and gives some indicator of the excess capacity that may exist within a system if the number of miles is relatively low. Again, there appear to be data collection problems for some systems where the values seem exceedingly high or low, but the data reported is very instructive. For example, consider if a vehicle is used eight hours a day, five days a week, and averages 16 miles per hour then a vehicle would travel 32,640 miles per year. The average for rural systems is just over one-third that level.

Recognizing that a system likely counts spare vehicles in its total count and also vehicles that are only used for peak or special occasions, the average number of miles for a high-use system would probably be less than the 32,640 figure. Systems that only report usage of 5,000 miles per vehicle undoubtedly have vehicles that are parked much of the time and therefore represent a resource for additional mobility. The rural average for this statistic is 12,628 miles per vehicle while the urban average is nearly double that amount except for the City of Minot that has a much lower average than the other urban systems because its service is primarily one of transporting school pupils.

In addition to supporting the observation that great excess capacity now exists in North Dakota's rural transit systems to provide more rides, the vehicle mile per vehicle statistic suggests that the state's investment in vehicles is being underutilized. Most of the vehicles used by rural systems have a federally determined useful life of four to five years and 100,000 miles. Obviously vehicles that are used 5,000 or even 15,000 miles per year will deteriorate from age before they are driven their useful miles.

One of the benefits of implementing the recommendations of the recent NDDOT - SURTC coordination study would be to identify customers that could use rides during times when vehicles are idle. This effort would increase both vehicle utilization and personal mobility across the state.

5.2.2 Service Fit Indicators

Table 5.4 also summarizes two "service fit" indicators. These indicators attempt to report on how well the service (vehicle miles or vehicles) is used by riders (one-way passenger trips). The passenger trips/vehicle statistic is particularly instructive. First, as a point of comparison, consider that if a system transports one person per vehicle hour, a low productivity figure, then a system average of 2,080 passengers per vehicle per year would result. The average for rural systems is about three times this figure suggesting either more hours of service than the five-day-a-week, eight-hour day, or higher ridership productivity in terms of passengers per hour.

Again, the type of service provided and the environment within which the systems operate will greatly influence the passengers per vehicle statistic. Nevertheless, it would seem that there are significant opportunities to improve system productivity and provide more rides if systems that now average less than say 2,000 passengers per vehicle (again, less than one passenger trip per vehicle hour) could increase utilization even up to the current state average.

5.2.3 Subsidy Policy Indicators

The final two indicators reported in Table 5.4 measure two aspects of transit system financial support; namely the percentage of expenses covered by fares and the subsidy per rider. The first statistic reflects transit system policy as to the portion of the operating cost that is expected to be borne by the passenger (or third-party funding agency). The other indicator, subsidy per trip, indicates the extent to which state, federal, and local government sources are required to pay for individual trips.

As with the other measures, care must be taken in interpreting the results given likely problems with the data. For example, systems vary in what they report as fare income. Sometimes payments from human service agencies are counted as fares from their client passengers, while other times it is counted as subsidy. With that said, the data for 2004 indicates that riders paid anywhere from none of the costs to nearly two-thirds the cost of the trips with about one-sixth being the average for both rural and urban areas. The subsidy per trip, a statistic closely correlated with the fare percentage indicator varies from less than $1 per trip to nearly $50 per trip with an average of $4.27 in rural areas to less than $2.50 per trip in urban areas. Again, the type of service needs to be considered when evaluating the appropriateness of the differences in this indicator.

5.2.4 Conclusions on Performance Indicators

This section of the chapter reports on several performance measures for North Dakota's public transit providers. Because of problems with data availability and lack of confidence in the accuracy of some data, few strong conclusions can be reached from this analysis. The most important conclusion and recommendation resulting from this analysis is that the NDDOT should consider adopting an expanded data-reporting requirement and providing the necessary guidance and training to help systems not only report more accurate information, but also develop their own evaluation processes to help them improve their operations. SURTC, either directly or through the RTAP program, can help with this effort. Doing so should help generate more reliable data that could be used to develop peer comparisons and guide state policy. Similar efforts should be undertaken by other agencies that support transit services with public funds.

In addition to the observations on the data, several other conclusions can be reached from this analysis. For example, the performance of North Dakota's public transit systems varies significantly and does so to an extent that cannot just be explained by different operating environments. Additionally, the vehicle miles per vehicle and passenger trips per vehicle statistics suggest that many systems are underutilizing the capacity they now have. That capacity could be directed toward more mobility for residents of their service areas. Also, many of the current grantee providers offer "niche" services to selected geographic areas or client groups. If a more coordinated and comprehensive approach to public transportation were applied in these areas, the same vehicles could be more fully used to meet additional mobility needs. Additional funds to pay for fuel and driver time would be required, but many of the resources for improved mobility are already in place.