Intercity Passenger Rail in the Context of Dynamic Travel Markets (2016)

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.

78 In the United States, Amtrak provides a range of services, including services that connect major metropolitan concentrations and services that consist of very long-distance routes with extremely sparse populations along the way. A third kind of service might be seen as something of a hybrid: moderate-volume corridors with major urban destinations at one end and communities of consid- erably less density at the other. Outside of its NEC service area, Amtrak operates service to Maine, New Hampshire, Vermont, and Western Massachusetts through a variety of specialized routes. 7.1 The University of Vermont Rural Intercity Transportation Survey This chapter concerns mode choice behavior for travel between rural communities and large metropolitan centers in the Northeast. This work uses data from a significant travel survey designed and undertaken by members of the research team in the spring of 2014 for the UVM TRC in collaboration with the NETI. The project was designed and managed by Professor Brian Lee of the University of Vermont, also a member of the NCRRP 03-02 research team. 7.1.1 Setting for the Rural Study In recent years, there have been several new research approaches developed for incorporat- ing attitudinal and behavioral components to the travel demand forecast and analysis process. Many of these research efforts have focused on intra-metropolitan trips and were very often undertaken on behalf of stakeholders in the metropolitan planning process. While this NCRRP project is primarily concerned with travel behavior between metropolitan centers, there is a need to better understand travel to/from less populated areas to large metropolitan ones (for the lack of a better term, in this chapter these trips are called “intercity” trips, even though they include travel originating from or traveling to towns and rural areas not commonly considered cities). In other studies, these intercity trips have been found to be typically of a distance greater than 100 miles, which in the Northeast usually means crossing at least one state border and often having several travel mode options to consider. Thus, Chapter 7 now examines trips for residents across the northernmost states of the Northeast—Maine, New Hampshire, Vermont, and Western Massachusetts—to improve basic understanding of the challenges in encouraging more sustain- able and efficient modes of intercity travel in a largely rural region. 7.1.2 Rural Intercity Survey Instrument The survey asked questions about actual trips taken; a hypothetical trip to New York City; and attitudes about traveling by automobile, intercity bus, and passenger rail. There were a total of C H A P T E R 7 The Role of Rail in a Rural Market

The Role of Rail in a Rural Market 79 98 questions plus a home zip code question that determined respondent eligibility for inclusion in the survey. The travel survey sampling protocol relied on commercial sample providers to recruit resi- dents from four New England states: Maine, New Hampshire, Vermont, and Massachusetts [outside of the Boston metropolitan area (Boston–Cambridge–Quincy Metropolitan Statistical Area)]. A total of 2,560 valid survey responses were collected. The survey was organized into four parts. Part 1 of the survey asked 13 questions about recent intercity travel and general travel preferences. Part 2 included 35 statements about intercity travel preferences, many regarding a specific utility or disutility pertaining to a certain mode. The content of these questions was strongly influenced by the TPB and were subject to several rounds of pre-testing. Many of these questions were used several months later in the NCRRP survey instrument. Part 3 presented a fictional scenario, in which the respondent has been asked to travel from his/her home to Manhattan, in New York City, for an important appointment during the following month and the respondent has decided to go. The respondent would stay one night at a hotel and travel alone. The host would pay for the hotel costs but not for travel. The respondent would be responsible for all costs of gas and parking or any fares. The respondent was asked to assume that, for one reason or another, she/he had already decided not to take any part of the trip by plane. She/he would then need to choose between taking the entire trip by car (whether or not it was his/her own vehicle) and taking at least part of the trip by intercity bus or train. All respondents were asked to select what mode(s) of transportation they thought were avail- able to them for this trip to New York City, how likely they would choose to take a bus or train for a trip like this to New York City, and whether learning that no Wi-Fi or electrical outlets were available on the bus or train would make them less likely to choose a bus or a train for this trip. The survey method also included an experimental design to test the effect of an advanced passenger information system, which was reported elsewhere. Part 4 included questions about what personal technology devices respondents own, and their demographics: age group, gender, level of education, and annual household income level. In addition to the information obtained from the survey data, several additional attributes were added, using available data and geographic information systems, for each zip code. These attributes included demographic information; land use; distances to destination cities; distances to the nearest urbanized areas within a metropolitan area; and distances to airports, rail stations, and bus stations of different sizes and types. 7.1.3 Intercity Travel Mode Distributions Part 1 of the survey asked respondents about recent intercity travel trips; this included a question about their recent trips from their home town to four major metropolitan centers in the Northeast: Boston, New York City, Philadelphia, and Washington, DC. The vast majority of intercity trips (N = 2,789) involved only one mode and a much smaller portion (N = 587) involved multiple modes. Not surprisingly, a much higher proportion of trips made by only one mode involved autos (73.6%) while the comparable number for trips with multiple modes is lower (45.7%). As such, the respondents are more likely to identify having used multiple modes if bus or rail was involved. Table 26 shows the travel mode distributions of these trips. As for comparisons between millennials and older age groups, there are differences for three distributions, as shown in Table 27. In general, older adults are more likely to use autos and air for their intercity trips, while younger adults are more likely to use bus and rail. This is similar to the

80 Intercity Passenger Rail in the Context of Dynamic Travel Markets differences between males and females (not shown). It is worth noting that differences between the age groups are bigger for the bus and rail modes than the differences between genders. 7.1.4 Components of the Rural Attitudinal Model The structural equation model (Figure 44) developed from the results of the rural surveying process focused on the Propensity to Take Rail or Bus as the outcome factor. Like the derivation of the shorter-term attitudes, the four latent factors for the four basic values (Figure 45) were derived by a process of confirmatory factor analysis, building upon the earlier results of the NCRRP Attitudinal Model. In this case, the optimal model fit was found to result when the observed variable “I like to be able to walk to a commercial or village center” was allowed to load on two separate latent factors. The content of the survey allowed the creation of an outcome factor (oval) for Propensity to Take Bus or Rail supported by four separate observed variables (rectangles) Auto Bus Rail Air Other Total Trips (N) Millennials 58.6% 12.9% 14.1% 12.2% 2.2% (100%) 618 Older group 66.3% 7.0% 11.1% 13.1% 2.5% (100%) 2,758 Table 27. Travel mode distributions for respondents’ recent intercity trip by age group. Figure 44. NCRRP Attitudinal Model, applied to a rural sample (2014, n  2,560). Auto Bus Rail Air Other Total Trips (N) All 64.8% 8.2% 11.7% 12.9% 2.4% (100%) 3,376 Table 26. Travel mode distributions for respondents’ recent intercity trips.

The Role of Rail in a Rural Market 81 Figure 45. Four latent factors for basic values in the NCRRP Attitudinal Model, applied to a rural sample. as shown in Figure 46. The shorter-term attitudes in the model reflect perceptions that the bus or rail option might be unsafe, inconvenient, and less stressful or more expensive than the auto trip. The full rural model includes the nine latent factors described here, and the addition of an observed variable for education level, and for residential density. 7.1.5 Rural Attitudinal Model Estimation The rural model was run on its full sample of 2,560 respondents, using the AMOS Version 22 software package, with maximum likelihood estimation. It had satisfactory levels of overall model fit, with a RMSEA of 0.044 (under 0.05 is considered good) and a comparative fit index of 0.94 and a Tucker-Lewis index of 0.92 (where values of over 0.90 are considered desirable for both). Details of the model output are included in the NCRRP Web-Only Document 2, Technical Appendix: Documentation for the Structural Equation Models. 7.1.6 Results The examination of the STE of key factors upon the outcome factor allows the early summary of the relative rank of factors in the explanation of Propensity to Take Bus or Rail. Table 28 shows that, for this rural sample looking at both bus and rail together, the perceived inconve- nience of public modes is the most powerful explanatory influence. This, of course, is consistent with the results from a very similar model of the NCRRP data set presented in Chapter 4 for rail only, and a model that will be presented in Chapter 8 for intercity bus only.

82 Intercity Passenger Rail in the Context of Dynamic Travel Markets Rank Factor* STE 1 Rail or Bus Trip Inconvenient 0.69 2 Rail or Bus Trip Unsafe 0.43 3 Rail or Bus Trip Expensive 0.31 4 Values Privacy in Travel 0.29 5 Rail or Bus Trip Less Stressful 0.22 6 Values Urbanism/Sociability 0.17 7 Values ICT 0.13 8 Educaon 0.13 9 Values Auto Orientaon 0.08 10 Density 0.04 *The four basic values are shown in italic bold; the four short- term atudes are shown in roman; and demographics are shown in italic. Table 28. Ranking of factors explaining rural propensity to choose bus or rail. Figure 46. Latent factors (ovals) for the four short-term attitudes and outcome factor, based on observed variables (rectangles).

The Role of Rail in a Rural Market 83 Impacted Factors Demographic Basic Longer Term Values Locaon Rural Bus or Rail Trip Perceived as … Educaon Privacy Auto Urban ICT Density Unsafe Inconvenient Less Stressful More Expensive Privacy 0.26 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Auto 0.09 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Urbanism 0.09 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 ICT 0.22 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Density 0.00 0.12 0.14 0.18 0.00 0.00 0.00 0.00 0.00 0.00 Unsafe 0.17 0.68 0.14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Inconvenient 0.10 0.33 0.05 0.21 0.00 0.00 0.49 0.00 0.00 0.28 Expensive 0.03 0.00 0.07 0.30 0.00 0.00 0.00 0.00 0.00 0.00 Less Stressful 0.10 0.41 0.08 0.02 0.00 0.00 0.61 0.36 0.00 0.10 Rural Propensity 0.13 0.29 0.08 0.17 0.13 0.04 0.43 0.69 0.22 0.31 Table 29. Impact of explanatory factors (columns) on impacted factors (rows). Exploring Privacy and Safety in the Rural Survey Beyond the expected role of inconvenience in the rural study, the next two ranking factors both reflect concern with safety and the unpleasantness of traveling with people one does not know. It should be pointed out that the rural survey was undertaken earlier than the NCRRP survey, and it focused on a trip to New York City only. Thus, there may be several explanatory phenomena at play at once: First, the rural population itself may fear the trip to the big city more than those originat- ing their trips from large metropolitan areas. Second, for the rural survey, the respondents were asked to think about New York City, not about Boston or Washington, DC, which might be seen as less intimidating destinations. For whatever reason, the calculated STE of “unsafe” is higher in the current survey than in the other two applications of the Attitudinal Model elsewhere in this report. Similarly, the factor for auto orientation was a less important factor in the rural study than it was for the more metropolitan NCRRP sample, while increasing level of auto orientation is modestly associated with lower levels of bus/rail use (-0.08). Summary for the Rural Sample Based on the relationships between factors revealed in Table 29, the following observations can be made (in this summary, the column values are shown in bold or italic bold font and row values are shown in roman font, reflecting the direction of STE implied in the table): • Higher levels of education are associated with lower need for privacy in travel (-0.26), less fear for personal safety on the trip (-0.17), and higher need for information technology (0.22). • Higher levels of need for privacy in travel are associated with much higher levels of fear for personal safety in the trip (0.68), with higher propensity to report that the trip is inconvenient (0.33), lower propensity to report that the trip is less stressful than the car trip (-0.41), and higher levels of residential density (0.12). • Higher levels of auto orientation are associated with lower levels of safety fear for the bus/rail trip (-0.14) and lower levels of residential density (-0.14). • People who value urbanism live in higher-density locations (0.18) and are less likely to per- ceive that the bus/rail trip is either expensive (-0.30) or inconvenient (-0.21). • People with the perception that the bus or rail trip is unsafe are much more likely to report that the trip is inconvenient (0.49) and much less likely to conclude that the trip less stressful (-0.61) than the car trip.

84 Intercity Passenger Rail in the Context of Dynamic Travel Markets • There is a positive association (0.28) between the bus/rail trip being seen as Expensive and seen as inconvenient. • There is a negative association (-0.36) between the trip being seen as Inconvenient and its being less stressful than the car. 7.2 Conclusion for the Rural Corridor Market The UVM TRC study of 2,560 residents of rural areas in the Northeast who have made trips to large cities concludes that the level of fear for personal safety for bus and rail trips is a significant factor in the explanation of the propensity to take bus or rail modes. This is, of course, in addition to the well-documented need for schedule convenience and quality of access to and from the terminals. Compared to the more metropolitan sample used in all other chapters of this report, basic values toward auto orientation seem to play somewhat less of a role in this rural context. Values held about urbanism/sociability seem to be somewhat more relevant to the modal decision than they were for the less rural sample. Concerns for privacy and safety seem to diminish with higher levels of education. And, again, the density of one’s residential location does not seem to be very much associated with the key determinant factors for choice of intercity mode.