Air Demand in a Dynamic Competitive Context with the Automobile (2019)

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.

91 Good News and Bad News It is impossible to predict the future, but, hopefully, the application of scenario analysis can help in understanding the range of possible futures for the competition between the modes of airplane and automobile for a total trip and for access to longer-distance air service. From a series of purposefully dramatic scenarios for the future, notes of optimism and notes of pessimism can be found. Optimistic? Seen from the point of view of an airport manager, the “optimistic” message is that, if for one reason or another, a new set of non-stop short- and medium-distance flights could be offered from non-hub airports, there would be a market for travelers to purchase those flights. An ambitious model of all long-distance travel in the lower 48 states shows significant market support for such new services in the scenario analysis. This, in turn, supports the logic of developing new kinds of aircraft designed to travel relatively short distances with lowered operating costs. Such an optimistic scenario produced increases in use of smaller airports of 30% to 50%, with the highest impacts on shorter trip distances and the smallest airports. Pessimistic? Also, for the airport manager, the “pessimistic” message is that massive erosion in shorter- distance airplane tripmaking has already occurred throughout the country, resulting in a 30% decrease in seat-miles per capita offered in short-distance segments and a 30% decrease in trips taken with origin-to-destination distances of 500 miles or less. Beyond these empirical observations, the scenario analysis clearly sends out a warning that this erosion is by no means completed; with the incremental improvement in the level of comfort associated with the automobile, there could be an incremental shift toward the automobile in the short-distance air market. Increase in Use of Automobile for the Mid-Distance Trip When Did It Start Happening? The research team has concluded that the roles of airline services and the roles of the private automobile have changed over the past two decades, particularly for trips of less than 1,000 miles in length. This market pattern is associated with drops in total traffic at smaller airports and C H A P T E R 7 Conclusions and Further Research

92 Air Demand in a Dynamic Competitive Context with the Automobile gains at larger airports. Arguably, this restructuring of American airline services away from the shorter-distance service commenced in the fall of 2001. Using 2000 as a base year, the fall in airplane use for the shorter trip has been dramatic. As analyzed by the regional jet manufacturer Bombardier, air trips for the market of less than 500 miles in distance has fallen by about 30% up to the present time. That analysis noted about a 50% drop in air trips of less than 200 miles, a 39% drop in air trips between 200 and 300 miles, and a 14% drop in air trips between 300 and 400 miles in distance. Using Dallas to Houston as a case study of what the author calls “the tenuous balance of the drive or fly decision,” the analysis traced the automobile access time to the origin airport and egress time from the destination airport. To that market analyst, the addition of hours of “hassle time” associated with earlier arrivals at the airport of origin tipped the scale over to the automobile competitor. Air volumes today are only 40% of their year 2000 level in the case study corridor. The main conclusion of the article is that “One thing is abundantly clear: something fundamental changed shortly after 2000 in the short-haul market” (Miller 2017). Shorter Flights Become Less Viable Airlines have found lower levels of revenue in shorter flight segments, which they can usually only justify when these segments serve as feeders to more profitable long-distance domestic and international flights. In the vulnerable distance-band categories, all of this results in a lowering of the number of trips taken entirely by airplane and the number of air segments used to gain access to the longest distance segment of trips. At the same time, growth in the longer- distance air markets has resulted in a level of overall enplanement per capita that is higher today than ever before, with roughly a 15% increase in the rate of trips per capita over the base year of 1995. As shown in Figure 2-4 in Chapter 2, the length of the airline trip segment has increased in a nearly linear pattern over time, reflecting the lower use of short flights and higher use of longer flights. The Automobile Share of the Mid-Distance Market Soars A long process of data analysis, including most recently the 2017 ACRP Project 03-40 survey, shows the growth of the automobile share of the long-distance market. As Figure 2-8 in Chapter 2 shows, an increase in automobile mode share since 1995 is most dramatic in the distance band between 700 and 1,500 miles. As expected, in the final simulation dataset, the scale of auto mobile share decreases in a relatively linear pattern as the distance of the trip increases. Automobile trips of less than 1,000 miles seem to be associated with the fall of the airplane mode, with less variation for the longer trips. Using the more commonly accepted definition of the long-distance trip, this analysis concluded that the role of the automobile trip of more than 100 miles seems to be changing. In the 1995 base year data, those trips represented 17% of yearly VMT per capita; today they may constitute 29% of VMT per capita. Selection of Departure Airports Is Changing Change in Automobile Use to Distant Airports The pattern of increased reliance on automobiles in the long-distance trip also includes the increased use of the automobile to access the airport of choice. This research has revealed the basic logic behind leakage. This research has established that, if total travel times are the same, most travelers would prefer to spend additional time in an automobile to gain access to a

Conclusions and Further Research 93 better-served airport rather than spend that same time in a feeder flight and its associated times. The results of the project’s Hybrid Choice model show that business travelers would be willing to spend $17 to save an hour of driving time but would be willing to spend $36 to save an hour of flying time. Leisure travelers would be willing to spend $10 to save an hour of driving time and $26 to save an hour of flight time. Thus, the traveler would drive an extra hour to save one-half hour of flight time. Proximity Is Not the Key Attribute for Selecting Airport of Departure In short, the research suggests that the question of the “closest” airport is not the key attribute when selecting the airport of departure. The question seems to turn on what airport will provide the best flight options, with the key desire being to minimize time in airplanes. In short, the customer may want to “leak” in spite of whatever local sense of loyalty might be in play. Given that circumstance, the research team estimated that between 16% and 32% of proxi- mate residents would not choose the closest airport. Broken down into non-stop and connecting travel, the median share of passengers captured at local airports in 2015 is 83% for direct itin- eraries and 42% for connecting itineraries. As reported in Chapter 4, this means that regaining passengers with the implementation of non-stop flights is more likely than with implementation of non-direct connections. This, in turn, provides support for the development of aircraft that could provide short-distance direct services, particularly for smaller airports. What’s in the Future? The major conclusion from this research is that the competition between the modes of auto- mobile and airplane for trips of short-to-moderate distance is not over. The already observed increases in the mode share of the car may be just the tip of the iceberg. The Fully Developed Autonomous Vehicle Which Trips Are Most Vulnerable to Change in Mode? Table 7-1, adapted from Figure 1-2 in Chapter 1, reveals the starkness of the implications of the possible development of autonomous automobiles. Moreover, as noted throughout this report, rates of loss for trips of less than 1,000 miles are greater than rates of loss for trips of more than 1,000 miles. The Scenario Testing Model process was not designed to forecast the future; it was designed to help determine a sense of scale for what might happen. Nonetheless, from data in Table 7-1 the implications for these mid-range air trips should be clear and a cause for concern in the aviation industry. Incremental Improvements to the Car There are several components to the threat posed by the car of the future. In addition to the examination of the autonomous car in five overarching scenarios, separate component elements of a future car scenario were examined. The concept of the automobile trip being less stressful than today was tested, resulting in decreases in airplane use ranging from 5% for the 800- to 1,000-mile trip distance band to 10% for the shortest trips. It is important to note that these sensitivities did not result from the case of full adoption of autonomous vehicles, which was shown in Figure 1-2; lowered stress of the automobile trip might come from more music, more entertainment, more perception of being connected, or even a feeling of greater safety associated with modern connected-vehicle technology.

94 Air Demand in a Dynamic Competitive Context with the Automobile What Has Been Learned About the Choice Between the Car and the Plane? Does Market Behavior Make Sense? This research has concluded that the traveling public acts rather logically when faced with trip options. Over long distances, the price per mile of air travel is lower than the price per mile by car. With an average trip party size, the plane becomes cheaper at about 1,500 miles. For trips of more than 1,500 miles, some 70% of trips are made by air. By contrast, while the plane is clearly cheaper for longer distances, the price per mile for a trip of less than 700 miles is more than twice the price per mile of trips of more than 700 miles. As a result, for trips of less than 700 miles, some 70% of trips are made by car. In terms of overall predilection toward a given mode, the research team’s market segmen- tation revealed that about 52% of the population belongs to market segments with positive attitudes toward the plane, with only 24% in market segments with pro-automobile attitudes. Logically enough, the segment most positive about the air trip is dominated by business travelers. In short, the study concludes that, while most passengers would prefer the plane, their choice of mode accurately reflects reasonable considerations of cost. What Drives the Modal Decision? This research has addressed the question of the competition between the car and plane by using a wide variety of research approaches, ranging from attitudinal (and subjective) to pre- dictive (and quantitative), but there are common threads running through the various types of analyses. The research has reinforced the concept that much of the decision to choose a long- distance mode is emotional and very much associated with attitudes and values. When a traveler with a positive inclination toward the plane faces the reality of this mode’s additional costs, attitudes toward making a car trip instead come into play. At that point, the mode choice is influenced by a somewhat polarized set of attitudes about the unpleasantness of the automobile trip, where one group (including the young) finds the automobile trip distaste- ful, and a second group (including those over 35) finds it more acceptable. In sum, the SEM process suggests that the choice of the long-distance mode is the result of a trade-off in the mind of the traveler, in which the price of the air trip is compared with perceived level of unpleasantness DISTANCE (MILES) DECREASE IN NUMBER OF AIR TRIPS PERCENT LOSS IN DISTANCE BAND 200–400 8.2 million −18% 400–600 8.4 million −15% 600–800 13.2 million −24% 800–1,000 11.1 million −22% 1,000–1,200 9.3 million −18% 1,200–1,400 6.2 million −16% 1,400–1,600 2.7 million −13% Adapted from Figure 1-2. Table 7-1. Simulated decrease in air trips as a result of autonomous vehicles.

Conclusions and Further Research 95 of the long-distance automobile trip. Compared with the explanatory power of these factors, other considerations modeled have been revealed to be relatively unimportant. The Positive Outlook Table 7-2 shows the simulated increase in air trips as a result of new short-distance flights, the most positive scenario for smaller airports (Scenario 3). In this scenario, the increase in absolute number of additional flyers would be spread rather evenly over all the distance bands under 1,000 miles, with the highest increase from the shortest trip distances. The modeling process in Scenarios 2 and 3 answered one question: assuming that the airlines elected to operate more direct, non-hub-transfer, services, would there be a market to pur- chase these services. Although the answer to this carefully phrased question is “yes,” there is no assumption by the research team that the airlines indeed would choose to offer this new set of short-to-medium distance services. This is a matter for further research. Further Research This report has presented the first results from a new modeling procedure that predicts trip- making behavior for all long-distance trips in the lower 48 states, in this case applied to the competition between airplane and automobile modes. This report has presented early interpre- tations of possible futures, each of which deserves more attention on its own merit. Similarly, each of the separate component factors of mode choice for long-distance trips were reviewed only briefly in this document and could be explored in more detail in further research. What About the Other Modes? This report has noted that for the vast majority of American city-pair corridors, the effective market competition is between airplane and automobile. However, in the Northeast and a few very select city-pair markets, rail and bus together represent meaningful alternative modes of intercity travel. Further development of the Scenario Testing Model used in this research should re-incorporate those trips into the analysis process for those corridors where inclusion would improve the accuracy of the modeling process. Although the subject of rail’s competition with the airplane has been explored in ACRP Report 118: Integrating Aviation and Passenger Rail DISTANCE (MILES) INCREASE IN NUMBER OF AIR TRIPS PERCENT INCREASE WITHIN DISTANCE BAND 200–400 11.7 million 25.4% 400–600 11.0 million 19.7% 600–800 9.7 million 17.7% 800–1,000 9.2 million 18.6% 1,000–1,200 7.4 million 14.7% 1,200–1,400 5.0 million 12.7% 1,400–1,600 2.0 million 9.9% Table 7-2. Simulated increase in air trips as a result of new short-distance flights.

96 Air Demand in a Dynamic Competitive Context with the Automobile Planning (Coogan et al. 2015) and NCRRP Report 4: Intercity Passenger Rail in the Context of Dynamic Travel Markets (RSG et al. 2016), widening the modal coverage of the Scenario Testing Model would have obvious benefits. Better Understanding of the Attributes of Automobile Travel In many parts of this research, policy attention was focused on the implications of a full and complete program of autonomous vehicles replacing the car as we know it. Such a mature program would offer an alternative to long-distance travel of all kinds. However, the next few decades will clearly see incremental improvements to the automobile, and the impacts will be more selective and more subtle. On the one hand, in upcoming decades there will be more hedonic improvements to the pleasure of the automobile trip, through more entertainment and quite possibly through measures to increase the ability to be connected to the Internet. Although it is difficult to quantify, the near future may bring improvements to the safety of travel through connected- vehicle technology, with more safety associated with spacing of, and interaction with, other vehicles on the roadway. This might be associated with a decreased level of stress associated with the driving experience. On the other hand, in upcoming decades there will almost inevitably be increased levels of con gestion in the portion of a long-distance automobile trip that occurs within urban/ metropolitan areas. The automobile of the near future may offer more pleasurable attributes while still spending more time stuck in congestion. How this trade-off will affect the competition of car trips with air trips is beyond the present ability to predict. Better Understanding of Why Airlines Do and Do Not Add New Service As emphasized in this report, Scenarios 2 and 3 both assumed that, for reasons established exogenously, the airlines would add more direct service between airports that are not major hubs. At the present time, both NASA, on the public-sector side, and private entrepreneurs are exploring the concept of smaller airplanes providing this kind of service. To the extent that such planes could mimic the quality of travel experience offered by larger planes today, this hypothetical service makes sense. However, if these smaller planes lack the ability to undertake instrument-based landing in lower visibility or the ability to fly at altitudes “above the weather,” their future role in replacing the automobile for short distances is highly questionable. This report, as noted earlier, has addressed the question of the market scale for such services, but not their economic viability. Separate from the high-visibility issue of new aircraft technology is the less dramatic question of airline-based decisions about the amount of service to provide between a smaller airport and even the dominant hub of that airline. Quite simply, the more the airport manager can understand the underlying economics of service provision to the smaller airports, the better that manager can be at creating a strategy to maintain and expand those services. The modeling pro- cesses developed in this research might be modifiable to address such near-term demand issues. Better Understanding of the Choice of the Airport of Arrival Much of the research team’s analysis has focused on the way in which a traveler selects the air- port of departure. This was not meant to imply that the selection of the airport at the non-home

Conclusions and Further Research 97 end of the trip is a symmetrical, mirror image of the selection at the home end of the trip. At the home end of the trip, most Americans have access to an automobile, whether by owning, sharing, or borrowing it. At the home end of the trip, the traveler is aware of all the details of the roadway network and quite capable of choosing the best automobile trip options. The opposite is true at the non-home end of the trip. An automobile must be rented, or a taxi/transportation network company must be obtained at some cost. The traveler may or may not know the roadway network and may not want to add several hours of driving after a long airplane trip. The research team proposes that additional research be undertaken to explore the hypothesis that the local airport, close to the center of the destination city, may be more relevant to the arriving visiting traveler than to the departing local traveler. Gaining supporting data for this hypothesis might be difficult, in that neither the BTS/FAA T-100 database, nor the DB1B database really focus on whether the tripmakers are local (going out) or non-local (coming in). The implications for the airport manager are critical. Finding that the true strength of the small airport market is the arriving passengers, rather than departing, would result in a signifi- cantly different marketing strategy than the opposite case. The Need for Better Data Collection for the Automobile Trips Building on significant advances made by FHWA, ACRP Project 03-40 has developed a new approach to understanding national long-distance travel patterns. It addresses the issue of modal competition directly. It is critically important for the future of this effort that coordination be maintained with ongoing efforts at FHWA to improve the quality of data used to describe origin-to-destination automobile tripmaking. Future improvements to a multimodal approach (in this case automobile vs. airplane) will require a significantly improved understanding of long-distance automobile patterns. Ongoing efforts at FHWA concerning the future of the NHTS, and the “NextGen” for automobile trip data collection must be care- fully coordinated with further improvements in understanding multimodal and intermodal passenger travel behavior.