Integrating Aviation and Passenger Rail Planning (2015)

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.

54 C H A P T E R 4 Introduction and Structure In Chapter 2, the Research Team examined the use of rail services in Europe to serve in a feeder function to give access to longer distance aviation flights, which it labeled a “complemen- tary” function for rail. Chapter 4 explores European experiences where rail operates in competition with air in the provision of corridor services dominated by center-city-to-center-city higher speed services. Consistent with the approach in Chap- ter 2, this project provides a significant update of the estimated total diversions from air to rail in the major European projects documented to date. Major Themes Emerging from the Research: Rail in a Competitive Role Chapter 4 focuses on the role of rail as a component of the intermodal system, this time making its contribution by lowering the number of passengers that need to be accom- modated at airports, allowing the managers of the aviation system to alter the supply of aviation services. The Research Team’s North American interviews consis- tently demonstrated that the rail system makes its most meaningful contribution to the overall intermodal system when it is attractive to long-distance city center-to-city cen- ter markets, which are influenced by terminal-to-terminal travel times. Chapter 3 presented a “business case” for the rail operator, which often tends to argue against the addi- tion of airport stops for higher speed intercity rail services. Chapter 2 demonstrated that strong markets exist for rail to carry people over long distances to airport rail stations in Frankfurt, Paris, and Amsterdam; however, such specialty markets are hard to initiate and hard to keep. This Chapter focuses on how HSR has already been a success in lessen- ing the volume of passengers at busy airports in Europe by focusing on providing high-quality center-city-to-center- city services. The Chapter reviews eight corridors in Europe and docu- ments the manner in which new HSR services influence the aviation market, noting the change in air volumes on a month- by-month basis. Rail Diversion from Air in Europe This section summarizes the results of reported diversions from air to rail as the result of improvements to rail travel time in eight European corridors, as shown in Figure 4-1. HSR has diverted an estimated three million air passengers from the several corridors connecting Paris with the rest of France, and Europe. A diversion of one million passengers from Paris to Marseilles, one million between Paris and Strasbourg, 0.7 mil- lion air passengers from Paris to London, and about 0.1 million between Paris and Lyon is observed. In addition, lesser amounts of diversion have occurred in other corridors; a diversion rate of 20% has been estimated between Paris and Bordeaux, with varying rates associated with early improvements to Paris– Amsterdam to 4.5-hour travel times (Wardman et al. 2002). In the newly developing system in Spain, the diversions have been similar, with rail now carrying about 46% of the 5 mil- lion passenger market between Madrid and Barcelona, repre- senting a shift of 1.8 million passengers from air to rail. For Madrid–Seville, air played a relatively small role in the corridor before the rail investment, with an estimated 0.25 million now diverted from the air to the rail. An estimated 0.35 million trav- elers have been diverted from air to rail on the Madrid–Malaga corridor (anna.aero 2010). In a rail corridor which had yet to open, it was reported that air officials believe that the Madrid– Valencia corridor will lose 0.67 million air passengers to rail (anna.aero 2010). Thus, both France and Spain are looking at levels of diver- sion in the range of 3 million riders each from air to rail. This scale of long-distance rail riders diverted from air is distinctly smaller than the 23 million European long-distance rail riders who are accessing a major airport, as documented in Chapter 2. Rail in a Competitive Role: Diversion from Air in Europe

55 This contrasts with the prevailing view accepted in the United States: that competitive rail is the most important concern. The corridors reviewed include: 1. Paris–London 2. Paris–Lyon–Marseille 3. Paris–Strasbourg 4. Madrid–Séville–Malaga 5. Madrid–Barcelona 6. London–Manchester 7. London–Glasgow 8. Amsterdam’s Corridors (predicted diversions) Relationship Between In-vehicle Rail Time and Share to Rail The relationship between the in-vehicle travel time of the rail service and the rail portion of the air plus rail market is well documented in the literature. Figure 4-2 documents all mode share data collected or updated for this research, including different mode shares over time in a given corridor. Because of possible differences in definitions over many sepa- rate sources, the data could be seen as illustrative rather than definitive. However, this figure attempts to show only those mode share values that represent the corridor rail passenger volume compared with the corridor air passenger volumes 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Ma dri d– Ba rce lon a Pa ris –M ars eil le Pa ris –S tra sb ou rg Lo nd on –M an ch est er Pa ris –L on do n Ma dri d– Se vil le– Ma lag a Ma dri d S ev ille Pa ris –L yo n Figure 4-1. Annual European rail riders diverted from air by city pair, in millions. Source: Case studies in this chapter. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 50 100 150 200 250 300 350 400 Ra il sh ar e of A ir Pl us R ai l M ar ke t Minutes of In-vehicle Travel European Rail Share by Travel Time 3 1/2 Hour Rail Travel Time Figure 4-2. European rail mode share by travel time.

56 Diversions in Corridors: France Figure 4-3 shows the change in mode share for two major rail corridors and single observations for service between Paris and several major cities. When the best rail travel times between Paris and London went from 6 hours to 2.5 hours, the mode share rose from 14% to 60%; further improve- ments in travel time increased mode share to 75%. When travel times between Paris and Marseille went (in total) from 4 hours 40 minutes to 3 hours 6 minutes, mode share rose from 22% to 69%. The graph shows that when the value on the X-axis shifts to the left, the value on the Y-axis will increase; in other words, as rail travel times improve, their mode share will increase. Corridor #1 Paris–London The Setting The present service between Paris and London takes between 2 hours 15 minutes and 2 hours 45 minutes based on the Spring 2011 schedule. London–Paris travel times shot from between 6 and 8 hours (depending on channel crossing vessel) to 3 hours, and mode share jumped from 14% to 54% after the first full year of operation. Annual air passengers have fallen from roughly 3.6 million (averaged) in the early 1990s to roughly 2.9 million in 2009, as shown in Figure 4-4 as defined to include those air passengers making connect- ing flights. Later, a smaller number of European mode share statistics that only focus on O-D markets is presented, which can then be compared with similarly defined American mode share data. In either case, it is notable that in this chart, the only corridors with rail mode shares higher than 50% are those whose rail in-vehicle travel times are lower than 3.5 hours, expressed on the X-axis of the graph as 210 minutes. The Research Team’s database shows only two situations in which a rail travel time of less than 3.5 hours failed to produce a mode share above 50% (see Table 4-1). Historically, during the incremental development of the London-Brussels service, there was a period where a travel time of 2 hours 45 min- utes was associated with only a 44% share of the air plus rail market—including all connecting air passengers in the cal- culations. Since the more recent improvement of rail speeds in the United Kingdom, the most recent reported mode share estimate is 75% to rail, by the same definitions. More impor- tantly, between Madrid and Barcelona, the mode share of rail to the air plus rail market is still under 50% when the con- necting air passengers are included in the calculation. As dis- cussed in Chapter 5 (structured to allow direct comparison with the best North American data), the rail share is above 50% when the connecting air passengers are excluded from the calculation. Minutes Share Minutes Share FRA-Cologne, 2006 (1) 70 99% Madrid- Barcelona, 2010 (8) 160 46% Paris-Brussels, 2008 (3) 85 99% Paris-Amsterdam, 2003 (7) 240 46% Paris-Lyon, 2003 (1) 120 88% London- Brussels, 2002 (3) 165 44% Madrid-Seville, 2006 (1) 150 84% Rome-Milan, 2005 (2) 270 38% Madrid-Seville, 1994 (4) 150 80% London-Edinburgh, 1999 (1) 260 29% London-Manchester, 2009 (9) 127 80% Madrid-Malaga, 2005 (10) 280 28% Paris-London, 2009 (3) 150 76% London-Edinburgh, 2004 (1) 260 18% London-Brussels, 2009 (3) 125 75% Paris-Marseille, 2000 (3) 300 22% Rome-Bologna, 2003 (7) 169 71% Madrid-Seville, 1991 (4) 390 28% Paris-Bordeaux, 2008 (5) 125 70% London-Glasgow, 2010 (9) 271 20% London-Manchester, 2008 (9) 135 70% London-Cologne, 2011 (6) 258 13% Paris-Marseille, 2006 (3) 180 69% Madrid-Barcelona, 2005 (1) 285 12% Madrid-Malaga, 2008 (5) 150 63% London-Amsterdam, 2011 (6) 257 11% Paris-London, 2003 (3) 183 61% Madrid-Barcelona, 2002 (1) 370 8% London-Manchester, 2004 (1) 150 60% London-Glasgow, 2004 (9) 310 6% Stockholm-Goteborg, 2003 (7) 181 59% London-Lyon, 2011 (6) 340 5% London-Frankfurt, 2011 (6) 343 4% Sources: (1) SDG, 2006a; (2) SDG 2006; (3) SKM 2010; (4) Wardman, et al 2002; (5) IARO 2009 (6) Atkins 2011; (7) M3 Systems 2004; (8) Spanish Civil Authority, 2010; (9) Virgin Trains 2011. Table 4-1. European rail mode share, by rail travel time.

57 travelers today (SDG 2006a). From examining Figure 4-4, the reader can observe the initial jump in rail mode share upon completion and stabilization of the service (to 1997), followed by a stable period (to 2004). This stable period was followed by market improvements, with improvements in in-vehicle travel time on the United Kingdom end of the corridor. What Happened to the Flights? Reportedly, the number of flights between London and Paris fell by about 40% with the opening and stabilization (anna.aero 2010). This occurred after the opening of the Eurostar service through the channel tunnel in November of 1994, which included new rights-of-way in France. This was followed by travel time improvements on the United King- dom portion of the trip. This loss of 0.7 million air passen- gers represents a decrease of about 20% of all air passenger traffic, which is generally consistent over sources (Wardman et al. 2002). Eurostar—as a portion of Eurostar plus all air in the corridor—has a mode share of about 75%. SDG has esti- mated that rail captured more than 80% of the O-D market in 2005, which implies that it is attracting about 85% of such 0 10 20 30 40 50 60 70 80 90 100 60 120 180 240 300 360 Ra il Sh ar e of Ai rP lu sR ai l Rail In Vehicle Travel Time, in Minutes Paris-London Paris-Marseilles Lyon Bordeaux Amsterdam Brussels Paris-Marseille, 2006 Paris Marseille, 2001` Paris London, Historic Paris London, 2009 Figure 4-3. Change in mode share by travel time (France). Figure 4-4. Yearly change in air traffic between London and Paris. Source: anna.aero, 2010.

58 stitution rate from planes to trains for the O-D market (Wardman et al. 2002). What Happened to the Flights? Air France operates approximately 25 flights between Paris and Marseilles, down from 45 flights at its peak before the rail improvements (a decrease of 45%) (SDG 2006). A low-cost carrier entered the market, but left in 2001. In the report of the International Air/Rail Organization titled HSR Competition, its Director General, Andrew Sharp, reports that the Marseille Airport had lost 1.1 million passengers, mainly because of the competition between air and rail on the Marseille–Paris routes (Zunino 2007). Sharp summarizes that “the Orly route had lost about 30%, but the impact on the Paris Charles de Gaulle route had been much less because of the volume of interlining traffic. Since 2001, Orly traffic has dropped by 43% (although it is still the top destination, with 1.34m passengers in 2005, 23% of the total, compared with Paris Charles de Gaulle at 0.57m and 9.7%) . . . percentage drops in passenger numbers were 8.2 (2001), 8% (2002) and 1.7% (2003)” (Zunino 2007). Corridor #3 Paris–Strasbourg The Setting Service between Paris and Strasbourg was improved in July 2007, improving rail travel times from 4 hours 20 minutes to 2 hour 20 minutes. As shown on Figure 4-5, the 12-month of the Eurostar service between the two city centers (SDG 2006). The Research Team found 40 flights between London and Paris via the legacy carriers, three flights via easyJet, and none on Ryanair. Importantly, this is in a market where the rail system largely absorbed the point-to-point market. Corridor #2 Paris–Marseille The Setting Paris–Marseille is one of the longest corridors—approxi- mately 420 miles—where rail improvements have impacted air volumes. Present travel times range from between 3 hours 6 minutes to 3 hours 20 minutes. Initial service improve- ments occurred in 2001, improving travel time to 4 hours 40 minutes, with reported decreases in air travel by about one million, from approximately three million passengers to two million (SKM 2010). With the improvement to the pres- ent travel times, mode share has increased from 22% in 2001 to 69% in 2006 (SKM 2010). DGAC estimated the present mode share as 67% (SDG 2006a). At present, the O-D market is estimated to be between 80% and 85% to the rail system (SDG 2006a). Within this larger corridor lies the historic Paris–Lyon service opened in the early 1980s. Before the HSR service, air volumes were reported as 180,000 per year (Wilken, reported in Wardman et al. 2002), which fell to 100,000 after the opening and stabilization of the TGV service, a diver- sion of 45% of air traffic. MVA has estimated a 90% sub- Source: anna.aero, 2010. Figure 4-5. Monthly change in air traffic, Paris—Strasbourg. Source: anna.aero, 2010.

59 Madrid and Barcelona, most ridership change occurred with the adoption of the present travel time of 2 hours 40 minutes, resulting in a mode share of 46%. Again, mode share num- bers reflect the use of air segment volumes, which include air passengers connecting with other flights. Corridor #4 Madrid–Seville–Malaga The Setting: Madrid–Seville The first HSR investment in Spain was made in 1992, in the relatively small market between Madrid and Seville, where travel times ranged between 2 hours 30 minutes and 2 hours 40 minutes (i.e., almost no range at all). The distance on the route of the train (via Cordova) is about 290 miles. With the HSR opening, the sole air carrier, Iberia, lost about 35% of its air passenger market (SDG 2006), and the rail vol- umes have grown considerably (e.g., 28% growth between 1999 and 2004) (SDG 2006). Before the new rail service, rail had 56% of the air plus rail market; afterward, rail had 93% of the air plus rail mar- ket. It is important to note that before the new HSR service, the buses were carrying more than the rail. For this reason, Table 4-2 presents the mode shares for all modes, which is not reliably available for most of the corridors presented in this Chapter. What Happened to the Flights? Four flights per day were found from Madrid to Seville, thus eight flights between the two cities. This represents what airlines need to keep network connections marketable, in a moving average of air ridership fell from about 1 million trips to about 325,000 in 2010, a decrease of about two-thirds (anna.aero 2010). During this period, volumes between Strasbourg and Orly fell nearly twice as much as volumes between Strasbourg and CDG, reflecting the lack of diver- sion by the network connecting traffic at CDG (SKM 2010). IARO’s Sharp noted that these travel time improvements “allowed Paris–Stuttgart times to be reduced from 6 hours to 3 hours 39 minutes, Paris–Zürich times to 4.5 hours, Paris– Nancy to 1½ hours, Paris–Metz to 1 hour 24 minutes, and Paris–Luxembourg to just over 2 hours” (Perrin 2007). What Happened to the Flights? IARO reports that “A reaction by Air France was to reduce Paris–Strasbourg frequencies from 12 to 8 return trips a day and to withdraw from the Paris–Metz market entirely ( Redman 2007). Anecdotally, it also damaged the Paris– Stuttgart air market (Laistner 2008). An article in Jane’s Air- port Review noted that Traffic to Lille and Nantes was also hit” (Citrinot 2010). Diversions from Air to Rail in Spain Figure 4-6 shows the changes in rail mode share resulting from travel time improvements in three markets: Madrid to Seville; Madrid to Malaga; and Madrid to Barcelona. It shows that rail share went to 80% when rail travel times improved from 4 hours to 2.5 between Madrid and Seville. Mode share rose to above 60% when rail travel times between Madrid and Malaga dropped from nearly 5 hours to under 3 hours. While there were two phases in the improvements between 0 10 20 30 40 50 60 70 80 90 60 120 180 240 300 360 420 480 Ra il Sh ar e of Ai rP lu sR ai l Rail In vehicle Travel Time in Minutes Madrid Seville Madrid Malaga Madrid Barcelona Figure 4-6. Change in mode share change by travel time—Spain.

60 loss of 25% of this air volume would represent a diversion of 350,000 air travelers to rail. Corridor #5 Madrid–Barcelona The Setting In the important Madrid to Barcelona corridor, some 1.8 million air travelers have been diverted away from air to the new HSR services. After London–Dublin, Madrid– Barcelona is the highest volume air corridor in Europe, and the scale of diversion is the largest reported in this report. Scheduled travel times were between 2 hours 43 minutes and 3 hours 18 minutes. The HSR service opened in Febru- ary 2008; since then, “numbers have fallen from 4.9 million in 2007 to just 3.1 million in 2009, a drop of 37%” in the cor- ridor’s air travel market (anna.aero 2010). Figure 4-7 shows that the air market took a hit between January 2008 and the beginning of 2009, but that it has stabilized between that time and the present. In fact, the air market share has doggedly remained higher than the rail share, as can be concluded from Figure 4-8, which traces the absolute volumes of both air and rail on the same graph. Visually, the rail volumes, on occasion, closely approach the air volumes but never surpass them, ending in a 46% rail mode share. It is reported that between 12–15% of passengers on the Iberia flights between the two cities were connecting network corridor where the air service is largely irrelevant for the O-D (i.e., non-transferring) market—in a corridor where the train diverted as many riders from bus as from air. Madrid–Malaga In late 2007, RENFE established a new high-speed link between Madrid and Malaga, using infrastructure from the original Madrid–Seville line, with new track between Cordoba and Malaga. Travel times were between 2 hours 32 minutes and 2 hours 50 minutes in the timetable for spring 2001—improved from 4 hours 20 minutes. Reportedly, air traffic has fallen by 25% (Railway Gazette International 2009). IARO reports that since 2007, market share has grown from around 50% to nearly 70%. IARO reports that the route had 20 flights each way each day, carrying over 1.4 million passengers in 2005. Thus, a Mode Share Before HSR Share Aer HSR Auto 60% 34% Bus 15% 8% Air 11% 4% Rail 14% 54% Source: SDG 2006. Table 4-2. Madrid-Seville mode share change. Source: anna.aero, 2010. Figure 4-7. Volumes of air and rail passengers between Madrid and Barcelona. Source: anna.aero, 2010.

61 revealed that three airlines were serving the market with a combination of Boeing 737 and Airbus 310 and 320 aircraft. There is no evidence of a shift to regional jets or propeller aircraft. Corridor #6 London–Manchester Figure 4-9 shows the change in mode share for three cor- ridors, and present mode share between London and sev- eral major destinations. While the London to Glasgow share reflects the fact that the service is not under 3.5 hours of ter- minal-to-terminal travel time, the other two corridors look remarkably similar. The corridors between London and Brus- sels and London and Manchester show marked share growth related to travel time improvement. passengers before the HSR service began (SDG 2006). Rail now captures approximately 55% of the air plus rail mar- ket for the O-D market of travelers between the two regions, assuming that those individuals did not divert to rail. What Happened to the Flights? The airlines are not giving up on this market. The time- table for spring 2011 showed three flights per hour, leav- ing Madrid to go to Barcelona—an intensity of service that could match almost any shuttle service in the United States. Both airports are relatively well located and served by reli- able public transportation; years of market loyalty to the “air-bridge” concept of continuous flight offerings will be tested in coming years. A review of off-peak flight schedules Note: The upper line represents air passengers; the lower line represents rail passengers. 0 50 100 150 200 250 300 350 400 450 500 ja n 06 m ar 06 m ay 06 ju l0 6 se p 06 no v 06 ja n 07 m ar 07 m ay 07 ju l0 7 se p 07 no v 07 ja n 08 m ar 08 m ay 08 ju l 0 8 se p 08 no v 08 ja n 09 m ar 09 m ay 09 ju l0 9 se p 09 no v 09 ja n 10 m ar 10 m ay 10 ju l1 0 se p 10 no v 10 ja n 11 m ar 11 m ay 11 ju l1 1 se p 11 no v 11 Commercial passengers in miles (departures MAD + departures BCN) Figure 4-8. Volumes of air and rail passengers between Madrid and Barcelona. Source: Spanish Civil Aviation Agency. 0 10 20 30 40 50 60 70 80 90 100 60 120 180 240 300 360 Pe rc en t Ra il Sh ar e of Ra il Pl us Ai r Rail In vehicle Travel Time, in Minutes London-Manchester London-Glasgow London-Brussels Edinburg Amsterdam Frankfurt London Glasgow, 2010 and Historic Figure 4-9. Mode share by travel time—United Kingdom.

62 that between 2006 and 2010, passenger journeys by air between Glasgow and London airports went down by 22.4%, while ATOC information confirms a remarkable 85.8% increase in rail pas- senger journeys between the cities. Rail’s share of the rail/air market between Glasgow and London had been as low as 6% in the early years of the 21st century but has grown significantly since then, reaching 10% in 2006, 12% in 2008, 16% in 2009 and 20% in 2010” (Virgin Trains–The Media Room 2011). Corridor #8 Amsterdam Just as HSR has become a reality in the corridors docu- mented, the full implementation of the national scheme in the Netherlands has been delayed by a variety of factors, with HSR service commenced only in 2012. However, during the recent period, considerable research has been examining the effect that HSR might have on Schipol, and the extent to which potential air passengers will be diverted away to rail systems. Forecasts, with a 4% to 13% decrease, have been made for Schipol air passenger volumes (Wardman et al. 2002). Amsterdam and the Potential for Substitution Looking at the present use of rail, Jorritsma has summa- rized that the rail market from Amsterdam has, “. . . a volume of about 3.5 million passenger journeys to France and Belgium. About one-third (1 million journeys) is accounted for by the high-speed train to Paris (Thalys). The other 2.5 million concern journeys to destinations in Belgium, includ- ing transfers to London (Eurostar) at Brussels. As a result of the introduction of the HST, the amount of journeys on the southern corridor will increase to 5.7 million journeys in 2010. In particu- lar, the market with London will be growing rapidly. Intraplan (Intraplan 2003) estimates a growth of 23% in railway journeys on the market to Brussels, 28% growth to Paris, and an increase of 42% to Frankfurt in the period 2010–2020” (Jorritsma 2009). Jorritsma has calculated the following volumes of passen- gers from Amsterdam in 2020 would be diverted from air to rail: • From the Brussels market: 125,671 • From the Paris market: 299,351 • From the London market: 1,152,092 This results in 1,500,000 diverted passengers. How would this level of diversion impact the airport and its levels of con- gestion? In dealing with a key issue examined in this report, Jorritsma concludes, “This results in a total of 1.6 million potential passengers to substitute from airplane to HST in 2020. The amount of flights that can be reduced on Schipol airport in 2020 would be about In London-Manchester, the corridor of principal concern, the train trip takes 2 hours 8 minutes (spring 2011 schedule) making the corridor a prime candidate for majority mode share over air, in a corridor of under 200 miles. Importantly, a quick survey of schedules showed three direct trains per hour. This represents a major difference in emphasis from the French/Spanish focus on the speed of the train versus the United Kingdom’s present experiment with “Very High Fre- quency” services. HSR services started in 2004, but major reli- ability problems plagued the operation for some years, until the period around 2004–2005. The present time of 2 hours 8 minutes is 22 minutes faster than service offered in 1999. The IARO reports that, “The rail share of the London– Manchester traffic changed from 49% in 2003 to 70.4% in 2008, with the total market growing from 3.8m to 4.6m pas- sengers (Modern Railways 2009).” In the first full year of oper- ation of the Virgin High Frequency service, the rail market share grew to over 80% (http://www.travelagentcentral.com). In a thoughtful analysis, SDG argues that the analysis of short-distance markets should cautiously apply the metric rail divided by air plus rail. For a given year (2004–2005), the use of one market definition (dominant rail station in Man- chester to dominant rail station in London) the rail volumes would make up only a 45% share of the air plus rail market. “If we define London–Manchester as all travel from a broad area of southeast England to all of Manchester and the area around it, the market share is substantially higher (60% com- pared with 45%)” (SDG 2006). Methodological issues such as this are key concerns to this report. In the Northeast Corridor in the United States, similar fluctuations on rail mode share are seen as a function of regional definitions. During the period between the commencement of service by the new operator in 1997 and 2004, rail travel times did not get better, and rail mode share did not get better. Then, a new strategy of higher frequency rail service, with improved reliability, has seen significant increases in rail ridership and reported mode share. If rail mode share has gone from 60% in 2004, to 70% in 2008, to 80% in 2010, this implies that between 2004 and 2010, rail volumes have increased by roughly one million persons who might have gone to air. However, not cal- culated is a single “rail diversion from air.” In a corridor of this length, it is not clear what portion of those million riders has been diverted from air or auto or induced demand. Corridor #7 London–Glasgow Trains between London and Glasgow take 4 hours 30 min- utes, travel approximately 350 miles, and occur hourly. In the spring of 2010, Virgin Trains reported, “Based on a comparison of rail industry figures with data published by the Civil Aviation Authority (CAA), a long-term change in travel patterns has been identified. CAA figures show

63 rail has gained market during the period of significant infra- structure (speed) improvement (Jorritsma 2009). Bibliography anna.aero. 2010. http://www.Anna.aero. (As of December 2010). Atkins Limited. 2011. “International Connections, a Report for HS2 Ltd.,” London. Citrinot, L. 2010. “Strasbourg Faces SEA Change.” Jane’s Airport Review, Vol. 22, No. 2, p. 16. Eisenkopf. 2006 “Competition between Low Cost Carriers and Railways in Germany,” quoted in Jorritsma (2009). Fox, J. T. 2010 (January 4). “Virgin Trains Announces First Year Report.” Available at http://www.travelagentcentral.com. (As of May 1, 2010). IARO. 2009 The International Air/Rail Organization, Andrew Sharp, author “IARO Report 12.08, High Speed Rail Competition.” (Draft). London, U.K. Jorritsma, P. 2009. “Substitution Opportunities of High Speed Train for Air Transport.” Aerlines Magazine, e-zine edition 43 (at http://www. aerlines.nl). Laistner, A. 2008 (November). Personal communication to Andrew Sharp. M3 Systems, Lead Consultant, 2004 “WP1 Review of the Current Inter- modality Situation,” Eurocontrol, European Organization for the Safety of Navigation, Brussels. Modern Railways. 2009 (August). “Britain’s Strategic Artery.” Issue 731, p. 74. Perrin, B. 2007 (June). “TGV Est Opens for Business.” Modern Railways, Vol. 65, June, page 56. Railway Gazette International. 2009 (January). “Traffic Grows by 900%.” Vol. 165, Issue 1, p. 10. Redman, C. 2007 (November 12). “The Fast Track.” Fortune, Vol. 156, Issue 9, pp. 36-39. Available at http://money.cnn.com. SKM. 2010 (February). “International Rail Travel Demand Model for HS2.” Sinclair Knight Merz, London, U.K. SDG. 2006. Steer Davies Gleave. “Air and Rail Competition and Comple- mentarity: Final Report” Prepared for the European Commission DG TREN. London, U.K. SDG. 2006a. Steer Davies Gleave. “Air and Rail Competition and Com- plementarity: Case Studies” Prepared for the European Commission DG TREN. London, U.K. Spanish Civil Authority. 2010. Personal Communication to Ashraf Jan, FAA. Virgin Trains—The Media Room. 2011 (March 28). “Former Fliers Get on Track.” Press release available at http://www.mediaroom. virgintrains.co.uk. Wardman, M., A. Bristow, J. Toner, and G. Tweddle. 2002 (May). “Review of Research Relevant to Rail Competition for Short Haul Air Routes.” European Organisation for the Safety of Air Navigation EUROCONTROL, Brussels, Belgium. Zunino, G. 2007 (March/April). “Marseille Gateway to Provence and the Mediterranean.” Airports of the World, Issue 7, page 68. 16,000 per year (assuming 100 passengers per flight). This means a total reduction of approximately 2.5% of all flights that will be handled in 2020” (Emphasis added) (Jorritsma 2009). Thus, the major rail investments surrounding the Schipol airport—and those directly connected to it—could lower the number of flights by 16,000 per year, or 2.5% of flights. The pro- fessional literature on this subject does not address the extent to which the airline managers will be motivated to lower the num- ber of flights or lower the number of seats on a fixed number of flights. This report addresses the airlines’ responses. Given that there could be a reduced number of flights between Schipol and Brussels, Paris, Frankfurt and London, the research ques- tion asks how airlines will respond to the decrease in O-D traffic while defending the highly prized hub-based network connections. As a reminder, even though rail claims an esti- mated 85% of the O-D air market between London and Paris, some 40 flights per day remain in the corridor. Major Factors Influencing the Aviation Market The available data show there has been a three-phase reaction to effective city-to-city services. In the first phase, the legacy carriers did what would be expected and sig- nificantly decreased the number of flights in the corridor (Madrid– Barcelona seems to be an exception). With esti- mated point-to-point rail market shares often above 85%, a pattern emerges in which the network connections desired by the dominant carrier at a major hub do not disappear, and take the form of day-long series of connections justi- fied by network demands. In the third phase, point-to-point low-priced carriers challenge the role of dominant hubs. The interest of these carriers in serving the traditional hub airport is not known; these carriers prefer airports like Luton and Stansted for London, Hahn for Frankfurt, and Charlerois for Brussels. The evidence suggests it may be misleading to look at transfer of volumes to rail without examining simultane- ously the nature of the network role and the shift in demand to aviation products less reliant on traditional hub airports. Eisenkopf (2006) has estimated that low-cost airlines have a substitution rate of air trips for rail trips at 5% from Cologne to Hamburg and 13% between Cologne and Munich, while