n° 9

Thematic Report of the French Air Transport Directorate december 2008

Air traffic between and the United Kingdom: air transport between the Regions, between capitals

Preamble

"-London" was the most important international air connection for a long time. Since 8 February 1919, when the first passenger air service in the world was launched by the Farman Company (Paris-London in 2:37), the traffic on this line has kept growing, reaching over four million passengers. With the construction of the Channel Tunnel and the TGV Nord in 1994 and the high-speed lines in England, rail traffic increased between Paris and London. From 2007, it was no longer the first airline departing from France. However, the dynamic offers from low cost companies led to strong growth in traffic between France and England, no longer driven by connections between capital cities. This review provides an original analysis of the impact that a new (low-cost) airline offer has on demand.

1. Cross -Channel traffic: …in which air transport a stagnating market since 2000… is winning market share (see figure 2 & 3)… After steady growth (9.3 % on average per year) during the four years following the launch of the Although it is still the most popular mode of trans- (November 1994), cross channel traffic languished ••port, passenger traffic on ferries decreased to 18.2 at 45 million passengers per year (figure 1). The million in 2007 from 28.8 million passengers in 1994 ferry remains the preferred cross channel mode of and its market share was halved (39 % in 2006). transport (17.6 million passengers in 2006 i.e. 39 % Railway traffic dropped by 13 % between 1998 and of the total). This is followed by train travel with 15.5 ••2007, from 18.6 million passengers in 1998 to 16.1 million passengers (34 %) and air travel with 12.1 million in 2007. This decline is attributable to the million passengers (27 %). decreasing number of Eurotunnel shuttle passengers. On the other hand, Eurostar saw its traffic increase 55 30 Eurotunnel shuttles* by 3.3 % per year between 1997-2007Sea transport and its share 50 Eurostar 25 45 Air (France - UK excl. Paris-London) in cross-Channel traffic amounts toAir transport17.7 % in 2007 40 Air (Paris-London) Ferries (9 French ports) 20 compared to 12.4 % in 1997. 35

30 In 2007, Eurostar passenger numbers rose by 5.1 %, a 15 25 rate slightly lower than 2006 (5.4 %), but, the traffic 20 10 15 has been increasing by 11 % since the middle of 10 5 November 2007, the date of the opening of high- 5

0 0 speed tracks in the UK. From January to September 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1990 1992 1994 1996 1998 2000 2002 2004 2006 * Passengers of vehicules on "tourism" shuttles 2008, Eurostar traffic grew by 13.9 % compared to Source : SNCF, DGAC, DTMPL, Eurotunnel the same period of 2007. As for the airline traffic between France (mainland) Figure12 1. Overall cross channel traffic (millions ofEurostar passengers) •• Air (Paris-London) and the UK, despite the launch of the Eurostar it has 10 Air (France - UK excl. Paris-London) regularly and steadily increased since 1995 (5.5 % 8 per year). This transport mode keeps winning market Present 6 share (27 % in 2007 vs. 21 % in 1994) thanks to low-cost activity development. for 4 the future 2

0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Ministry of Ecology, Energy, Sustainable Development and Town and Country Planning

www.aviation-civile.gouv.fr THEMATIC REPORT | december 2008

100% 0,10 and the United Kingdom (excluding the Paris-London line) is 90% 0,17 21% 0,24 marked by three major phases (see figures 5 and 6): 0,11 35% 80% 36% 0,18 70% 0,12 From 1986 to 1990, passenger traffic increased at a very high 60% 0,09 •• 15% 0,22 rate of 17 % per year on average mainly due to the strong 26% 50% 0,06 increase of flight frequencies (+20 % per year). During this same 0,05 40% 0,78 period, passenger traffic between France and other European 30% Eurotunnel shuttles countries rose by an average of 11.5 % per annum; Eurostar 0,49 20% Plane (France–Uk excl. Paris-London) 0,39 Plane (Paris-London) 10% Ferries (9 French ports) From 1990 to 1997, traffic increased by only 4.4 % per year 0% ••on average and the number of aircraft movements increased 1994 1998 2007 slightly (2.5 % per year). At the same time, passenger traffic Figure 2. Distribution of cross-channel traffic by transport mode between France (outside Paris) and the rest of Europe 1 Note: from January to October 2008, air traffic between France and UK decreased by 1.6 % compared to the same period of 2007. increased by 6.9 % per year. The slowdown in traffic can be partly explained by the consequences of the Gulf War and lower economic growth;

55 30 Eurotunnel shuttles* Sea transport 50 Eurostar Rail transport From 1997 to 2007, traffic increased dramatically (10 % per year 25 45 Air (France - UK excl. Paris-London) Air transport ••on average, compared to 7.2 % for the rest of the European 40 Air (Paris-London) Ferries (9 French ports) 20 35 Union) with growth of the airway offer - (5.8 % per year in 30 number of movements) - and in particular the development 15 25 of low-cost carriers whose market share rose to nearly 65 % 20 10 in 2007. However, the traffic of traditional scheduled airlines, 15

10 5 facing strong competition, has been decreasing since the early 5 2000s. The growth of traffic between France and the United 0 0 Kingdom remained steady even between 2000 and 2003, 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 1990 1992 1994 1996 1998 2000 2002 2004 2006 * Passengers of vehicules on "tourism" shuttles a period during which air transport was affected by several Source : SNCF, DGAC, DTMPL, Eurotunnel Figure 3. Sea, air, and rail cross-Channel traffic external events (September 2001, Iraq conflict, SARS). (millions of passengers) 12 Eurostar 55 3012 000 000 140 000 Eurotunnel shuttles* aagr 1997-2007Sea transport : +10,2%/year aagr 1997-2007 : +5,8%/year Air (Paris-London) 50 10 Eurostar Rail transport Total Air (France - UK excl. Paris-London) 120 000 45 Air (France - UK excl. Paris-London) 2510 000 000 Low-cost cies Total Air transport ...mainlyAir (Paris-London) departing from regional Trad cies 8 40 Low-cost cies 100 000 charter Ferries (9 French ports) 208 000 000 Trad cies 35 airports, the Paris-London air connection charter 6 aagr 1986-1997 : +7,6%/year 30 80 000 suffers from Eurostar competition. 156 000 000 25 aagr 1986-1997 : +8,9%/year 4 60 000 20 The air transport between both countries has been marked by 104 000 000 15 40 000 2 two contrasting changes (cf. figure 4): 10 52 000 000 Paris-London air traffic has dropped by half: 4 million pas- 20 000 0 5 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 • 0 0 0•sengers in 1994, only 2.2 million in 2007 ; 0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 19901986 1992 1988 1994 1990 1992 1996 1994 1998 1996 2000 1998 2000 2002 2002 2004 2004 2006 2006 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 aagr : annual average growth rates * only passenger aircrafts - aagr : annual average growth rates * PassengersAir traffic of vehicules between on "tourism" French shuttles regions and the UK (excluding Paris- •Source•London : SNCF, DGAC flights), DTMPL increased, Eurotunnel 2.6-fold between 1997 and 2007. Figure 5. Air traffic between France and UK Note: from January to October 2008, air traffic between France and UK (excl.

12 Paris-London) grew by 1.0 % compared to the same period of 2007: +0.7 % Eurostar for low-cost compagnies, -0.01 % for traditional scheduled airlines and Air (Paris-London) 10 +9.0 % for charter compagnies. Air (France - UK excl. Paris-London)

8 12 000 000 140 000 aagr 1997-2007 : +10,2%/year aagr 1997-2007 : +5,8%/year Total 6 120 000 10 000 000 Low-cost cies Total Trad cies Low-cost cies 4 100 000 charter 8 000 000 Trad cies charter aagr 1986-1997 : +7,6%/year 80 000 2 6 000 000 aagr 1986-1997 : +8,9%/year 60 000

0 4 000 000 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 40 000

2 000 000 Figure 4. Air and rail cross-Channel traffic (Eurostar) 20 000

(millions of passengers) 0 0 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 Note: from January to October 2008: Eurostar traffic increased by 13.9 %, aagr : annual average growth rates * only passenger aircrafts - aagr : annual average growth rates air traffic between Paris and London decreased by 13.7 % and air traffic between France and UK (excl. Paris-London) grew by 1.0 % compared Figure 6. Number of aircraft movements between France and UK to the same period of 2007. (Excl. the Paris- London line, thousands of passengers) Note: from January to October 2008, air traffic between France and UK (excl. Paris-London) decreased by -0.04 % compared to the same period of 2007: -2.2 % for low-cost compagnies, -2.5 % for traditional scheduled airlines 2. Strong growth of air traffic between and +19.5 % for charter compagnies. the French regions and the United Kingdom: the role of low-cost companies 1. International connections between France and 26 member states of the Since 1986, low-cost company traffic has posted over 10 % (excluding the United Kingdom) without the key competitive growth per year. The development of air traffic between France TGV lines.

Ministry of Ecology, Energy, Sustainable Development and Town and Country Planning This traffic growth has been accompanied by a sharp increase 2 10 000 000 in the number of scheduled routes - which have increased observed 9 000 000 estimated threefold between 1996 and 2007 (see figure 7) – in 2007 Without low-cost cies (with a number of movements equal to 90 000 in 2005) low-cost companies operated almost 80 routes compared to 8 000 000 only thirty for traditional scheduled airlines. 7 000 000 6 000 000

5 000 000 100

90 4 000 000 Total Scheduled 80 Low-cost companies 3 000 000 70 Traditional scheduled cies 2 000 000 60 1 000 000 50 0 40 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 30 * excluding Paris-London

20 Figure 8. Estimation of air traffic between France and UK* 10 with and without low-cost companies (passengers) 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Figure 7. Number of scheduled routes2 between France and UK (excl. the Paris-London line) Low cost companies have created extra Note: from January to October 2008, nearly 70 routes are operated by low- cost compagnies and about 30 by traditional scheduled airlines. traffic, which accounts for one third of the total traffic. An econometric explanation If there hadn’t been the low-cost companies, the number of aircraft movements between France and the United Kingdom GDP and price variables alone cannot explain the (excl. the Paris-London line) would have been lower than in strong growth in air traffic since 1997 3 2005 (116 000). The average carriage would also have been lower. If the number of movements was 90,000, the traffic in Several econometric models were developed to explain the air traffic 2005 would have decreased by 34 % compared to the traffic between France and the United Kingdom. The best model takes into account the average GDP of both countries by volume (GDP) and the observed the same year, which would correspond to a loss average price of air transport without inflation (Price TA), and also the of 3.1 million passengers (see figure 8). According to these number of aircraft movements (Mvt) and the share of movements of assumptions, traffic would have grown from 1996 to 2005 at low-cost companies (Part BC). As for the models explaining the traffic an average of 5.6 % per year instead of 10.6 %. using only GDP and price of air transport variables, they underestimate the increase in traffic between 1996 and 2005. The price “effect" is taken into account through variables "price of air transport" and "share of low-cost". Why doesn’t the variable 3. What future for the Paris-London “price of air transport” alone enable to take into account the full connection… price effect caused by the development of low-cost companies? Two reasons can explain this: Before Eurostar, the growth of air traffic between Paris and The calculation of average unit revenue for all companies, both London was steady: 7 % per year between 1986 and 1994. ••low-cost and traditional, could minimize the effect linked to Between 1990 and 1994, traffic grew by only 3.2 % per year, the drop in price offered by low cost companies only, which have an average unit revenue about 30 % lower than the unit due in particular to the gloomy economic situation. In 1997, three revenue of traditional companies. The passengers’ response years after the launch of Eurostar, air traffic had already lost 1.2 would not be in line with the price changes. million passengers on the Paris-London route (from 4 million in The average unit revenue is calculated for the company’s entire 1994 to 2.8 million in 1997), i.e. 30 % of total traffic. The traffic ••network. However, it is likely that traditional companies have then remained at about 3 million passengers per year until 2002. offered, on lines in competition with low-cost companies, With the commercial launch of the British high-speed line in higher price reductions than on other lines. In this case, the two stages (in September 2003 and November 2007), traffic tariff index selected would minimize the actual price drop between 1997 and 2005. dropped to 2.2 million passengers in 2007 even with low cost companies such as easyJet that operate four times with flights Apart from the price effect, the variable “share of low-cost going there and back between Roissy and Luton. Meanwhile, companies " takes into account some of the effect in relation Eurostar traffic grew from 6 million passengers in 1997 to over with the great range of offers of low-cost companies. 8 million in 2007, this transport mode attracts nearly 80 % of The model is as follows: the Paris-London travellers. (see figure 9). ln traffict = -1.62 + 1.44 ln PIBt – 0.20 ln TAt price + 0.53 ln (6.0) (-1.8*) (9.6) 90% Mvtt + 0.64 BCt share+ ut (5.3) 80% The numbers in brackets are Student ‘t’ 70%

Period: 1987-2005 60%

Observations: 19 50% Adjusted R2: 0.997 Durbin-Watson: 1.82 40% 30% *by applying a “boostrap” test, the elastic nature of the - 20% rage air price is significantly different from 0 at a 5 % level. 10%

0% 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2. Airport-airport connections - only the routes with over 240 return flights per year. Figure 9. Paris and London Eurostar market share 3. Traffic France-UK excluding Paris-London. ( % in the air traffic and Eurostar)

32 …as the Eurostar has diverted nearly With increasing competition from the 5 million passengers from air in 2007 Eurostar since November 2007, air traffic should fall by 15 % to 20 % in 2008 Without Eurostar, air traffic between Paris and London would have been higher than 7 million passengers in 2007: traffic Since the 14 November last year, London is only 2h15 from Paris loss due to the launch of the Eurostar is estimated at about 5 with Eurostar, instead of 2h35, Saint-Pancras being the new million passengers (see figure 10). With 2.2 million passengers London Eurostar terminal instead of Waterloo. This twenty minute a year, in 2007 the Paris-London line lost its position as the first reduction in the journey time should cause an approximate international air connection departing from France, in terms of diversion rate of air passengers to Eurostar of 15 % 4. By taking traffic volume. The Paris- line (2.3 million passengers in into account the drop of traffic observed before November 2007) has now overtaken the Paris-London line. 2007 and the improvement of the Eurostar offer, the flow of air passengers between Paris and London should reach 1.76 to 8,0 1.90 million passengers in 2008 (figure 11). Orly 7,0 CDG +4,6% per year without Eurostar 6,0 8,0 Traffic diverted from +7% / year Paris- EurostarOrly Time plane to Eurostar * 7,0 5,0 Loss : approx 5 million CDG London +11% / year Diversion rate * passengers in 2007 +4,6% per year without Eurostar 6,0 4,0 +7% / year 2008 20 minutes time About 15 % 3,0 5,0 Loss : approx 5 million +11% / year saving (2h15 300 000 passengers in 2007 2,0 4,0 instead of 2h35) to 350 000

1,0 3,0 passengers

2,0 0,0 The impact of the improved Eurostar offer on Paris-London air

1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 traffic1,0 in 2008 * Traffic loss (relative and absolute value) compared with the base situation 0,0 Figure 10. Air traffic between Paris and London equals the lack of improvement of the Eurostar offer. This calculation is based

(millions2,5 of passengers) on the 1986 assumption1987 1988 1989 that1990 1991 easyJet1992 1993 continues1994 1995 1996 using1997 1998 the1999 Paris-Luton2000 2001 2002 2003 line.2004 2005 2006 2007 Drop by 15% to 20% of the traffic btw 2007 and 2008 2,0 2,5 Nearlyobserved one third of the Paris-London Drop by 15% to 20% forecasted (high scenario) of the traffic btw 1,5 2007 and 2008 line passengersforecasted (low scenario) have connecting 2,0 flights departing from Paris observed 1,0 forecasted (high scenario) 1,5 or London. forecasted (low scenario)

0,5 These customers are less interested in the Eurostar 1,0 offer,0,0 which does not serve Heathrow or Roissy directly 2007 2008 0,5

0,0 Millions 2007 2008 of passengers Distribution

Traffic origin-destination 1,6 6 9 % Figure 11. Air traffic between Paris and London (millions of passengers) Interconnecting passen- 0,7 31 % gers in Paris or London

Paris-London Line 2,4 100 % 4. This diversion rate was calculated with a logit model explaining the plane/ train market share from times, frequencies and fares of both modes and by differentiating the OD traffic and the interconnecting air passengers. The OD traffic and the interconnections on the Paris-London The traffic caused by the improvement of the railroad offer was also taken airline in 2006 into account in the calculations. This diversion rate is the variation of the traffic between the base situation (without improvement of the railway Source: MIDT offer) and the project situation (with improvment of the railway offer) in 2008.

For this last issue of the "notes thématiques de la Dast "which will soon be back in a new format, I would like thank the whole team of "Prospective et veille stratégique" led by Elisabeth Bouffard-Savary, who was determined to combine, in these documents, a recent and "thorough" analysis of the major themes of civil aviation in the most communicating way possible, thanks especially to the valuable help of the DGAC’s communication team.

Philippe Ayoun, Deputy Director Sub Directorate of Studies, Statistics and Foresight

Direction générale de l'Aviation civile – Direction du Transport aérien – 50 rue Henry Farman - 75720 Paris cedex 15 www.aviation-civile.gouv.fr – Phone: 00 33 1 58 09 49 54 – Fax: 00 33 1 58 09 48 49 Publisher: Paul Schwach – Chief Redactor: Elisabeth Bouffard-Savary – Redactor: Franck Lisio –Layout: Raphaëlle Vial – Printing: French Civil Aviation

www.developpement-durable.gouv.fr