Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

Connectivity Developments in Air Transport Networks at Primary Asian

Hidenobu MATSUMOTO a, Koji DOMAE b a Graduate School of Maritime Sciences, University, Kobe, 658-0022, ; E-mail: [email protected] b College of Global Communication and Language, Kansai Gaidai University, Hirakata, 573-1008, Japan; E-mail: [email protected]

Abstract: This paper measures and compares the connectivity developments in air transport networks at the primary airports in Asia. To determine how the connectivity at these airports has developed in the specific markets, the connectivity figures are broken down by regions. For an assessment of the model and its application, the paper conducts scenario analyses for Chubu Centrair International in , Japan, on the connectivity impacts of an additional flight from this airport to large hub airports in Europe, North America and Asia, and of moving all domestic flights from , the other airport in Nagoya, to this airport. The results reveal that the most striking growth of air network connectivity developments has been found at the three airports in Mainland (Beijing, Shanghai and Guangzhou) and International Airport. The model is helpful for airports to assess their network performance and their competitive hub status vis-a-vis other airports.

Keywords: Air network performance, Competitive hub status, NetScan connectivity model, Scenario analysis, Chubu Centrair International Airport, Asia

1. INTRODUCTION

The growth of hub-and spoke operations has changed the competition among airports in a structural way. Due to the rise of hub-and-spoke networks, compete directly as well as indirectly. Traditional measures on airport performance, such as passenger enplanements and aircraft movements, fail to address in particular indirect connectivity via hubs. To date, many studies have analyzed hub-and-spoke networks. One branch of research is from the viewpoint of economic perspectives, with a focus on economies of density and scope (Caves et al., 1984; Brueckner and Spiller, 1994), hub premiums (Borenstein, 1989; Oum et al., 1995), entry deterrence (Zhang, 1995) and the role of hub-and-spoke networks in alliances (Oum et al., 2000; Pels, 2001). Another branch of research is the field of operations research, where the cost-minimizing approach is used to determine spatial optimization of air networks (Kuby and Gray, 1993; O’Kelly and Miller, 1994; O’Kelly and Bryan, 1998). A third branch uses the geographical approach, in which the structures, performance and spatial dimension of hub-and-spoke networks are analyzed empirically (Ivy, 1993; Shaw, 1993; Bania et al., 1998; Burghouwt et al., 2003). These studies, however, take into consideration air traffic flows purely from the demand aspect, without capturing the airline network structures, schedule coordination and its resulting hub performance from the supply aspect. Consequently, some studies have included the level of schedule coordination in the measurement of performance and structure of hub-and-spoke networks. Veldhuis (1997)

 Corresponding author.

2240 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

analyzes Amsterdam Airport Schiphol, focusing on the quality and frequency of indirect connections. Burghouwt and Veldhuis (2006) evaluates the competitive position of West European airports in the transatlantic market from this viewpoint, followed by Burghouwt et al. (2009) and De Wit et al. (2009) which assess the competitive hub status of primary airports in East and Southeast Asia. The main purpose of this paper is to measure and compare the air network performance and competitive hub status of primary airports in Japan and elsewhere in Asia between 2001 and 2017. Its special focus of attention is Chubu Centrair International Airport in Nagoya, Japan. In this paper, the NetScan connectivity model is used to measure the connectivity developments at these airports, taking into account the quantity and quality of both direct and indirect connections. For an assessment of the model and its application, the paper conducts scenario analyses for Chubu Centrair International Airport on the connectivity impacts of an additional flight from this airport to large hub airports, and of moving all domestic flights from Nagoya Airfield, the other airport in Nagoya, to this airport. The remainder of this paper is organized as follows. The next section provides an overview of the NetScan connectivity model. In Section 3, the connectivity developments in air transport networks at the primary airports in Asia are measured and compared. In Section 4, after describing the dual airports system in Nagoya Metropolitan Area, scenario analyses are conducted on the connectivity impacts of an additional flight to a large hub airport in Europe, North America and Asia, and of moving domestic flights from Nagoya Airfield to Chubu Centrair International Airport, followed by discussion and conclusion in Section 5.

2. MEASUREMENT OF NETWORK QUALITY

2.1 Four Types of Network Connectivity

In our approach, four types of network connectivity are distinguished as described in Figure 1. 1. Direct connectivity: flights between airports A and B without a hub transfer 2. Indirect connectivity: flights between airports A and B, but with a hub transfer at airport H 3. Onward connectivity: connections with a hub transfer at airport B between airports A and D 4. Hub connectivity: connections with a hub transfer at airport A between airports C and B

Direct connectivity A B

Indirect connectivity

A H B

Onward connectivity A B D

Hub connectivity C A B

Figure 1. Four types of network connectivity Note: This paper does not consider onward connectivity.

2241 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

The quality of an indirect connection between airports A and B with a hub transfer at airport H is not equal to the quality of a direct connection between airports A and B. In other words, the passenger traveling indirectly will experience additional costs due to longer travel times, consisting of transfer time and detour time. Transfer time equals at least the minimum connecting time, or the minimum time needed to transfer between two flights at airport H. The measurement of indirect connectivity is particularly important from the perspective of passenger welfare; how many direct and indirect connections are available to passengers between airports A and B? The concept of hub connectivity is particularly important for measuring the competitive hub status of airports in a certain market; how does airport A perform as a hub in the market between airports C and B?

2.2 Concept of Connectivity Units

Many passengers transfer at hub airports to their final destinations, even in case good direct connections are available. Passengers’ choices depend on the attractiveness of the available alternatives. When measuring the attractiveness of a certain alternative, we consider frequencies and travel time. As for fare differentiation, fares on non-stop direct routes are generally higher than those on indirect routes. Fares on indirect routes are generally lower for on-line (or code-shared) connections than for interline connections. Fares on a route are generally lower if more competitors are operating on these routes. And finally, fares are ‘carrier-specific’ and are depending on the ability of carriers to compete on fares. Therefore, it can be concluded that fares are generally depending on the number of competitors on the route and the product characteristics, like travel time, number of transfers, kind of connection (on-line or interline) and the carrier operating on the route. So, fare differentiation is partly reflected in the route characteristics. The route characteristics mentioned are to be operationalized in a variable indicating connectivity, expressed in so called ‘connectivity units (CNU’s)’. This variable is a function of frequencies, travel time and the necessity of a transfer.

2.3 NetScan Connectivity Model

The NetScan connectivity model, developed by Veldhuis (1997), has been applied here to quantify the quality of an indirect or a hub connection and scale it to the quality of a theoretical direct connection. The model assesses the level of direct connectivity based on the Official Airline Guide (OAG) flight schedules. Based on the direct connections, the model builds viable indirect and hub connections. The model weighs these for their quality based on transfer time and detour time involved, which results in the level of indirect and hub connectivity provided. Figure 2 shows the scheme of NetScan Model. First, direct connections have been retrieved from the OAG flight schedules (Step 1). Then, indirect and hub connections have been constructed using an algorithm, which identifies each incoming flight at a hub airport and the number of outgoing flights that connect to it. The algorithm takes into account the minimum connecting time and puts a limit on the maximum connecting time. In our case, we assume 30 minutes between domestic connections and 45 minutes between domestic and international connections and between international connections for the minimum connecting time and 420 minutes for the maximum connecting time. Next, NetScan assigns a quality index to every individual connection, ranging from 0 to 1 (Step 2). A non-stop direct connection is given the maximum quality index of 1. The quality

2242 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

index of an indirect or a hub connection will always be lower than 1, since extra travel time is added due to transfer time and detour time for the passenger. The same holds true for a multi-stop direct connection. Passengers face a lower network quality because of en-route stops compared to a non-stop direct connection. If the additional travel time of an indirect or a hub connection exceeds a certain threshold, the quality index of the connection equals 0. The threshold between two airports depends on the travel time of a theoretical direct connection between these two airports. In other words, the longer the theoretical direct travel time between two airports, the longer the maximum indirect travel time can be. The travel time of a theoretical direct connection is determined by the geographical coordinates of origin and destination airport and the assumptions on flight speed and time needed for take-off and landing. Furthermore, additional time penalties for transfer time have been incorporated into the model. Passengers generally perceive transfer time as more inconvenient than flying time, as additional risks exist of missing connections and loss of baggage. By taking the product of the quality index and the frequency of the connection per time unit (day, week or year), the total number of connections or connectivity units (CNU’s) can be derived.

1. Retrieval of direct connections from OAG Step 1: 1. Minimum connecting time: 30 minutes (Domestic>>>Domestic) 45 minutes (Domestic>>>International International>>>Domestic International>>>International) 2. Maximum connecting time: 420 minutes 2. Indirect and hub connections building

Step 2: NetScan assigns to each connection a quality index, ranging from 0 to 1.

3. Calculation of CNU

Figure 2. Scheme of NetScan connectivity model

Summarizing, the following formulas have been applied for each individual direct, indirect and hub connection (Airports Council International, 2014). The air network developments at each airport are assessed by calculating the direct, indirect and hub connectivity.

flight, nonstop txy = (40 + 0.068*gcdkm)/ 60 (1) perceived, max flight, nonstop flight, nonstop txy = txy + 5ln (txy + 0.5) (2) flight,actual txy for direct flights perceived, actual tx(h) y = (3) flight, actual flight, actual transfer (txh + thy ) + pxy *th for indirect flights

2243 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

perceived,actual flight,nonstop 1 if tx(h) y  txy perceived,actual flight,nonstop t − txy q = 1− x(h) y if t flight,nonstop  t perceived,actual  t perceived,max (4) x(h) y perceived,max flight,nonstop xy x(h) y xy txy − txy perceived,actual perceived,max 0 if tx(h) y  txy where, flight, nonstop txy : non-stop flight time between airports X and Y in hours, gcdkm : great-circle distance in kilometers, perceived, max txy : maximum allowable perceived travel time between airports X and Y in hours, perceived, actual tx(h) y : actual perceived travel time between airports X and Y (via airport H) in hours, flight,actual txy : actual travel time between airports X and Y in hours, transfe r th : transfer time at airport H in hours, flight, nonstop pxy : penalty for transfer time (= = 3 − 0.075txy ), and

qx(h) y : quality index.

First, the maximum allowable perceived travel time is calculated. The maximum allowable perceived travel time between airports X and Y depends on the non-stop flight time between airports X and Y and a factor which decreases with distance. The non-stop flight time is determined by the geographical coordinates of origin and destination airport and the flight speed of an average jet aircraft taking into account the time needed for take-off and landing. Over longer distances, passengers are willing to accept longer transfer and circuity time. Therefore, the maximum allowable perceived travel time also depends on a factor which decreases with distance, indicating that the further apart two airports are, the longer the maximum allowable perceived travel time will be. Second, the actual perceived travel time is determined. For direct connections, the actual perceived travel time between airports X and Y equals the actual flight time. For indirect connections, the actual perceived travel time equals the flight time on both flight legs and the transfer time at airport H. As the transfer time is considered more uncomfortable than the flight time, the transfer time is penalized by a factor which decreases with distance. If the actual travel time is smaller than or equal to the average non-stop flight time, then the weight of the connection equals one. In practice, this is only the case on direct flights operated by aircraft that are at least equally fast as the average jet aircraft on which the non-stop flight time is based. When the perceived travel time becomes larger than the maximum allowable perceived travel time, then the weight of the connection is zero and the connection will be considered unviable. In other cases, the perceived travel time lies between the non-stop flight time and the maximum allowable perceived travel time. In these cases, the weight of the connection depends on the relative difference between the perceived travel time and the maximum allowable perceived travel time. When the perceived travel time is relatively small compared to the maximum allowable perceived travel time, then the weight of the connection will be high, and vice versa. The connectivity of an individual direct or indirect connection equals its quality. The CNU is calculated for each individual direct and indirect connection. This means

2244 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

that when a flight is offered with a daily frequency, the CNU’s for each of these seven flights as well as for each possible connection have been calculated. The reason for distinguishing between individual flights is twofold. First, the flights might be carried out by different types of airplane during the week, leading to different flight time and therefore differing CNU’s. Second, the same flight might connect to different flights, for example, on Monday than on Friday.

2.4 Data and Study Airports

The data used in this analysis are from the OAG flight schedules for the third week of September in 2001, 2009 and 2017. In this study, only online connections are considered as viable connections. In other words, the passenger transfer has to take place between flights of the same airline or the same global airline alliance partners. For the years 2004 and 2007, three global airline alliances are distinguished: , SkyTeam and Star Alliance. For the year 2001, an additional alliance, Wings Alliance, is also distinguished, which submerged into SkyTeam in 2004. In addition, actual codeshare agreements between airlines are also considered when building connections. The study area is specified as East Asia and Southeast Asia. The airports, selected and analyzed in our study, are 15 primary airports in this area: Narita International Airport (hereafter, Tokyo/Narita), Tokyo International Airport (Tokyo/Haneda), Kansai International Airport (), Chubu Centrair International Airport (Nagoya), Incheon International Airport (), Beijing Capital International Airport (Beijing), Shanghai Pudong International Airport (Shanghai), Guangzhou Baiyun International Airport (Guangzhou), Hong Kong International Airport (Hong Kong), Taoyuan International Airport (), Ninoy Aquino International Airport (Manila), Suvarnabhumi Airport (Bangkok), Kuala Lumpur International Airport (Kuala Lumpur), Singapore (Singapore) and Soekarno-Hatta International Airport (Jakarta). The analysis considers the connectivity between or via these airports and airports worldwide.

3. COMPARISON OF NETWORK PERFORMANCE AND HUB COMPETITIVE STATUS AMONG PRIMARY AIRPORTS IN ASIA

3.1 Total Network Connectivity

Figure 3 shows the total network connectivity split up in direct, indirect and hub connectivity at the primary Asian airports in 2017. As for direct connectivity, Chinese airports definitely provided many direct connections: Beijing (5,737 CNU), Guangzhou (4,364 CNU), Shanghai (4,239 CNU) and Hong Kong (3,308 CNU). Jakarta was the second largest airport in this region with regard to direct connectivity and accommodated 4,827 direct flights in this year. Furthermore, Tokyo/Haneda (4,143 CNU), Kuala Lumpur (3,728 CNU), Singapore (3,477 CNU), Bangkok (3,257 CNU) and Seoul (3,027 CNU) offered more than 3,000 direct flights. Indirect connectivity was remarkable at Beijing (11,000 CNU), Singapore (10,561 CNU) and Tokyo/Narita (10,029 CNU), followed by Shanghai (9,903 CNU), Hong Kong (8,739 CNU) and Bangkok (7,953 CNU). With respect to hub connectivity, Beijing, Hong Kong and Singapore were in the first tier, with 16,400 CNU, 14,520 CNU and 12,077 CNU, respectively. Bangkok (10,766 CNU), Tokyo/Haneda (10,117 CNU) and Seoul (10,014 CNU) were in the second tier. Indirect and hub connectivity at Guangzhou and Jakarta, in general, were not so high, compared with direct connectivity.

2245 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

CNU 18,000 Direct connectivity Indirect connectivity Hub connectivity 16,000

14,000 12,000 10,000

8,000 6,000

4,000

2,000 0

Figure 3. Direct, indirect and hub connectivity at primary Asian airports, 2017

3.2 Connectivity Developments

3.2.1 Direct connectivity

Figure 4 describes the direct connectivity developments between 2001 and 2017 at the primary airports in Asia. The highest growth can be found at the three airports in Mainland China. In particular, the figure at Shanghai increased by 598%. That of Guangzhou increased by 262% between these years. One reason concerns the opening of new international airports in 1999 and in 2004 in each of these cities, respectively. Jakarta (288%), Kuala Lumpur (218%) and Seoul (218%) also experienced remarkable growth levels. CNU 7,000 2001 2009 2017 6,000

5,000 4,000

3,000 2,000

1,000 0

Figure 4. Direct connectivity at primary Asia airports, 2001, 2009 and 2017

2246 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

3.2.2 Indirect connectivity

Figure 5 describes the indirect connectivity developments between 2001 and 2017 at the primary airports in Asia. The high growth of indirect connectivity at Tokyo/Haneda (2,494%) can be definitely attributed to the resumption of international air services in 2010. In addition, Guangzhou demonstrated quite a high growth (2,955%). In contrast, other airports in Japan showed negative growth rates: Tokyo/Narita (-19%), Osaka (-32%) and Nagoya (-11%).

CNU

14,000 2001 2009 2017 12,000

10,000

8,000 6,000

4,000

2,000

0

Figure 5. Indirect connectivity at primary Asia cities, 2001, 2009 and 2017

3.2.3 Hub connectivity

CNU 18,000 2001 2009 2017 16,000 14,000

12,000 10,000 8,000

6,000 4,000 2,000

0

Figure 6. Hub connectivity at primary Asia airports, 2001, 2009 and 2017

2247 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

Figure 6 describes the hub connectivity developments between 2001 and 2017 at the primary airports in Asia. Shanghai and Guangzhou experienced the highest growth percentages, with 12,876% and 5,630%, respectively. This was because these two airports had very low levels of hub connectivity in 2001. The high growth of hub connectivity at Tokyo/Haneda (1,528%) can be again due to the resumption of international air services in 2010. Osaka experienced the negative growth percentage (-17%).

3.3 Directional Connectivity

3.3.1 Direct and indirect connectivity

The direct and indirect connectivity figures have been broken down by geographical regions of the final destination to determine how the direct and indirect connectivity at each airport has developed in specific markets: domestic, Japan, China (including Hong Kong and Macau), East Asia, Southeast Asia, West Asia (Middle East), Other Asia, Oceania, Europe, North America, South America and Africa. Figure 7 shows the directional total connectivity (direct and indirect) at the Asian primary airports in 2001, 2009 and 2017. These figures demonstrate in which market each airport has a competitive status. The first observation is that all airports are rather poorly connected to Other Asia (South Asia and Central Asia), South America and Africa. Tokyo/Narita has the best competitive status in the transpacific market and a strong network to European destinations. In addition, it is the gateway to South America, looking at the total connectivity. Yet, the connectivity to domestic destinations is low because of the split-up of operations between Tokyo/Haneda. It experienced the negative growth in the total connectivity to North America over these years. This was because downsized its operations as a consequence of its bankruptcy in 2010, resulting in a considerable loss of connectivity to this region. The total connectivity from Tokyo/Haneda to almost all regions increased drastically over these years, definitely owing to the resumption of international air services at this airport in 2010. The growth in the connectivity to North America and Europe is remarkable. Osaka had, on the other hand, the largest connectivity to Europe in 2017 and high connectivity to domestic and Asian destinations, in addition to North America. However, Osaka experienced negative growth rates over the years. Meanwhile, Nagoya had the largest connectivity to domestic destinations. Seoul increased its connectivity especially to North America and Europe during the period of analysis, mainly because of the successful network growth of Lines and to these regions. Seoul has little connectivity to domestic destinations because of the split-up between Gimpo International Airport, the other airport in Seoul Metropolitan Area. With respect to the four Chinese airports, Beijing and Guangzhou are highly accessible from a domestic point of view (directly or indirectly), whereas Shanghai shows, besides domestic connectivity, strong connectivity to North America. Hong Kong is serving, on top of domestic destinations, North America and Europe, East and Southeast Asia and Oceania. These Chinese airports demonstrate high percentage growth rates during the years analyzed. Taipei had the largest connectivity to North America in 2017. As for the five ASEAN airports, the connectivity of Bangkok and Singapore are characterized by a strong hub status in the European market with the modest growth rates. On the other hand, Manila, Kuala Lumpur and Jakarta are much more oriented towards domestic and Asian destinations. Kuala Lumpur and Jakarta show high growth rates throughout the period of analysis.

2248 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

Domestic Domestic Domestic 10,000 2001 4,000 2001 2,500 2001 Africa China 2009 Africa China 2009 Africa China 2009 8,000 3,000 2,000 2017 2017 2017 6,000 1,500 2,000 South America 4,000 East Asia South America East Asia South America 1,000 East Asia 2,000 1,000 500 0 0 0 North America Southeast Asia North America Southeast Asia North America Southeast Asia

West Asia West Asia West Asia Europe Europe Europe (Middle East) (Middle East) (Middle East)

Oceania Other Asia Oceania Other Asia Oceania Other Asia 1. Tokyo/Narita 2. Tokyo/Haneda 3. Osaka

Domestic Domestic Domestic 1,000 2001 4,000 2001 7,000 2001 Africa Japan Africa China 2009 2009 Africa 6,000 Japan 2009 800 3,000 2017 2017 5,000 2017 600 South America 2,000 China 4,000 South America 400 East Asia South America 3,000 East Asia 1,000 2,000 200 1,000 North America East Asia 0 0 0 North America Southeast Asia North America Southeast Asia Europe Southeast Asia West Asia West Asia Europe Europe (Middle East) West Asia (Middle East) Oceania (Middle East) Oceania Other Asia Other Asia Oceania Other Asia 4. Nagoya 5. Seoul 6. Beijing

Domestic Domestic China 4,000 2001 6,000 2001 4,000 2001 Africa Japan 2009 Africa 5,000 Japan 2009 Africa Japan 2009 3,000 3,000 2017 4,000 2017 2017 2,000 3,000 2,000 South America East Asia South America East Asia South America East Asia 2,000 1,000 1,000 1,000 0 0 0 North America Southeast Asia North America Southeast Asia North America Southeast Asia

West Asia West Asia West Asia Europe Europe Europe (Middle East) (Middle East) (Middle East)

Oceania Other Asia Oceania Other Asia Oceania Other Asia 7. Shanghai 8. Guangzhou 9. Hong Kong

Domestic Domestic Domestic 2,500 2001 1,500 2001 4,000 2001 Africa Japan Africa Japan Africa Japan 2,000 2009 2009 2009 3,000 2017 1,000 2017 2017 1,500 South America China South America China South America 2,000 China 1,000 500 500 1,000 North America 0 East Asia North America 0 East Asia North America 0 East Asia

Europe Southeast Asia Europe Southeast Asia Europe Southeast Asia

West Asia West Asia West Asia Oceania Oceania Oceania (Middle East) (Middle East) (Middle East) Other Asia Other Asia Other Asia 10. Taipei 11. Manila 12. Bangkok

Domestic Japan Domestic 2,000 2001 5,000 2001 5,000 2001 Africa Japan Africa Japan 2009 Africa China 2009 2009 1,500 4,000 4,000 2017 2017 2017 3,000 3,000 South America 1,000 China South America China South America 2,000 East Asia 2,000 500 1,000 1,000 North America 0 East Asia 0 North America 0 East Asia North America Southeast Asia

Europe Southeast Asia Europe Southeast Asia

Europe West Asia West Asia West Asia Oceania Oceania (Middle East) (Middle East) Other Asia Oceania Other Asia Other Asia 13. Kuala Lumpur 14. Singapore 15. Jakarta

Figure 7. Directional direct and indirect connectivity at primary Japanese Airports, 2001, 2009 and 2017

2249 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

3.3.2 Hub connectivity

Meanwhile, the hub connectivity figure has been broken down by geographical regions of the final destination to determine how the hub connectivity at each airport has developed in specific markets: domestic-Domestic, Asia-Asia (Within Asia), Asia-Oceania, Asia-Europe, Asia-North America, Asia-South America, Asia-Africa and intercontinental. Figure 8 shows the directional hub connectivity at the Asian primary airports in 2001, 2009 and 2017. There are some geographical differences with respect to the hub connectivity among these airports. For example, Tokyo/Narita shows the strongest hub connectivity to North America, whereas Tokyo/Haneda has the largest hub connectivity in the domestic market. Seoul has the large hub connectivity to North America, Europe and within Asia. Meanwhile, Beijing shows the strong connectivity to North America and Europe, in addition to domestic airports. Hong Kong demonstrates the strong intercontinental hub connectivity and Singapore the large hub connectivity to Oceania, Europe and within Asia. Jakarta, on the other hand, has specialized in domestic and Asian hub connectivity, in other words, connecting domestic airports and those in Asia. Meanwhile, the three Chinese airports in Mainland China demonstrate the impressive growth in hub connectivity to all geographical regions between 2001 and 2017. This indicates that these Chinese airports are quickly developing as hubs, though hub connectivity levels at these airports were rather low in 2001. In addition, hub connectivity via Tokyo/Haneda also increased drastically during the period of analysis. Overall, there is only small intercontinental hub connectivity at many of these airports.

2250 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

Domestic Domestic Domestic 8,000 2001 6,000 2001 500 2001 2009 5,000 2009 400 2009 Intercontinental 6,000 Within Asia Intercontinental Within Asia Intercontinental Within Asia 4,000 2017 2017 300 2017 4,000 3,000 200 2,000 2,000 1,000 100 Asia-Africa 0 Asia-Oceania Asia-Africa 0 Asia-Oceania Asia-Africa 0 Asia-Oceania

Asia-South Asia-South Asia-South Asia-Europe Asia-Europe Asia-Europe America America America

Asia-North Asia-North Asia-North America America America 1. Tokyo/Narita 2. Tokyo/Haneda 3. Osaka

Domestic Domestic Domestic 400 2001 5,000 2001 6,000 2001 2009 4,000 2009 5,000 2009 Intercontinental 300 Within Asia Intercontinental Within Asia Intercontinental Within Asia 4,000 2017 3,000 2017 2017 200 3,000 2,000 2,000 100 1,000 1,000 Asia-Africa 0 Asia-Oceania Asia-Africa 0 Asia-Oceania Asia-Africa 0 Asia-Oceania

Asia-South Asia-South Asia-South Asia-Europe Asia-Europe Asia-Europe America America America

Asia-North Asia-North Asia-North America America America 4. Nagoya 5. Seoul 6. Beijing

Domestic Domestic Domestic 4,000 2001 1,500 2001 4,000 2001 2009 2009 2009 Intercontinental 3,000 Within Asia Intercontinental Within Asia Intercontinental 3,000 Within Asia 2017 1,000 2017 2017 2,000 2,000 500 1,000 1,000 Asia-Africa 0 Asia-Oceania Asia-Africa 0 Asia-Oceania Asia-Africa 0 Asia-Oceania

Asia-South Asia-South Asia-South Asia-Europe Asia-Europe Asia-Europe America America America

Asia-North Asia-North Asia-North America America America 7. Shanghai 8. Guangzhou 9. Hong Kong

Domestic Domestic Domestic 2,500 2001 1,500 2001 6,000 2001 2,000 2009 2009 5,000 2009 Intercontinental Within Asia Intercontinental Within Asia Intercontinental Within Asia 1,000 4,000 1,500 2017 2017 2017 3,000 1,000 500 2,000 500 1,000 Asia-Africa 0 Asia-Oceania Asia-Africa 0 Asia-Oceania Asia-Africa 0 Asia-Oceania

Asia-South Asia-South Asia-South Asia-Europe Asia-Europe Asia-Europe America America America

Asia-North Asia-North Asia-North America America America 10. Taipei 11. Manila 12. Bangkok

Domestic Within Asia Domestic 8,000 2001 4,000 2001 2,000 2001 2009 2009 2009 Intercontinental 6,000 Within Asia Intercontinental 1,500 Within Asia 3,000 2017 2017 Intercontinental Asia-Oceania 2017 4,000 1,000 2,000 2,000 500 1,000 Asia-Africa 0 Asia-Oceania Asia-Africa 0 Asia-Oceania 0 Asia-Africa Asia-Europe Asia-South Asia-South Asia-Europe Asia-Europe America America Asia-South Asia-North Asia-North Asia-North America America America America 13. Kuala Lumpur 14. Singapore 15. Jakarta

Figure 8. Directional hub connectivity at primary Japanese Airports, 2001, 2009 and 2017

2251 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

4. SCENARIO ANALYSES FOR CHUBU CENTRAIR INTERNATIONAL AIRPORT

For an assessment of the model and its application, the paper conducts scenario analyses for Chubu Centrair International Airport on the connectivity impacts of an additional flight from this airport to large hub airports in Europe, North America and Asia, respectively, and of moving all domestic flights from Nagoya Airfield, the other airport in Nagoya, to this airport.

4.1 Dual Airport Systems in Nagoya Metropolitan Area

4.1.1 Chubu Centrair International Airport

Chubu Centrair International Airport is an international airport on a manmade island, 35 km south of Nagoya in the central region of Japan (see Figure 9). It is classified as a first class airport and is the main international gateway for the third largest metropolitan area in Japan with its population size of more than twenty million in the catchment area. The airport had reached its maximum capacity and currently processes around four million international passengers and six million domestic passengers, ranking 8th busiest in the nation. When Chubu Centrair opened in 2005, it took over almost all of Nagoya International Airport’s (now Nagoya Airfield) commercial flights. However, there were several withdrawals from Chubu Centrair after the airport commenced its operation. operated a route to Chicago for less than seven months in 2005. In 2008, after a few years of service from Chubu Centrair, several airlines cancelled certain flights, including ’ suspension of flight to Kuala Lumpur, Jetstar ending its operation, stopping its flight and ’ suspension of flight to San Francisco. Emirates and Hong Kong Express Airways left the airport in 2009, although the latter resumed its service from 2014. and EVA Air left the airport in 2013. V Air withdrew from the airport and ended its operation in 2016. Meanwhile, new international services started mainly to Chinese airports to service the influx of inbound tourists from China. Furthermore, AirAsia Japan, a Japanese low-cost carrier, launched in 2017 from its base at Chubu Centrair. A low-cost carrier terminal is scheduled to be completed at this airport in 2019.

Chubu Centrair International Airport

Figure 9. Location of Chubu Centrair International Airport and its aerial photo

2252 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

4.1.2 Nagoya Airfield

Nagoya Airfield, former Nagoya International Airport, served as the main international airport for Nagoya Metropolitan Area until the opening of Chubu Centrair in 2005. It is also known as Airport or Nagoya Airport. Figure 10 shows the location of Nagoya Airfield and its aerial photo.

Nagoya Airfield

Figure 10. Location of Nagoya Airfield and its aerial photo

During the 1980s and early , Nagoya International Airport was a busy airport because of the overflow from other international airports in Japan, New Tokyo International Airport (now Narita International Airport) and Osaka International Airport (). Since the opening of Kansai International Airport in 1994, the airport’s main traffic source has been the nearby automotive and manufacturing industries, causing carriers such as United Airlines to stop flying to Nagoya. In addition, the airport was hampered by its location in a residential area, limiting the number of flights, as well as the operating hours. Because of these reasons, a new airport, Chubu Centrair, was built on an island south of Nagoya. In 2005, nearly all of Nagoya International Airport’s commercial flights moved to Chubu Centrair. On the same day, the old airport became a general aviation and airbase facility, as well as was renamed to the current names. It is now a domestic secondary airport, which is the main hub for , the only airline that offers scheduled domestic services from this airport. Therefore, the domestic air services in Nagoya Metropolitan Area are split up between Chubu Centrair and Nagoya Airfield, which deteriorates the domestic network connectivity and hub competitive status of Chubu Centrair.

4.2 Connectivity Impacts of Additional Flights

In this section, scenario analyses for Chubu Centrair are conducted on the connectivity impacts of an additional flight from this airport to large hub airports: Frankfurt by , by and Kuala Lumpur by AirAsia.

4.2.1 Frankfurt

Table 1 shows the indirect connectivity impacts of an additional flight of Lufthansa from Chubu Centrair to Frankfurt by countries. In other words, this indicates the additional onward

2253 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

connectivity from Frankfurt by countries (see Figure 1). According to this table, one additional flight of Lufthansa from Chubu Centrair to Frankfurt increases the indirect connectivity to domestic destinations in Germany by 10.89 CNU from 32.54 CNU to 43.44 CNU. Other remarkable examples include Italy (additional 6.82 CNU), France (additional 4.03 CNU), United Kingdom (additional 3.95 CNU), Austria (additional 3.70 CNU), Switzerland (additional 3.04 CNU) and Spain (additional 3.00 CNU).

Table 1. Additional indirect connectivity from Chubu Centrair by countries (CNU) Country Before After Difference Argentina 0.28 0.37 0.09 Austria 11.08 14.78 3.70 Belgium 2.86 4.02 1.16 Bulgaria 0.12 0.16 0.04 Belarus 0.65 0.85 0.20 Canada 0.08 0.10 0.03 Switzerland 9.12 12.16 3.04 Czech Republic 1.41 1.88 0.47 Germany 32.54 43.44 10.89 Denmark 3.38 4.51 1.13 Spain 9.00 12.00 3.00 France 13.34 17.37 4.03 United Kingdom 10.91 14.86 3.95 Greece 0.72 1.00 0.29 Croatia 3.72 5.14 1.42 Hungary 1.62 2.16 0.54 Ireland 1.79 2.39 0.60 Israel 0.13 0.17 0.04 Italy 21.24 28.05 6.82 Luxembourg 1.53 2.04 0.51 Netherlands 2.94 3.93 0.99 Norway 1.37 1.85 0.49 Poland 7.31 9.75 2.44 Portugal 0.35 0.47 0.12 Russia 0.57 0.76 0.19 Sweden 2.99 3.99 1.00 United States 0.24 0.32 0.08

Form a hub connectivity perspective shown in Table 2, the strongest gains are found in Japan (additional 4.37 CNU), because connections between a Lufthansa flight and flights of are constructed at this airport, both of which are the Star Alliance members. In addition, hub connectivity at this airport to (additional 0.52 CNU), (additional 0.40 CNU), Hong Kong (additional 0.38 CNU) and China (additional 0.33 CNU) becomes larger by the connections between a Lufthansa flight and flights of the Star Alliance members. Table 2. Additional hub connectivity at Chubu Centrair by countries (CNU) Country Before After Difference China 0.98 1.31 0.33 Guam 1.21 1.61 0.40

2254 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

Hong Kong 1.14 1.52 0.38 Japan 13.10 17.46 4.37 0.34 0.46 0.11 Philippines 1.55 2.06 0.52 Singapore 0.46 0.62 0.15 Thailand 0.33 0.44 0.11 Vietnam 1.74 1.91 0.17

4.2.2 Detroit

Table 3 shows the indirect connectivity impacts of an additional flight of Delta Air Lines from Chubu Centrair to Detroit by countries. In other words, this indicates the additional onward connectivity from Detroit by countries. As shown in this table, one additional flight of Delta Air Lines from Chubu Centrair to Detroit increases the indirect connectivity mainly to domestic destinations in the United States by 24.92 CNU from 118.81 CNU to 143.72 CNU. Form a hub connectivity perspective shown in Table 4, the strongest gains are found in China, because connections between a Delta Air Lines flight and flights of the SkyTeam members, including and , are constructed at this airport.

Table 3. Additional indirect connectivity from Chubu Centrair by countries (CNU) Country Before After Difference Brazil 1.10 1.66 0.55 Canada 3.15 3.78 0.63 United Kingdom 0.17 0.20 0.03 United States 118.81 143.72 24.92

Table 4. Additional hub connectivity at Chubu Centrair by countries (CNU) Country Before After Difference China 12.93 16.01 3.09 South Korea 3.75 4.50 0.75 0.24 0.36 0.12

4.2.3 Kuala Lumpur

Table 5 shows the indirect connectivity impacts of an additional flight of AirAsia from Chubu Centrair to Kuala Lumpur by countries. In other words, this indicates the additional onward connectivity from Kuala Lumpur by countries. Currently, there is no direct flight between these two airports, so this scenario analysis suggests the impact on indirect connectivity (onward connectivity) of launching a new route to Kuala Lumpur by AirAsia Japan or the affiliate airlines of AirAsia, such as AirAsia X. As shown in this table, one flight of AirAsia from Chubu Centrair to Kuala Lumpur increases the indirect connectivity to Malaysia (4.69 CNU), Indonesia (2.99 CNU), Thailand (1.16 CNU) etc. As explained Section 4.1.1, AirAsia Japan launched out in October 2017 with a first flight from this airport to (), so there was no hub connectivity during the period of analysis (September).

Table 5. Additional indirect connectivity from Chubu Centrair by countries (CNU)

2255 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

Country Before After Difference Australia － 0.003 0.003 Indonesia － 2.99 2.99 India － 0.37 0.37 Cambodia － 0.35 0.35 Sri Lanka － 0.23 0.23 Myanmar － 0.36 0.36 Malaysia － 4.69 4.69 Singapore － 0.57 0.57 Thailand － 1.16 1.16 Vietnam － 0.27 0.27

4.3 Impacts on Connectivity of Moving Domestic Flights from Nagoya Airfield to Chubu Centrair International Airport

Another scenario analysis is furthermore conducted on the connectivity impacts of moving domestic flights from Nagoya Airfield to Chubu Centrair. This means that the domestic flights by Fuji Dream Airlines, the only carrier providing scheduled services at Nagoya Airfield, are moved to Chubu Centrair. The first observation is that the number of 154 direct domestic flights is added at Chubu Centrair, as shown in Table 6 by destination airports. To be more specific, six new destinations are added, including , Yamagata, Hanamaki, Izumo, Kochi and . Direct connectivity to , and increase by 35 CNU, 7 CNU and 14 CNU, respectively.

Table 6. Additional direct connectivity from Chubu Centrair by destination airports (CNU) Destination airport Before After Difference Aomori － 21.00 21.00 Fukuoka 84.00 119.00 35.00 Yamagata － 14.00 14.00 Hanamaki － 28.00 28.00 Izumo － 14.00 14.00 Kochi － 14.00 14.00 Niigata 14.00 21.00 7.00 Kitakyushu － 7.00 7.00 Kumamoto 21.00 35.00 14.00

As for indirect domestic connectivity, the total number of 5.46 CNU is added to New Chitose (4.40 CNU) and Fukuoka (1.06 CNU), as shown in Table 7. Table 8 shows them form a hub connectivity perspective, indicating that these additional indirect domestic connectivity are constructed at Yamagata (4.40 CNU) and Niigata (1.06 CNU).

Table 7. Additional indirect connectivity from Chubu Centrair by destination airports (CNU) Destination airport Before After Difference Aomori 1.19 1.19 0.00 Akita 1.99 1.99 0.00 New Chitose 16.50 20.91 4.40

2256 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

Fukue 6.96 6.96 0.00 Fukuoka 0.72 1.78 1.06 Hakodate 2.98 2.98 0.00 Ishigaki 14.71 14.71 0.00 Miyazaki 2.63 2.63 0.00 Memanbetsu 1.20 1.20 0.00 Miyako 5.58 5.58 0.00 16.38 16.38 0.00 Rishiri 0.09 0.09 0.00 Nakashibetsu 4.39 4.39 0.00 Tsushima 0.20 0.20 0.00 Wakkanai 6.12 6.12 0.00

Table 8. Additional indirect connectivity from Chubu Centrair by hub airports (CNU) Hub airport Before After Difference Akita 0.43 0.43 0.00 New Chitose 18.18 18.18 0.00 Fukuoka 17.29 17.29 0.00 Yamagata － 4.40 4.40 Hakodate 0.07 0.07 0.00 Haneda 36.48 36.48 0.00 Ishigaki 3.13 3.13 0.00 Niigata 0.02 1.08 1.06 Kumamoto 1.27 1.27 0.00 1.36 1.36 0.00 Matsuyama 0.51 0.51 0.00 Nagasaki 3.41 3.41 0.00 Narita 134.84 134.84 0.00 Naha 19.89 19.89 0.00 4.92 4.92 0.00

Regarding hub connectivity, additional 57.73 CNU are found in the domestic market, including the routes between Fukuoka and Yamagata (7.49 CNU), between Hanamaki and Fukuoka (6.78 CNU), between Kumamoto and Aomori (4.84 CNU) and between Kumamoto and Hanamaki (4.11 CNU). Among the domestic routes shown in Table 9, Chubu Centrair functions as a hub airport on the route between Niigata and Fukuoka (2.45 CNU), and hub connectivity on this route increases to 5.43 CNU after moving the domestic flights from Nagoya Airfield to this airport.

Table 9. Additional hub connectivity at Chubu Centrair by domestic routes (CNU) Domestic route Before After Difference Aomori-Izumo － 1.56 1.56 Fukuoka-Aomori － 2.95 2.95 Fukuoka-Yamagata － 7.49 7.49 Yamagata-Fukuoka － 3.85 3.85 Yamagata-Kochi － 2.55 2.55 Yamagata-Kitakyushu － 2.22 2.22

2257 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

Yamagata-Kumamoto － 2.11 2.11 Hanamaki-Fukuoka － 6.78 6.78 Hanamaki-Kochi － 1.22 1.22 Hanamaki-Kitakyushu － 0.96 0.96 Hanamaki-Kumamoto － 3.31 3.31 Kochi-Hanamaki － 1.75 1.75 Kochi-Niigata － 2.65 2.65 Niigata-Fukuoka 2.45 5.43 2.98 Niigata-Kochi － 1.37 1.37 Niigata-Kitakyushu － 1.17 1.17 Niigata-Kumamoto － 3.86 3.86 Kumamoto-Aomori － 4.84 4.84 Kumamoto-Hanamaki － 4.11 4.11

5. CONCLUSION

In this paper, we applied the Netscan connectivity model for the analysis of the network performance and hub competitive status of fifteen selected primary airports in Asia over the period from 2001 to 2017. The Netscan connectivity model measures the number of direct and indirect connections for each airport and weighs it for its quality in terms of transfer time and detour time. We classified network connectivity into four: direct, indirect, onward and hub. All connectivity is expressed in one indicator, connectivity units (CNU’s). The results revealed that the most striking connectivity growth took place at the three major airports in Mainland China, which are quickly developing into major hubs: Beijing, Shanghai and Guangzhou. The number of direct, indirect and also hub connectivity at these three airports increased at a much higher rate than at other airports in our sample. As for Shanghai and Guangzhou, opening of new international airports boosted network performance. All connectivity at Tokyo/Haneda increased drastically over these years analyzed, definitely owing to the resumption of international air services at this airport in 2010. On the contrary, other airports, such as Osaka, Nagoya and Taipei, experienced a deteriorating network performance during the period of analysis. The results presented here may be useful for airports in the assessment of their network performance as well as benchmarking their competitive hub status vis-à-vis other airports, as demonstrated in the scenario analyses in this paper. We have presented our results at a fairly aggregated level. The NetScan connectivity model allows for much more detailed analysis of an airports’ competitive hub status, including at the level of the geographical corridor (the Transpacific or Europe-Asia market, for example) or even at the individual market level. These analyses are, however, left for future research.

ACKNOWLEDGEMENTS

This research was subsidized by the Japan Society for the Promotion of Science (JSPS), Grant-in-Aid for Scientific Research (C) (Grant Number: 17K03688) and Grant-in-Aid for Scientific Research (B) (Grant Number: 17H02039), and Kansai Airport Research Institute.

REFERENCES

2258 Journal of the Eastern Asia Society for Transportation Studies, Vol.13, 2019

1. Airports Council International (2014) Airport industry connectivity report 2004-2014. ACI Europe. 2. Bania, N., Bauer, P.W., Zlatoper, T.J. (1998) U.S. air passenger service: A taxonomy of route networks, hub locations and competition. Transportation Research E, 34(1), 53-74. 3. Borenstein, S. (1989) Hubs and high fares: Dominance and market power in the U.S. airline industry. RAND Journal of Economics, 20(3), 344-365. 4. Brueckner, J.K., Spiller, P.T. (1994) Economies of traffic density in the deregulated airline industry. Journal of Law and Economics, 37, 379-415. 5. Burghouwt, G., Hakfoort, J.R., Ritsema-Van Eck, J. (2003) The spatial configuration of airline networks in Europe. Journal of Air Transport Management, 9(5), 309-323. 6. Burghouwt, G., Veldhuis, J. (2006) The competitive position of hub airports in the transatlantic market. Journal of Air Transportation, 11(1), 106-130. 7. Burghouwt, G., De Wit, J., Veldhuis, J., Matsumoto, H. (2009) Air network performance and hub competitive position: Evaluation of primary sirports in East and Southeast Asia. Journal of Airport Management, 3(4), 384-400. 8. Caves, R.E., Christensen, L.R., Tretheway, M.W. (1984) Economies of density versus economies of scale: Why trunk and local service airline costs differ. RAND Journal of Economics, 15(4), 471-489. 9. De Wit, J., Veldhuis, J., Burghouwt, G. Matsumoto, H. (2009) Competitive position of primary airports in the Asia-Pacific rim. Pacific Economic Review, 14(5), 639-650. 10. Ivy, R.J. (1993) Variations in hub service in the US domestic air transportation network. Journal of Transport Geography, 1(4), 211-218. 11. Kuby, M.J., Gray, R.G. (1993) The hub network design problem with stopovers and feeders: The case of Federal Express. Transportation Research A, 27(1), 1-12. 12. O’Kelly, M.E., Miller, H.J. (1994) The hub network design problem; A review and synthesis. Journal of Transport Geography, 2(1), 31-40. 13. O’Kelly, M.E., Bryan, D.L. (1998) Hub location with flow economies of scale. Transportation Research B, 32(8), 605-616. 14. Oum, T.H., Zhang, A., Zhang, Y. (1995) Airline network rivalry. Journal of Economics, 18(4a), 836-857. 15. Oum, T.H., Park, J.-H., Zhang, A. (2000) Globalization and strategic alliances: The case of the airline industry. Pergamon. 16. Pels, E. (2001) A note on airline alliance. Journal of Air Transport Management, 7, 3-7. 17. Shaw, S.-L. (1993) Hub structures of major US passenger airlines. Journal of Transport Geography, 1(1), 47-58. 18. Veldhuis, J. (1997) The competitive position of airline networks. Journal of Air Transport Management, 3(4), 181-188. 19. Zhang, A. (1995) An analysis of fortress hubs in airline networks. Journal of Transport Economics and Policy, 30(3), 293-308.

2259