This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

Journal of Air Transport Management 14 (2008) 205 212 Contents lists available at ScienceDirect Journal of Air Transport Management journal homepage: www.elsevier.com/locate/jairtraman The geographical efficiency of Spain s regional airports: A quantitative analysis Francisco J. Tapiador a,, Ana Mateos a, Jordi Martí-Henneberg b a ICAM, University of Castilla-La Mancha, Avda. Carlos III s/n, 45071 Toledo, Spain b Department of Geography, University of Lleida, Plaza Victor Siurana s/n, 25003 Lleida, Spain article info Keywords: Spanish regional airports Low-cost carriers Data envelopment analysis abstract Previous studies of the role of regional airports in the Spanish economy have focused on calculating efficiency using air traffic as the output and airport infrastructure data as the input. We present an alternative quantitative analysis based upon the geographical structure within the airport s market basin. We formulate a geographical efficiency model that considers territorial variables within the airport hinterland. Input data for our model include the socio-economic structure of the population, intermodal transport links, industrial and tourism potential and existing leisure-related services. The output is the annual number of passengers associated with such constraints. The result is a relative efficiency estimate that shows uneven patterns of geographical efficiency amongst Spain s regional airports and also provides insights into opportunities for expanding some of these critical items of infrastructure. They suggest that some of Spain s regional airports may be better placed than others to compete in a liberalized market that exhibits a clear tendency in favor of coastal, tourism-based airports. & 2008 Elsevier Ltd. All rights reserved. 1. Introduction The airport sector in Europe is facing important changes associated with the liberalization of services and infrastructure. The slot allocation regime has been modified (EC Regulation 793/ 2004) to adapt to the current characteristics of the market, facilitate the entry of new companies, improve the use of airport capacities and establish quasi-market values for determining airport capacity (Madas and Zografos, 2006). EU policy governing the liberalization of air transport includes four principal areas: access to the market, control of capacity, tariffs, and the issuing of operating licenses. This policy began in 1980 and was implemented in several stages until 1997, when the transition period came to an end. As in other European countries, liberalization affected the Spanish transport system in many ways, with perceptible effects upon economic sectors such as tourism and trade, regional imbalances, and also social tendencies and cultural links. Within this context, an analysis of the efficiency of airports may help to increase competitiveness by identifying both the factors that may hinder future growth and the elements that ensure that airports work efficiently. Corresponding author. Tel.: +34 925 268 800. E-mail address: francisco.tapiador@uclm.es (F.J. Tapiador). The Spanish airport system is managed in an integrated, centralized way through Aeropuertos Españoles y Navegación Aérea (AENA), a publicly owned organization that depends on the Ministerio de Fomento and has full control over Spain s airports. In its role as airport manager, AENA assigns slots to airlines and manages the airports. This centralized system has been justified with arguments of territorial redistribution. It is claimed that the profits generated by the major airports can be used to compensate the losses of the smaller, less profitable ones, effectively subsidizing services to remote areas or to destinations that would not otherwise have enough market potential to attract private operators. Despite its public role, AENA s statutory regime allows this institution to work as a private company as far as its contractual and labor relations are concerned. In recent years, the organization has taken a more commercial direction and has established several commercial associations with the private sector (Martín and Roman, 2001). Despite its efforts to be more competitive, the operations of AENA must be analyzed in the context of its original role (Costas-Centivany, 1999). With the initial situation ending some time ago, the new framework, which originated with the emergence of the low-cost carriers (LCC), has generated a renewed interest in the way that the Spanish airport system is managed. Whether public control implies more or less efficiency remains open to debate (Oum et al., 2006): monopolies may be associated with a reduction in services, higher costs, lower levels of 0969-6997/$ - see front matter & 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.jairtraman.2008.04.007

206 F.J. Tapiador et al. / Journal of Air Transport Management 14 (2008) 205 212 maintenance and insufficient investment in services (Martín and Roman, 2001). The emergence of the LCCs has been the single most important event in recent times, not only due to its impact on the Spanish air business, but also on the European market. The growth of LCC air traffic (Barbot, 2006) has helped to revive and revamp secondary, regional airports, which had previously been maintained for purely strategic reasons, providing them with thriving business and bringing economic opportunities to their hinterlands. The reasons for most LCC choosing these regional airports are well known: their initial infrastructures were primitive, which meant that new developments could be customized to meet the LCC s needs; there were no problems of slot availability or congestion, thereby saving costs associated with delayed flights; and marginal costs were very low. The regional airports and their hinterlands have also benefited from the LCCs. The international situation has witnessed an increase in air traffic with changes in its patterns of demand (Forsyth, 2007). Airports are very important items of infrastructure for these territories as they play a very important role in the economic development of their respective communities (Sarkis, 2000). Almost overnight, areas with previously little appeal have been flooded with convenience travellers, some of whom make overnight stays in the local area. Similarly, local populations have been blessed with direct links to major economic areas such as London and Brussels. In the case of Spain, secondary airports such as Girona and Valladolid, and those serving other large cities, such as Valencia and Bilbao, have experienced a notable increase in the number of operations, mostly associated with the activity of lowcost companies (Table 1). The impact of low-cost companies has also been notable at airports competing with the high-speed train (HST), such as Seville and Zaragoza, even after taking into account the opportunities for co-operation and intermodal potentials of the two modes (Givoni and Banister, 2006). Airports whose access to important economic centres has been improved by the arrival of the HST, such as Valladolid and Santander and their connections with Madrid, have also experienced an increase in airport traffic. Given this growth, international air transport requires modern infrastructure with sufficient capacity to meet anticipated future demands (Martín and Román, 2001). This study relates the market potential of Spain s regional airports to their current traffic and allows us to identify the resources that have been best exploited by these airports and those that currently remain underexploited. One potential benefit of this analysis is the possibility that it offers to link territorial policies (such as industrial location and the development of complementary infrastructure) to the needs of airports, thereby strengthening the role played by secondary airports within their respective regional economies. Another area of application relates to the criteria that LCC use to select new airport destinations. While Warnock-Smith and Potter (2005) concluded that access to low-cost services was the single most important factor governing decisions related with the selection of regional airports by these companies, other researchers have shown that the characteristics of their respective hinterlands also constitute a very important factor (Guillen and Lall, 2004). LCCs target secondary airports that are attractive in terms of their regional economies, population resources and any other characteristics that could be attractive for potential passengers, including proximity to beaches and tourist resorts. This paper provides an analysis on the potential of Spain s secondary airports taking all these factors into account and conceptualizing them as geographical constraints. Below, we describe the model that we formulated and used and explain the input data and different variables considered in the analysis. In the following section, we analyze the results obtained from the model. Finally, we discuss the potential consequences of our findings in terms of improving the efficiency of Spain s airports and examine the potential consequences of the foreseen liberalization of the airport management scenario. Table 1 Low-cost operations and passengers in 2005, and growth in traffic for the period 1995 2005 for Spanish airports Number of low-cost flights (28 LCCs) in 2005 Number of passengers in 2005 (%) Growth in traffic 1995 2005 (%) (National average: 74.11%) Palma mallorca 37,283 21,218,897 44.0 Barcelona 19,523 26,941,215 129.7 Màlaga 16,862 12,591,501 70.2 Alicante 14,614 8,768,730 124.9 Girona 10,252 3,513,612 541.5 Valencia 6999 4,633,490 159.6 Madrid 6683 41,560,552 108.2 Murcia 3872 1,409,701 1490.9 Tenerife Sur 3046 8,320,291 31.7 Ibiza 2791 4,130,806 22.4 Gran Canaria 2741 9,561,300 21.35 Bilbao 2691 3,831,362 139.2 Sevilla 2353 3,491,796 140.4 Reus 2137 1,351,608 180.1 Jérez 1660 1,222,280 225.1 Lanzarote 1602 5,309,005 35.6 Almería 1260 1,062,226 49.6 Santander 1049 644,046 247.8 Fuerteventura 709 3,996,672 61.4 Valladolid 702 435,993 145.7 Zaragoza 659 375,936 41.7 Granada 613 854,089 130.6 Santiago 395 1,798,941 49.5 The other 16 airports 1237 Data source: AENA and Instituto de Estudios Turísticos. 2. Data Spain s current airport network includes 47 airports; 40 of these offer purely commercial services, while the other 7 operate as joint civil-military airports. This study is limited to major airports in mainland Spain (Fig. 1). This selection was motivated by the different socio-economic contexts of the two island territories and the autonomous cities of Ceuta and Melilla with respect to mainland Spain. As an ultra-peripheral region of the UE, the Canary Islands have their own specific characteristics and like the Balearic Islands they constitute a traditionally active tourist destination. Ceuta and Melilla airports also represent a rather special case due to the extremely small geographical size of these two cities and their respective territories. We first selected the variables to be included in the model. The criteria were that the values should be objectively estimated, mathematically well-defined (i.e. with no singularities), and should also be available for the whole national territory. Our a priori choice therefore included population, European resident population, a leisure-related services activity index, an economic activity index, a commercial activity index, an industrial activity index, a tourist activity index, the length of railway (km), the length of roads (km) and an estimate of intermodality (the length of motorways/dual carriageways, railways and roads). The demographic variables (population and European foreign population, including both EU and non-eu residents) were obtained from official statistics provided by the National Institute of Statistics (INE). Data relating to communications infrastructure, such as main roads and railways was obtained from the National

F.J. Tapiador et al. / Journal of Air Transport Management 14 (2008) 205 212 207 Fig. 1. The Spanish airport system. Aerodromes are not considered. Geographic Institute (IGN). All the data used were available at disaggregated, municipality level. The indexes were calculated using data from the Economic Yearbook of Spain (2006) published by La Caixa, and also based on municipal data from Caja España for municipalities with fewer than 10,000 inhabitants. In the case of leisure-related services, the index reflected the relative weights of restaurants, cafeterias, bars, etc. in given areas compared with national levels. This index value reflected the relative weight of restaurant and bar activities for the whole of Spain. The base used equated the value of taxes (in euros) collected in Spain to 100,000 units. The industry index, which includes construction, was calculated in a similar way. The economic and professional activity index was based on tax income corresponding to all forms of economic activity (industrial, commercial and services), and therefore included all forms of economic activity except that relating to the agricultural sector. Finally, the tourist index was based on the income associated with the number of rooms available and annual occupation of hotels, motels, hotel-apartments, inns, boarding houses, guest houses, camp sites and apartments. All the indexes were based on elaborated using data relating to the tax on economic activities (IAE) levied on the corresponding economic sector; they did not take into account the black economy (Table 2). Once the data had been collected, we calculated the integral of each variable corresponding to municipalities located within 75 km of each airport. We then analyzed covariance between the different variables to select the ones with the greatest correlation with the number of passengers using Spain s airports during the same period, as provided by AENA. The results are shown in Table 3. We decided to include any variables with a correlation higher than 0.60 in the model. This implied excluding two variables from the analysis: the integrals of the lengths of the railway and road networks. Intriguingly, while the lengths of the railways and motorways were not individually correlated, variable accessibility (which was simply the sum of the length of the railways, motorways and roads) was well correlated with airport traffic. Table 2 Ranking of Spanish airports in terms of the number of passengers in 2006 Airport Passengers (2006) Airport Passengers (2006) 1 Madrid-barajas 45,530,010 22 Vigo 1,187,730 2 Barcelona 30,008,152 23 La palma 1,175,328 3 Palma de mallorca 22,408,302 24 Fgl granada-jaen 1,086,221 4 Malaga 13,076,252 25 Almeria 1,055,545 5 Gran canaria 10,286,635 26 A coruña 1,014,780 6 Alicante 8,893,749 27 Santander 649,447 7 Tenerife Sur 8,845,668 28 Valladolid 457,618 8 Lanzarote 5,626,337 29 Zaragoza 435,887 9 Valencia 4,969,113 30 Pamplona 375,309 10 Ibiza 4,460,141 31 San sebastian 368,009 11 Fuerteventura 4,424,880 32 Melilla 308,313 12 Tenerife norte 4,025,601 33 Vitoria 173,607 13 Bilbao 3,876,062 34 El hierro 171,444 14 Sevilla 3,870,600 35 Leon 126,648 15 Girona 3,614,223 36 Badajoz 80,464 16 Menorca 2,690,992 37 Logroño 55,427 17 Santiago 1,994,519 38 La gomera 38,846 18 Murcia-san javier 1,645,886 39 Salamanca 29,308 19 Reus 1,385,157 40 Ceuta 22,127 20 Jerez de la frontera 1,381,560 41 Cordoba 19,568 21 Asturias 1,353,030 42 Albacete 17,520 Data source: AENA. The empirical basis of this study is presented in Table 4. The first column includes passengers in 2006 and the other columns present integral values for the geographical variables considered. 3. Methods We analyzed the efficiency of the Spain s airports by applying a geographic variant of data envelopment analysis (DEA, Charnes et al., 1978; Srinivas Talluri, 2000; Charnes et al., 1995). As shown by Seiford (1994), this kind of analysis has been widely used in economic applications and has proved particularly useful in

208 F.J. Tapiador et al. / Journal of Air Transport Management 14 (2008) 205 212 empirical studies requiring non-parametric approaches. Previous airport-related research on this topic evaluated the efficiency of airport infrastructure by focusing on capacity (Oum et al., 2003; Bazargan and Vasigh, 2004; Martin-Cejas, 2002; Lin and Hong, 2006; Martin and Roman, 2001). The underlying hypothesis in these studies was that the improved efficiency could be achieved by improving the airport infrastructure itself. We follow a complementary approach and used the term geographical efficiency to refer to how efficiently an airport benefits from its location. We assumed that this geographical efficiency is linked to certain key characteristics of the size of an airport s catchment area, such as population, level of economic activity, accessibility or tourism potential. Some of these variables, such as population, are linked with the traffic from the airport, whereas others, such as the tourism potential, account for potential trips to the airports. As some of these characteristics can be improved through public and private initiatives such as promoting science parks or improving accessibility, there is evident interest in such analysis for planners. The actual formulation is a DEA model: maxðs l i y ia Þ i s:t: S o j x ja ¼ 1 j S k l k y ki S o j x ji p0 j 8i l i ; o j p0 8i; j (1) where x i represents the geographical parameters of each airport a; y i, i ¼ 1 is the productivity of the airport in terms of passenger volume per year (a value which, in our case, had a correlation of 0.9 with the number of operations carried out); and l and o are the weights. The technical issues of DEA modelling are sufficiently well-known to readers of this journal and so do not need to be repeated here. Even so, three short notes are due: the system is solved by first calculating the dual of Eq. (1); cross-efficiency methods were also used to avoid sub-optimal points; and we formally defined geographical efficiency as the ratio between the location assets of the airport and the traffic generated. Table 3 Correlations of the a priori variables considered in the efficiency model Code in Tables 4 and 5 Variables (integral values for a 75 km hinterland) 1 Leisure services index 0.84 2 Total population 0.83 3 Economic activity index 0.81 4 Commercial activity index 0.80 5 Tourism activity index 0.77 6 Intermodality (Motorways+Road+Rail) 0.66 7 European resident population 0.65 8 Industrial activity index 0.63 Not used Length of railways 0.44 Not used Length of national roads 0.02 Correlation (R 2 ) 4. Results and discussion DEA modelling results are shown in Table 5. The table shows the extent to which each airport benefits from the geographical resources within its hinterland. A zero value indicates full exploitation of this resource. Values above zero indicate the degree to which the resource is wasted. The last two columns, respectively, represent the overall efficiency index whether the scale of the airport is considered or not (equivalent to variable returns to scale: VRS) and constant returns to scale (CRS) models (Banker et al., 1984). Individual conclusions can be derived from Table 5 for each airport. The results allow us to cluster the airports into four groups on the basis of their respective efficiencies: (1) optimal, Table 4 Empirical basis of the efficiency model Passengers in 2006 Var 7 Var 2 Var 8 Var 4 Var 1 Var 3 Var 5 Var 6 A coruña 1,014,780 7044 1,316,610 3067 3141 3293 2746 1624 933 Albacete 17,520 8944 5,84,993 1098 1136 956 948 283 1301 Alicante 8,893,749 2,57,367 2,901,148 6080 6604 6688 6034 6061 1533 Almeria 1,055,545 33,807 6,69,349 1101 1343 1325 1241 1810 532 Asturias 1,353,030 8045 1,076,183 2651 2439 3034 2114 1331 1274 Badajoz 80,464 3904 6,75,122 1058 1487 1308 1096 450 945 Barcelona 3,000,8152 1,31,566 5,575,866 16,501 14,324 12,915 15,518 9913 2347 Bilbao 3,876,062 27,285 2,408,224 8347 5118 4825 6992 2243 1443 Cordoba 19,568 6078 1,229,217 2040 2339 2025 1728 860 1353 Granada-jaen 1,086,221 50,959 2,164,680 2601 4884 4391 3459 2966 1191 Girona 3,614,223 1,19,004 4,399,335 11,808 12,089 10,582 12,612 10,269 1460 Jerez 1,381,560 17,395 2,181,625 3633 5670 5059 4943 5919 1044 Leon 1,26,648 6831 7,01,056 1577 1259 1676 1219 604 1370 Logroño 55,427 25,018 1,434,333 5781 3273 2901 4304 1469 1871 Madrid 45,530,010 2,02,918 6,549,492 11,418 14,941 15,573 17,359 10,823 2582 Malaga 1,307,6252 1,41,868 1,786,105 2147 4092 4110 3246 6906 938 Murcia 1,645,886 1,63,858 2,621,602 5298 6031 5965 5340 2648 1351 Pamplona 3,75,309 29,054 1,724,891 6717 4056 3578 5198 1866 1343 Reus 1,385,157 57,389 1,813,241 6591 3580 3454 4292 3210 1928 Salamanca 29,308 4574 6,11,914 993 1081 1398 1029 813 1247 San sebastian 3,68,009 17,875 1,263,041 4538 2718 2417 3623 1175 955 Santander 6,49,447 12,543 1,510,881 3941 2842 3376 3627 2045 1193 Santiago 1,994,519 15,078 2,300,714 4611 5365 5586 4512 2732 1524 Sevilla 3,870,600 13,551 2,302,896 3104 4933 4283 4083 3650 1345 Valencia 4,969,113 93,921 2,897,606 7428 6866 6385 6793 3127 1562 Valladolid 4,57,618 11,067 7,71,109 1740 1452 1756 1415 568 1398 Vigo 1,187,730 12,875 1,504,174 2844 3440 3320 2818 1954 1029 Vitoria 1,73,607 28,856 2,497,374 8881 4871 4948 6716 2078 2120 Zaragoza 4,35,887 24,707 1,105,276 3210 2744 3196 2772 1068 1530

F.J. Tapiador et al. / Journal of Air Transport Management 14 (2008) 205 212 209 Table 5 Exploitation of geographical resources and efficiency index both regardless of scale (next-to-last column) and considering scale (last column) for a perfect competition scenario Var 7 Var 2 Var 8 Var 4 Var 1 Var 3 Var 5 Var 6 Efficiency Scale efficiency A coruña 0 0 0 0 0 0 0 0 1 0.525 Albacete 0 0 0 0 0 0 0 0 1 0.015 Alicante 136,702 28,9126 1977 691 662 0 0 100 0.512 0.967 Almeria 0 0 0 0 0 0 0 0 1 0.259 Asturias 0 0 1057 92 955 223 105 228 0.998 0.658 Badajoz 0 0 0 0 0 0 0 0 1 0.085 Barcelona 0 0 0 0 0 0 0 0 1 1 Bilbao 0 756974 5743 1455 1355 3517 0 237 0.673 0.907 Cordoba 0 342,193 699 398 310 212 0 355 0.026 0.496 Granada-jaen 0 587,893 86 1359 842 0 135 0 0.139 0.821 Girona 0 1,073,686 6152 4584 2811 4212 3995 0 0.168 0.837 Jerez 0 5,34,220 141 1812 1078 1368 3694 0 0.399 0.771 Leon 0 54,759 484 0 376 122 0 167 0.490 0.162 Logroño 0 1,81,294 3666 490 215 1589 0 693 0.012 0.827 Madrid 0 0 0 0 0 0 0 0 1 1 Malaga 0 0 0 0 0 0 0 0 1 1 Murcia 1,16,509 7,01,036 2013 1714 1653 780 0 0 0.168 0.881 Pamplona 0 2,55,058 4290 760 378 1961 0 179 0.064 0.891 Reus 0 2,23,358 3927 89 0 748 395 515 0.166 0.814 Salamanca 0 0 0 0 0 0 0 0 1 0.025 Sansebastian 0 3,00,299 2940 588 394 1709 0 0 0.156 0.578 Santander 0 1,76,122 1969 0 714 1159 206 0 0.259 0.782 Santiago 0 4,49,480 1999 1362 2038 1117 0 285 0.549 0.924 Sevilla 0 0 0 0 0 0 0 0 1 1 Valencia 20110 1,07,6987 139 891 244 0 0 159 0.475 0.892 Valladolid 0 55,833 471 0 414 127 0 148 0.526 0.364 Vigo 0 1,28,284 143 273 168 0 212 0 0.494 0.740 Vitoria 0 9,22,684 6337 1358 1574 3314 0 931 0.029 0.901 Zaragoza 1316 76,058 1343 580 1151 602 0 134 0.128 0.758 Note: 0 indicates inefficiency and 1 indicates the theoretical maximum efficiency. (2) airports whose efficiency could be improved, (3) airports for which both scale and efficiency could be improved, and (4) airports that are efficient, but lacking in scale. The optimal airports would be those capable of efficiently benefiting from all of the resources available. They would operate at maximum level to extract the full potential of their population resources and also those associated with tourism, industry and available services. The airports whose efficiency could be improved are those of a suitable scale to allow further improvements in their performances, possibly associated with specific policies aimed at making better use of currently wasted resources. The airports with poor scales and levels of efficiency would therefore need an impulse in both directions, which is not always possible, while the airports that are effective but of limited scale could present deficiencies from the point of view of how they take advantage of scarce resources. Table 6 summarizes the four categories. According to this scheme, Malaga and Seville are examples of optimal regional airports (Madrid and Barcelona are used as the reference frontier for the model). The efficiency of the airport of Malaga is explained by the large number of passengers using this airport with respect to its geographical constraints. The airport is ranked fourth amongst Spain s airports in terms of its volume of traffic, mainly due to the attraction of the Costa del Sol resorts for foreign tourists. It is similarly possible to explain the efficiency of Seville airport, which has almost 4 million passengers per year, due to its function as the capital city of Andalusia and the benefits associated with its large population and dynamic economy. Vitoria, Pamplona, Logroño and Murcia are examples of airports that could improve their performances. They are all scale-efficient, but only take advantage of some of their resources. Vitoria airport is an example of an airport that has greatly benefited from the LCCs. Even so, its efficiency remains relatively low, with our analysis revealing that it only operates satisfactorily with respect to its European foreign population and tourist potential. The same applies to the airports of Pamplona and Logroño, whose performances are poor compared with their demographic potential. Murcia airport, on the other hand, is scaleefficient, but only extracts full potential from its tourist resources and intermodality. In recent years, this airport has experienced a continuous growth in traffic, even so, its resident European population resource seems to remain underexploited despite the airport having become a fundamental node for holiday and leisure transport in the local region. Then there are efficient airports with scale deficits, which take full advantage of their limited resources but lack scale. This is the case, for example, of the airports of Albacete, Salamanca, Badajoz, Almeria and La Coruña. Albacete is the airport with fewest passengers per year of all Spain s airports. Even so, it could be considered as efficient, since it takes full advantage of its limited resources despite its small scale. The airports of Salamanca and Badajoz present similar cases, with low passenger numbers and activity mainly centered on charter flights. The fourth category corresponds to airports that could improve in terms of both efficiency and scale. This is the case of such airports as Córdoba, San Sebastián, Valladolid and Leon. The charter flight volume of Córdoba does not exceed 20,000 passengers per year, which is below its potential. This airport does not take advantage of its own, local population. The airports of Valladolid and San Sebastián carry each more than 350,000 passengers per year, but also lack efficiency and scale with differences in their underexploited variables. San Sebastián airport does, however, seem to benefit from the intermodality of its immediate area. These results invite the question of what would happen if the inefficient airports were capable of realizing their full potentials

210 F.J. Tapiador et al. / Journal of Air Transport Management 14 (2008) 205 212 Table 6 Summary of the efficiency factors involved, with a quantitative indication for each element Airports Efficiency Scale efficiency Working at optimum efficiency Efficiency can be improved Scale and efficiency can both be improved Efficient but lacking scale A coruña 1 0.525 Albacete 1 0.015 Alicante 0.512 0.967 Almeria 1 0.259 Asturias 0.998 0.658 Badajoz 1 0.085 Barcelona 1 1 Bilbao 0.673 0.907 Cordoba 0.026 0.496 Granada-jaen 0.139 0.821 Girona 0.168 0.837 Jerez 0.399 0.771 Leon 0.490 0.162 Logroño 0.012 0.827 Madrid 1 1 Malaga 1 1 Murcia 0.168 0.881 Pamplona 0.064 0.891 Reus 0.166 0.814 Salamanca 1 0.025 San Sebastian 0.156 0.578 Santander 0.259 0.782 Santiago 0.549 0.924 Sevilla 1 1 Valencia 0.475 0.892 Valladolid 0.526 0.364 Vigo 0.494 0.740 Vitoria 0.029 0.901 Zaragoza 0.128 0.758 within their respective scale parameters. This could be the case in a perfect competition scenario in which each airport was individually managed. This hypothesis can be simulated as a CRS model. Within such a scenario, airport managers would design policies to attract both locals and passengers from abroad, employing a wide range of strategies to avoid wasting resources. Targeted marketing strategies could be used to combat lowseason diseconomies in tourist areas, while business strategies could be devised to provide tailored-made solutions and improve the level of regional economic specialization. Another important issue relates to the strategy for coupling public and private interests: the low-cost model drives airlines to negotiate contracts. This significantly reduces aeronautical charges from the airport, which must then seek to make up for this shortfall by obtaining commercial revenue from increased passenger numbers (Francis et al., 2004). The results offered by Oum et al. (2006) for large airports show that privately owned airports have significantly higher profit margins (56%) than airports with other type of property/management structure, despite their smaller aeronautical taxes. The result of this modelling is shown in Fig. 2, which simulates the potential growth of Spain s airports in an ideal case under perfect competition. Further examination of the data reveals that there are airports that do not offer too many expectations due to their scale problems (these include the airports of Albacete, Badajoz, Salamanca and Leon). On the other hand, Córdoba and Pamplona airports could both potentially increase their number of passengers by one order of magnitude. The values indicate that the current air traffic of 128,731,401 passengers could ascend to 200,212,775: an increase

F.J. Tapiador et al. / Journal of Air Transport Management 14 (2008) 205 212 211 Fig. 2. Comparison of the number of passengers in 2006 and the estimated flow for a perfect competition scenario. 5. Conclusions Fig. 3. Comparison of the number of passengers in 2006 and the estimated flow for a perfect competition scenario, but excluding Madrid and Barcelona airports. Note: correlation estimates are included. of 55.5%. However, this great potential is concentrated in just a few key airports. While a qualitative appraisal of the situation could yield similar results, the quantitative results shown here allow us to make a precise estimation of the factors involved, including a ranking of the elements that need to be improved in order to increase efficiency. Moreover, as shown in Fig. 3, there is a nonlinear relationship between the current number of passengers and airport potential for a perfect competition scenario, with a weak R 2 linear correlation. This reveals the added value provided by modelling with respect to a qualitative estimate based upon present-day traffic estimates, as suggested by DEA theory. The results obtained from our modelling indicate that Spain s regional airports have different geographical constraints and that it may therefore be necessary to design individual strategies in order to unleash their full potential. The possibility of a single manager being able to coordinate both local and regional needs is a matter open to debate. However, individual management strategies could prove more efficient for exploiting available resources, while competition between neighboring airports may also encourage the development of new market strategies. Our findings support the hypothesis that individual management may help to unleash the latent potential of many of Spain s regional airports. Simulations for a perfect competition scenario predict a sharp increase in the volume of operations linked with improved exploitation of local resources and the possibility of developing specialized strategies to attract new passengers. We also found an uneven distribution of this potential with greater possibilities for airports serving Mediterranean areas. These airports tend to be more attractive to private investors and public policies aimed at liberalizing management should take this in account. Within the current framework in which airports are considered as strategic elements within their respective territories, we propose a cluster strategy, in which the private sector should be allowed to manage not only individual airports but a mixture of high and low potential airports: this could help to mitigate existing and foreseeable regional disparities. Acknowledgments This research was partially funded by the research Grants FP-005 Water, Road and Rail 2005SGR01089. FJT acknowledges research Grants PAI06-0102-7466, CGL2006-03611 and the Ramon y Cajal program. JMH research was also partially funded by the research Grants SEJ2004-02824 and SEJ2007-64812. The usual disclaimers apply.

212 F.J. Tapiador et al. / Journal of Air Transport Management 14 (2008) 205 212 References Banker, R.D., Charnes, A., Cooper, W.W., 1984. Some models for estimating technical and scale in sciences in data envelopment analysis. Management Science 30, 1078 1092. Barbot, C., 2006. Low-cost airlines, secondary airports, and state aid: an economic assessment of the Ryanair Charleroi airport agreement. Journal of Air Transport Management 12, 197 203. Bazargan, M., Vasigh, B., 2004. Size versus efficiency: a case study of US commercial airports, 2003. Journal of Air Transport Management 9 (3), 187 193. Charnes, A., Cooper, W.W., Rhodes, E., 1978. Measuring the efficiency of decision making units. European Journal of Operational Research 2, 429 444. Charnes, A., Seiford, L., Cooper, W., Lewin, A., 1995. Data Envelopment Analysis: Theory, Methodology and Application. Kluwer Academic Publishers, Norwell, MA, USA. Costas-Centivany, C.M., 1999. Spain s airport infrastructure: adaptations to liberalization and privatization. Journal of Transport Geography 7 (3), 215 223. Forsyth, P., 2007. The impacts of emerging aviation trends on airports infrastructure. Journal of Air Transport Management 13, 45 52. Francis, G., Humphreys, I., Ison, S., 2004. Airports perspectives on the growth of low-cost airlines and the remodeling of the airport airline relationship. Tourism Management 25, 507 514. Givoni, M.Y., Banister, D., 2006. Airline and Railway Integration Transport Policy 13, 386 397. Guillen, D., Lall, A., 2004. Competitive advantage of low-cost carries: some implications for airports. Journal of Air Transport Management 10, 41 50. Lin, L.C., Hong, C.H., 2006. Operational performance evaluation of international major airports: An application of data envelopment analysis. Journal of Air Transport Management 12 (6), 342 351. Madas, M.A.Y., Zografos, K., 2006. Airport slot allocation: from instrument to strategies. Journal of Air Transport Management 12, 53 62. Martin-Cejas, R.R., 2002. An approximation to the productive efficiency of the Spanish airports network through a deterministic cost frontier. Journal of Air Transport Management 8 (4), 233 238. Martin, J.C., Roman, C., 2001. An application of DEA to measure the efficiency of Spanish airports prior to privatization. Journal of Air Transport Management 7, 149 157. Oum, T.H., Adler, N., Yu, C., 2003. A comparative analysis of productivity performance of the world s major airports: summary report of the ATRS global airport benchmarking research report 2002. Journal of Air Transport Management 9, 285 297. Oum, T.H., Adler, N., Yu, C., 2006. Privatization, corporatization, ownership forms and their effects on the performance of the world s major airports. Journal of Air Transport Management 12, 109 121. Sarkis, J., 2000. An analysis of the operational efficiency of major airports in the United States. Journal of Operations Management 18, 335 351. Seiford, L.M., 1994. A DEA bibliography (1978 1992). In: Charnes, A., Cooper, W., Lewin, A.Y., Seiford, L.M. (Eds.), Data Envelopment Analysis. Theory, Methodology and Applications. Kluwer Academic, Boston. Srinivas Talluri, 2000. Data envelopment analysis: models and extensions, Decision Line, May 2000. Warnock-Smith, D., Potter, A., 2005. An exploratory study into airport choice factors for European low-cost airlines. Journal of Air Transport Management 11, 388 392.