Currently, tourism is one of the most important issues in Spain. During the year 2013, the number of journeys, excluding urban proximity displacements, reached the 140 million; tourism journeys, at 52% of these, constituted the highest percentage [1]. In this sense, the transport system plays an important role because, to an extent, the appeal of cities as tourist destinations is influenced by their accessibility through different transport networks [2, 3]. The choice of destination is in part influenced by the ways each city can be reached, which is mainly related to the transportation system (travel times, timetables, travel costs…). There are also considerations related to accommodations and other travel expenses.

There is a very large imbalance in the transportation mode distribution in Spain. Private vehicles are the clear winner; they involve around the 84% of journeys. Other transport modes, such as trains, buses and airplanes, are used in around 5% of journeys [1]. The clear superiority of car travel can be attributed to several reasons, such as the possibility of beginning the trip whenever desired, the ability to use the car for transport at the destination and its competitive travel costs, especially for travels as a couple or in groups. However, car users perceive travel costs incorrectly, usually underestimating them. It is common to consider only the fuel expenses, but private vehicles also incur other costs related to acquisition fees, car insurance and maintenance costs [4, 5]. Also, difficulties of parking the car at destination and its additional costs when using a secure parking lot may be considered as an important disadvantage of this mean of transport.

In the last year, RENFE (the Spanish rail operator) has started exploring recreational travel as an interesting new market in which the High Speed Rail (HSR) system should compete. Although the initial conception of HSR was primarily oriented to business travel, RENFE is reorienting its services to the tourism market in an attempt to compete against private vehicles.

RENFE’s commercial policies are focused on increasing the occupancy of all the trains by offering more economical ticket prices in order to attract travellers who would otherwise use other transport modes. In 2013, they introduced the new AV City service to expand traditional long-distance high speed train (AVE) service to off-peak hours. Fares were discounted up to 70% depending on the train (timetables and connections) and how far in advance the ticket was purchased. Since being launched in 2013, these policies have generated a 23% increase in the number of passengers and around a 12% increase in the trains’ occupancies [6].

Additionally, RENFE entered the tourist package market, offering the ‘Spain Pass’, which allows foreign tourists unlimited travel by HSR so they can visit several cities using the train. Another tourism product is the ‘Avexperience’ that offers train tickets between two cities served by HSR, hotel accommodations, and city cards for discounts on museums, transport, restaurants, etc. for weekend visits.

The present paper is focused on assessing how competitive the HSR and its new services for weekend tourism journeys are compared to private vehicles. Weekend trips represent around 55% of tourism in Spain [1]; therefore, evaluating their efficiency in terms of the money invested can contribute to simplifying the choice of an appropriate mode of transportation for each destination.

The arrival of HSR systems generates changes in mobility patterns in the cities involved. Ex-post research studies on the assessment of high-demand transportation corridors have demonstrated the variations in modal shares that have been produced by the opening of new HSR lines. These studies are mainly centred on the competition between air and rail transportation because of the initial conception of HSR as a potential alternative to air travel in routes of approximately 400-600 km [7, 8]. Some high-demand routes, such as London - Paris [9], Paris - Lyon, or Madrid - Seville [7], have experienced a remarkable modal substitution by transferring passengers from air travel to HSR services.

Nevertheless, HSR is currently not only an alternative to air travel but also to road transportation. The development of the HSR network, through the involvement of many different cities and the introduction of different types of services, makes it possible for HSR to also compete in the short- and medium-distance travel market. Literature on the topic has identified different ranges of travel times for which HSR can compete against other transportation modes and found a limit of 2 hours of travel time for road transportation and 2.5-3 hours for air transportation [10]. These short- and medium-distance HSR connections, generally between smaller intermediate cities and a main metropolis, have also displayed variations in modal shares and mobility patterns, experiencing a significant decrease in private vehicle demand in favour of HSR service [11]. In addition, the modal share of a specific route may be conditioned by certain characteristics of the travel, such as the final destination (arriving to or leaving from the destination city), trip length, use of luggage, and temporal restraints in terms of travel and arrival times for certain trip purposes [10].

These ex-post studies and the effects they discovered on the modal share based on the travel context lead to questions about the different factors that affect choices for modes of transportation from the perspective of the user. The studies centred on modal share assessments are concerned about the characteristics of the service that directly affect the traveller. They are usually carried out using logit models that represent the probability of a traveller choosing among several transport options. They consider different factors that influence the traveller’s choice, which are generally related to both socioeconomic and transportation factors [12, 13].

Factors such as the travellers’ age, gender, income, socioeconomic background or level of education may influence their choice of transportation mode. Some studies have analysed the effects of socioeconomic characteristics on modal choice and their variations across different trip purposes, such as business, tourism and commuting [14, 15]; and distances and lengths [16].

However, the most important factors included in the literature are those related to transportation itself, such as travel times, frequencies and costs [17, 18]. First, travel time could be considered as the most representative factor in the modal choice, as it could be a useful tool to identify the most convenient transport mode between origins and destinations. It also makes it possible to pre-establish what ranges of travel times within each mode of transport have the highest demand [19]. Regarding travel time, some studies have analysed the effect of shrinkage of space due to the arrival of new infrastructure in terms of travel time reduction as a measure of network accessibility [20]. In addition, in some means of transport, including air and rail transport, it is necessary to spend extra time in reaching the train station/airport, increasing the total travel time. That is why, apart from the travel time, another variable related to the access time has been introduced in some mode choice studies [18]. Access to stations is not only a relevant factor in terms of time spent but also in terms of travel comfort. Private vehicles allow a journey to start directly from home without concerns about how to reach the station or how far in advance it is necessary to be there. Additionally, travel cost is another relevant factor that affects transport mode preferences. Although it is closely related to socioeconomic characteristics, it can also be considered as an independent feature to compare the competitiveness of different transportation options. In this factor, access to a station also plays an important role because, in many cases, the money spent on reaching the station represents a high percentage of the total travel expenses. Finally, the third factor used in the literature is frequency. This could also be useful, as it is closely related to waiting times when passengers arrive at the station. However, it is more representative for means of transport such as urban buses or undergrounds where the number of services per day is very high. For transport modes such as the HSR system, it loses representativeness because of the lower frequency and important variations in the number of services depending on the time of the day. In those cases, the convenience of timetables becomes a key factor [21, 22], especially when travelling for a weekend, because people usually organise their journeys and purchase train tickets some time in advance [12]. Therefore, timetables are much more important than frequencies in this transport mode.

In the transport mode choice process, assessing the convenience of schedules, which will influence the time available at the destination, is an approach based on individuals’ needs that has its origins in the Time Geography principles [23]. This concept refers to the useful time a person has for doing certain activities through the available transport systems with consideration for the temporal restraints of daily life. The application of these principles was undertaken in 1970 by Törnqvist [24], who developed the concept of ‘potential contacts’, that is, cities that could be reached from a certain origin and allow for more than 6 hours at the destination. Coming back to this study, French researchers carried out an assessment of European transport networks in the framework of an ESPON project [25]. The application of the Time Geography principles to the HSR system was undertaken by Coronado et al. (2013) [26] by means of an efficiency assessment, which is used in this paper to evaluate transport mode choices with a focus on travellers’ needs. This efficiency assessment involves both the time available at the destination, using real timetables, and the associated costs, which will determine the real competitiveness of the available service supply.

The case study encompasses the 27 cities served by HSR that were included in the Spanish HSR network at the beginning of 2015. There are actually 30 HSR cities in Spain, including the Galician corridor, but the latter is not connected with rest of the network at this time (Fig. 1). Therefore, the assessment carried out in this paper takes into account all of the Spanish HSR connections with Madrid as their origin for travel over a whole weekend. The choice of Madrid is due to its location in the core of the radiocentric HSR network and its favourable service supply, which is the widest in the network. Additionally, the capital of the country covers almost 35% of the total population linked by the HSR system, so it supposes the biggest potential tourism market that can be attracted.

Currently, the Spanish HSR supply offers four main services:

- AVE: Long-distance HSR services with speeds of up to 310 km/h, in place since the HSR system was conceived in Spain.

- ALVIA: Long-distance HSR service with a top speed of 250 km/h that has the ability to use both Iberian and standard gauges, which allows it to utilise both high-speed and conventional rail lines in the same journey.

- AVANT: Regional HSR services that are slower than AVE services but have more economic fares and are aimed at commuters

- ‘AV City’ is a new service whose aim is to complete the HSR supply in off-peak hours in areas with no AVANT service. They offer competitive fares and normally stop in all cities on the lines on which they run.

In addition to these services and related to the new tourist-oriented fares, RENFE has introduced:

- ‘Promo’ fares: The ‘Promo’ ticket is available for AVE and AV City services and is based on a dynamic pricing system that allows for discounts of up to 70% depending on the train (type of train and timetables) and how far in advance the ticket is purchased. In fact, there are a limited number of tickets with this ‘Promo’ fare, so its availability is influenced by the demand and varies depending on the HSR line and destination. This fare is neither reimbursable nor exchangeable.

- ‘Mesa’ fare: This is intended for trips for families and small groups of 1 to 4 people and is available for AVE, ALVIA and AV City services. The ticket includes four seats facing each other in pairs. They are sold as a package with a 60% discount on the general rate. They can be occupied by fewer travellers if desired, though the total cost of the space is still charged.

Fig. (1). Spanish HSR network in 2015. Source: Authors.

This new arrangement of the Spanish HSR services and fares makes it interesting to carry out an assessment of the capability of HSR as an alternative transport mode for tourism travel. Therefore, the present paper is focused on the analysis of the HSR supply efficiency [26] for weekend tourism trips in comparison to private vehicles. To achieve this, a person-based assessment has been considered in order to introduce the points of view of travellers and not only the characteristics of the infrastructure.

The measurement of this efficiency (1) is based on the time available at destination (tdest) and the associated costs (€travel) following the formula [26]:

(1)

To compute this efficiency indicator, some hypotheses must be established in relation to the departure and return times (total time budget) in order to obtain the time available at the destination (Fig. 2) and the associated costs.

Fig. (2). Transport chain hypotheses for a weekend journey. Source: Authors (adapted from L’Hostis and Leysens, 2012)

For travel by car, the efficiency calculation is quite direct. Given a total time budget, the available time at the destination will only be reduced by the travel time (Fig. 2). According to the travel costs, an average consumption value is used. The expenses of travelling by car are related not only to the fuel but also to the proportional costs of maintenance, insurance, the initial purchase, etc. [4, 5, 27]. Including all of these expenses, travel costs will be: €travel = K (€/km) · D (km), where, for a certain private vehicle, K varies only depending on the fuel price .

The fluctuations in fuel prices make it necessary to assess how much they influence travel costs for a private vehicle. The variations in K for different fuel prices are presented in the following table (Table 1).

In the period between January 2010 and January 2015, fuel prices reached a maximum in September 2012 (1.52 €/L for gasoline and 1.44 €/L for diesel) and a minimum at the beginning of 2010 (approximately 1.1€/L). On average, fuel prices have fluctuated around 1.30 - 1.40 €/L in the last two years [28], which led to a value of 1.4 €/L (K = 0.304) being chosen for this research. However, an analysis of sensitivity will be carried out in order to check the influence of fuel prices on the main results of the paper.

When travelling by HSR, the time available at the destination will depend on the time budget, travel times, access and security times and, especially, the timetables. The efficiency will be higher when the timetables fit the established departure and return times. The access time is considered the time needed to travel from the HSR station to the city centre and vice versa using public transit (urban buses). A period of fifteen minutes was added at all of the stations to account for security checks.

For associated costs, the HSR ticket prices for different fares will be included. For both return and ‘Mesa’ fares, the ticket prices maintain their values; however, the ‘Promo’ fares utilise a dynamic pricing system, as previously mentioned. Fig. (3) displays the price evolution of the ‘Promo’ fares depending on the destination city, the time of purchase and the day of the week. It can be observed that the variations in the ticket prices are higher on Fridays (Fig. 3a) because the demand is influenced not only by weekend journeys but also by same-day business travel or even commuting. However, on Sundays (Fig. 3b), the ticket prices rise, in general, in relation to how far in advance the ticket was purchased, although there are also differences between destinations.

Fig. (3). ‘Promo’ fare evolution on Fridays (a) and Sundays (b) for purchases in October 2014 Source: Authors.

Having established the previous hypotheses, 3 different scenarios are considered (Table 2). They correspond to 1, 2 and 4 travellers, respectively, as the differences in efficiency depending on the number of travellers must be considered. Each of these scenarios are evaluated considering only the transportation costs in order to exclusively assess the influence of the transportation network and services and both the transportation and accommodation costs, which is a more realistic assumption for assessing a weekend tourism trip.

The Spanish rail operator, RENFE, has reoriented its service supply to compete in the tourism market. New services, fare reductions and tourist packages are some of the strategies that have been introduced. Until now, tourism travel has been clearly linked to private vehicle use, but this new arrangement is making HSR an interesting transportation alternative for weekend journeys.

The HSR system in Spain is becoming a very complex transportation option, and it has different properties between cities. Having an HSR station is not linked with being well connected, which is why the analysis of the service supply and real timetables seems so relevant. In general terms, the HSR connections have higher efficiencies than private cars for one and two travellers and rise to a balanced situation when four people travel together due to the availability of ‘Promo’ and ‘Mesa’ fares.

In the latter scenario, short-distance travel is slightly dominated by private vehicles because the travel expenses are reduced, especially for four people. However, for medium and long-distance travel, the HSR system becomes more competitive, except in those cases where the service supply is very deficient.

The methodology used in this paper is centred on a service supply analysis, considers the principles of Time Geography, and focuses on travellers’ needs. It is a powerful approach as it recognises the real constraints found when travelling using different transport systems. However, it presents some limitations concerning the previous hypotheses. For instance, some destinations benefit from convenient timetables or competitive fares, even when their service supplies are reduced, only because the departure or arrival time hypotheses are well adapted. Apart from that, it is necessary to underline that the HSR’s ‘Promo’ and ‘Mesa’ fares are not always available and that their competitiveness varies depending on the time of purchase. In this sense, the new tourism arrangement for Spanish HSR is very similar to the air travel pricing system, making it difficult to assess it in a categorical way. Additionally, it must be highlighted that the efficiencies assessment carried out in this paper is exclusively focused on the service supply analysis, but there are other factors, such as travel comfort, availability of a private car, and even the willingness to drive, that could influence the choice of transport mode. For instance, access to stations is a weak point of HSR in relation to private car because it increases the total travel times and expenses. Additionally, carrying luggage to the train station may not be comfortable. However, travelling by car has the drawback of needing to find parking close to the hotel and in a sufficiently safe area. Also, traffic congestions, especially in highly demanded holiday periods, may be a determinant negative factor when considering the car travel efficiency. All of these factors may tip the scale in favour of one or another mode of transportation when travelling for a whole weekend.

Besides, this new approach to addressing transportation mode analyses should be developed further, as it could become a useful tool for simplifying the choice of the most convenient transport mode for each destination. In future research, this could also be compared to the real demand for each option.