Keywords

1 Introduction

Peripheral island tourism destinations rely significantly for their development on airport infrastructure. For the majority of cases, airports are the only gateway these destinations have with the rest of the world. At the same time, airports are the first attribute of the destination actual visitors’ encounter and the last thing that would shape their post-trip memories of the destination [1, 2]. As a result, airports play a pivotal role in shaping the safety standards and profile of the destination to actual and potential visitors [3, 4].

The aim of the paper is twofold. First, to identify the key policy initiatives, and their configurations, that influence visitors’ preferences between different security and safety scenarios at an airport context. Aligned with this aim is the intention to calculate marginal willingness to pay (MWtP) estimates for specific aspects of safety and security policy initiatives. Second, the paper aims to provide a number of potential policy recommendations to policy makers, practitioners and airport managers. Since, security and safety are now major aspects of visitors’ perceived satisfaction at a destination, such policy recommendations could serve as potential feedback for securing a competitive advantage for the destination [3].

2 Literature Review

The linkage between airport infrastructure and safety issues has started to emerge very prominently in the literature lately due to the covid-19 pandemic [13], the resulting revenge travel fad [12] and the heightened concern for terrorist attacks and incidents [10]. Thus, safety and security nowadays is considered as a competitive feature of the destination [10].

The significance of individual preferences for evaluating different safety and security scenarios at an airport context is also depicted in the abundance of studies appearing in the literature in the last few years. More specifically, the first effort to establish a connection between individual preferences and safety/security within a transportation setting was by Rizzi and Ortuzar [9]. Albeit not focusing on an aviation setting and not using a SPDCM approach, yet it was the first effort to direct attention towards a preference-based safety/security policy formulation. Leo and Lawler [5] were the first ones to actually examine passengers’ perceptions about safety technology at an airport setting.

From there onwards, Robinson et al. [8] and Potoglou et al. [7], Veisten et al. [11], and Patil et al. [6] were critical at providing the impetus for the emergence of other like-minded studies in the field. At the same time, [3] provided the theoretical justification for the use of preference-based methodological instruments for the examination of aviation security at a wider scale. More recently, [4] have also explored the interplay between airport security and individual preferences.

3 Research Methods

3.1 Research Methodology

The current paper makes use of stated preferences discrete choice modelling (SPDCM) methods to evaluate individual preferences for alternative/future policy configurations regarding safety and security at an international mass tourism island airport/destination in Greece. The SPDCM experiment was based around a survey based questionnaire. Considering the fact that airports are a major factor on visitors’ perceived satisfaction levels [2], and that safety perceptions are deemed as a competitive advantage nowadays ([4, 10]), SPDCM facilitates capturing and measuring respondents’ preferences for safety and security at airports.

Research Methods. The current study collected empirical data from a SPDCM experiment conducted during the summer months of 2021. The study was conducted in person, following all the hygienic protocols that were in place due to the covid-19 pandemic at the time across Europe. Data was collected over a sample of 500 individual respondents (tourists) travelling to Crete for holiday purposes. Respondents were selected through convenience sampling. They were approached during their stay in Crete (the sample was collected from all four prefectures of the island), as well as from the airport during their departure from Crete (airport departure lounges). For the purposes of the analysis, a binary logit was estimated.

The dependent variable in our study involved asking individual participants to state their preferred setting when traveling by air (i.e., the most preferred airport setting from a security/safety perspective). What is more, each alternative setting was described in terms of seven (7) product or policy attributes, all related to safety and security processes within an airport. These policy attributes included the ‘type of security personnel at the airport’ (qualitative approach), the ‘type of CCTV cameras at the airport’, the ‘security processes’ at the airport, the ‘time required for a passenger to go through security checks’, the ‘number of visible security personnel at the airport’, ‘who has access to information collected at the airport’, and finally a ‘security surcharge (equivalent to the price coefficient). The process of selecting the attributes and their respective levels was informed by focus group discussions, and a small qualitative study that included various stakeholders (citizens, industry experts, as well as academics in the field).

Furthermore, each one of the abovementioned seven policy attributes was described based on four (4) levels or configurations. For example, the ‘type of security personnel at the airport’ attribute is described in terms of ‘airport security personnel (status quo, or base)’, ‘airport police department personnel’, ‘undercover police department personnel’, and ‘military personnel’. Random configurations of levels across all seven attributes differentiated one airport setting over another in a given choice scenario. Each choice scenario was generated randomly through a main—effects routine in SAS. The random generation of choice scenarios through the experimental design ensured that each scenario would appear only once in the whole choice experiment and that it would facilitate the trade—off between different attributes and their configurations. Table 1 presents a SPDCM (choice) experiment faced by a participant.

Table 1 Example of a SPDCM experiment
Full size table

4 Empirical Findings

4.1 Descriptive Analysis

Table 2 summarizes the samples socio-demographic profile. There is an even distribution between males and females in our sample, whereas there is a clear majority of late middle aged (46–65) respondents. More than half in the sample are married and in full time employment. As expected, the overwhelming majority of respondents were travelling either with their partner, or with the family (and kids). Finally, the majority of them originated from either Germany or the UK, and their reported net income was in the mid to high end range.

Table 2 Sample’s socio-demographic profile
Full size table

Econometric Analysis. The empirical findings from the stated preferences discrete choice modelling (SPDCM) experiment are summarized in Table 3. Table 3 presents the beta coefficients from the homogeneous preference specification. First, one could observe the negative and statistically significant ‘price’ coefficient. For the purposes of the choice experiment, the security surcharge takes up the price role in the SPDCM experiment, primarily in order to assist in the derivation of the marginal willingness to pay (WTP) estimates later on. The negative and statistically significant ‘price’ coefficient aligns with standard microeconomics theory. From this perspective, it also acts as a sign of internal validity for the SPDCM experiment.

Table 3 Beta Coefficients and marginal WtP estimates from the SPDCM experiment
Full size table

The empirical findings in Table 3 also indicate that “people” attribute, referring to the type of security personnel patrolling the airport seems to be extremely important for passengers/international visitors. In other words, a qualitative aspect of security seems to be particularly relevant on respondents’ preference patterns. Hence, in both cases where the police is assigned the patrolling of the airport (as part of the airport police department, or as undercover agents), visitors expressed positive and st. significant preferences. On the contrary, when the military is assigned the same role, then visitors expressed negative and st. significant preferences. In other words, respondents tend to perceive the presence of the armed forces in airports as (probably) an unwanted escalation that causes concern and distress among passengers.

The findings on Table 3 also indicate that respondents hold quite strong preferences regarding technology used in airport security. They exhibit negative and st. significant preferences for no cameras (a deterioration from the status quo) and at the same time, positive and st. significant preferences for cameras capturing personal biometrics information. These two points could be taken to imply that travelers are particularly sensitive about their security at airports nowadays, and the absence of cameras is perceived by many as a backtrack for their personal security. Briefly, the majority perceives that no cameras at all would pose a serious threat to their security at an airport. What is more, the estimated parameters indicate that respondents prefer CCTV cameras with advanced features (ability to capture bio information, over standard CCTV cameras. This is a finding supported in other studies as well [6].

On the other hand, respondents expressed positive and st. significant preferences for both a 50% decrease in personnel, as well as a 50% increase in personnel numbers in airports. Combined these findings may be an indication of heterogeneity in the sample. This is something that would have to be investigated at a later stage.

Finally, respondents expressed positive and st. significant preferences for both personal biometrics cards as a security measure/process at the station, and the disclosure of relevant information to all police departments worldwide. Compared to the status quo (random selection of passengers to go through metal detectors), respondents were very positive towards personal biometrics cards, hence they feel more comfortable with a substantial tightness regarding security processes at the station. At the same time, they also feel very strongly regarding the wider disclosure of information regarding suspicious passengers to all police departments worldwide. This finding could potentially have significant policy implications regarding the trade—off between security and privacy of individual passengers.

5 Policy Implications

The analysis of the empirical findings from the homogeneous preference specification indicates that respondents strong very positive and st. significant preferences for the type of CCTC cameras at an airport. They oppose any possible policy deterioration (no cameras) and at the same time, they support CCTV technology that would capture individual passengers’ biometric information. At the same time, they are also very positive towards security processes that record passengers’ biometric information. This is due to the higher degree of familiarity with this process, since, frequent flyers already have already registered their biometric information at a similar card [11]. Hence, familiarity and exposure seems to exert a strong influence on individual preferences. Combined the abovementioned findings indicate that visitors are very positive towards the initiation of technologies and processes that proceed much deeper into personal privacy issues, in order to enjoy greater levels of security. Second, the empirical findings would suggest that there is an apparent uniformity of opinions regarding qualitative aspects surrounding security personnel, but not so much regarding quantitative aspects of security personnel at airports. Thus, one major takeaway point from this is the high likelihood of preference heterogeneity in our sample that may call for further examination of individual preferences for safety at greater depth.

6 Conclusions

The paper adopts a bottom-up approach to evaluate individual preferences for alternative policy initiatives within an aviation context. In particular, the analysis utilizes a stated preferences discrete choice modelling experiment to evaluate respondents’ preferences for various aviation scenarios with different safety and security attributes. Data was collected through a survey-based instrument on a sample of 500 individuals and was analyzed using a binary logit model.

Assuming homogeneous preference specification in the model, overall, the empirical results indicate that respondents exhibit much lower inhibitions regarding the trade—off between privacy and security/safety when travelling. In particular, in the majority of cases, respondents were in favor of policy initiatives that would greater levels of security, but would diminish their individual privacy levels. This finding is contrary to the reported findings in the literature. At the same time, the empirical results indicate that preference heterogeneity could be introduced in the model (in order) to account for different preferences across individuals.