Terrorist attacks and risk management implications in the
aviation industry
Master's Thesis
Cristina Cătălina Fogoroș
S2910047
Supervised by Dr. Melih Astarlioglu Co-Assessed by Dr. A. Visscher
Faculty of Economics and Business MSc. International Business and Management
University of Groningen
JUNE 20, 2016
ACKNOWLEDGEMENTS
I would like to thank all of those who helped me during the process of working on and completing the thesis. I am most grateful to my supervisor, Dr. Melih Astarlioglu for helping me by providing valuable comments and constructive feedback on my work. I would like to thank the participants. They shared their time and thoughts with me, and without them, this project would not have been possible. Similarly, I would like to thank my classmates and friends with whom we often discussed how to improve the quality of our papers. The moral effect was significant.
Last but not least important, I owe more than thanks to my family which includes my parents, my mother, grandmother and my clever cousin, Elvira in particular, for their support and encouragement throughout my life and especially during my college life.
Cristina Cătălina Fogoroș
ABSTRACT
The events of September 11th, 2001 have proved what a major threat terrorist attacks represent for aviation. In particular, the attacks demonstrated how civilian aircrafts can and have been used as weapons of mass destruction, killing hundred of civilians and destroying ground facilities. The tragic events highlighted vulnerabilities of the civil aviation system and led to significant changes in the design, provision and financing of aviation security at a global level. Terrorist attacks continue to play a crucial role in risk assessment and resource allocation to reduce potential threats. Security measures also have intangible costs such as privacy implications of security screening or discrimination issues.
This thesis aims to address the underlying issue of how officials in charge with security on airports perceive risk and what are the implications regarding passenger discrimination and security investment decisions. Investigation will be carried on Frankfurt International Airport and Bucharest's Henri Coanda International Airport.
Regardless of the complexity of the risk assessment methodology, due to the uncertainties associated with assessing risk in a dynamic counterterrorism context such as an airport, understanding how officials in charge with security personally perceive risk is necessary because ultimately, officials take decisions with further implications on air safety. A combination between an airport's profile in terms of number of previous attacks, their consequences, security system endowments, passenger traffic and security official's experiences (Juliusson, Karlsson, & Gärling, 2005), beliefs (Acevedo & Krueger, 2004) and expectations (Marsh & Hanlon, 2007; Nestler & von Collani, 2008; Stanovich & West, 2008) lead to personalized frameworks for selecting risk criteria. These have further implications on resource allocation and passenger discrimination.
This research aims to aid risk analysis and policy-making by providing a basis for understanding public responses to airport's discrimination behavior as well as officials' perception of risk. This work assumes that the understanding of how people think about and respond to risk is critical for the implementation of effective policies.
Table of Contents
1.0 INTRODUCTION ... 1
1.1 21st Century aviation industry: a challenge for risk management & security strategy ... 1
1.2 Objective of the thesis ... 2
1.3 Thesis structure ... 3
2.0 LITERATURE REVIEW AND PROPOSITION DEVELOPMENT ... 4
2.1 Risk management and the perceived level of risk ... 4
2.2 Security measures may restrict freedom, invade privacy and create discrimination ... 8
2.3 People or technology? Cognitive skills or X-ray? Preferred means of enhancing security ... 10
3.0 METHODOLOGY ... 14
3.1 Discussing qualitative and quantitative research ... 14
3.2 Reliability and validity: verification of qualitative and quantitative methods ... 15
3.3 Selecting the proper research method ... 16
3.4 Research Approach ... 16
3.5 Cross-Case analysis: an alternative approach ... 17
3.6 Data collection ... 17
3.7 Sampling ... 19
3.8 Conducting the interviews ... 20
3.9 Analysis ... 20
3.10 Conceptual Model ... 22
4.0 RESULTS AND FINDINGS... 23
4.1 Perception of risk ... 23
4.1.1 Testing proposition P1 a ... 24
4.1.1.1 Level of Acceptability for each scenario ... 25
4.1.1.2 Relation between acceptability, risk perception and number of incidents ... 26
4.1.2 Testing proposition P1 b... 26
4.2 Discrimination degree as perceived by passengers ... 28
4.2.2.1 Discrimination perception of major ethnic categories ... 32
4.2.2.1.1 Independent Sample t-test on discrimination as perceived by White/Caucasian passengers in Bucharest and Frankfurt Airport ... 33
4.2.2.1.2 Independent Sample t-test on discrimination as perceived by Asian passengers in Bucharest and Frankfurt Airport ... 34
4.2.2.2 Discrimination perception of major religious categories ... 35
4.2.2.2.1 Independent Sample t-test on discrimination as perceived by Christian passengers in Bucharest and Frankfurt Airport ... 36
4.2.2.2.2. Independent Sample t-test on discrimination as perceived by Muslim passengers in Bucharest and Frankfurt Airport ... 37
4.3 Investment decision-making... 38
4.4 Word Repetition analysis ... 40
5.0 DISCUSSIONS ... 44
5.1 How is risk associated to attacks perceived? ... 44
5.2 Are passengers discriminated? ... 45
5.3 What drives investment decision-making? ... 46
6.0 LIMITATIONS AND FUTURE RESEARCH ... 48
6.1 Research limitations ... 48
6.2 Future research suggestions ... 49
7.0 CONCLUSION AND IMPLICATIONS ... 51
References ... 53
1. INTRODUCTION
1.1 21st Century aviation industry: a challenge for risk management and security strategy
In the past several decades, trade, technology, and economic growth have created an age of globalization in which people, firms, and nations become ever more interconnected. Civil aviation is not an exception to the rule. This industry has evolved from a highly regulated system of national airlines, and government operated airports to a much liberalized global industry (Gillen & Morrison, 2015). Nowadays, according to ATAG (2014), the aviation sector supports 58.1 million jobs worldwide (through direct, indirect, induced and catalytic impacts) and contributes with $2.4 trillion (3.4%) to global GDP. It enables 3.1 billion passengers per year to travel over 5.4 trillion kilometers. Thus, it can easily be considered the most powerful engine for world economic growth, employment, trade links, tourism and support for sustainable development.
nature of risk can not be fully known, the terrorist activity being particularly difficult to assess as it is an intentional catastrophe that can occur anywhere, at any time, repeatedly (Ineichen, 2006). Thus, overall, airports' primary objective concerning risk management is to reduce different risks to an acceptable level.
One of the paper's main points is investigating the way in which "acceptability", regarding risk, can be quantified. What is an acceptable risk for one airport might as well be seen as an unacceptable risk for another; what influences their behavior when it comes to assessing the degree of risk? Could antecedents, such as the number of previous incidents and experienced events shape the dimension of risk tolerance? A second issue investigates whether risk management measures might restrict freedom, invade privacy and create passenger discrimination. Finally, a third subject discussed in this research takes issue with airports security investment decisions and tries to identify whether airports are more inclined to invest in human capital or technology. For the purpose of this paper, the terms "incidents" and "attacks" will be used to refer to the sum of false alarms, terrorist threats, shootings, and other such events which could endanger passenger, staff and airport integrity and safety.
Having explained the development of the thesis, the following research question is going to guide and center the current research:
1.2 Objective of the thesis
According to the research question, the main objective of the thesis is to investigate how the term "risk" is assessed by officials in charge of security within two European airports followed by the impact on passenger discrimination and security-related investments. The risk might be evaluated differently by different people. The way officials assess the term might be connected to antecedents, such as the number of incidents, their
frequency, and consequences, thus, such experiences will be taken into consideration. As the study is conducted with Romania's Henri Coanda International Airport (also known as Bucharest International Airport) and Germany's Frankfurt International Airport, it will be interesting to compare results from two distinctive airports, regarding security endowments, the number of occurred incidents, passenger traffic, etc.
1.3 Thesis structure
2. LITERATURE REVIEW AND PROPOSITION DEVELOPMENT
In order to grasp both the theoretical and practical scope in this paper, one has to firstly understand concepts such as: risk management, discrimination and investment decision-making in the current aviation environment. First, theories and previous research on these key components of the thesis will be elaborated. Subsequently, propositions will be elaborated, by combining the theoretical framework with the matters which are to be investigated throughout this study.
2.1 Risk management and the perceived level of risk
According to the Global Terrorism Database, within the period 1997-2007, there were five bombing attacks on European terminals, inflicting zero fatalities. These types of incidents are regarded as extremely low risks and are considered "acceptable" based on a fair degree of agreement about acceptable risk (Stewart & Melchers 1997). However, terrorism is a hazard where risk acceptability is not only a matter of fatality. Literature argues that patterns of risk evaluation are difficult to elaborate as there are numerous factors that might change the degree of risk acceptability from "acceptable" to "unacceptable" such as loss of airport reputation and customer confidence, a shift in airport preferences among passengers, etc. (Deodatis et. al., 2013). These levels of risk are established based on the risk's three components: threat, vulnerability, and consequences as shown in Figure 1. The threats to a target can be measured as the probability that a particular target is attacked in a particular way during a particular period. The vulnerability can be measured as the probability that damage occurs when exposed to a threat. Damages could include fatalities, injuries, property damage, or other similar consequences (Willis et. al., 2005).
Figure 1. Risk acceptance matrix (Kolesár & Petruf, 2012)
acceptable risk. On the other hand, in cases of a security system failure as a result of a probable scenario taking place and causing a catastrophic outcome, it is considered an unacceptable risk (Kolesár & Petruf, 2012).
At the individual level, knowledge, experience, values, attitudes, feelings and personal beliefs all influence the thinking and judgement of people about the gravity and acceptability of risks. These elements might lead to an individual becoming more sensitive and alert concerning future events or adopt a more habituated behaviour. Moreover, people use different mental models and other psychological mechanisms to judge risks, all being embedded through social and cultural learning and constantly moderated (reinforced, modified, amplified or attenuated) by the media, other people's influences and similar communication processes (Morgan, Fischhoff, Bostrom, & Atman, 2001).
Literature argues that what people perceive as a potential threat affecting their well-being and how they evaluate probabilities and magnitudes of unwanted consequences is a sum of attitudes, social influences, as well as cultural identity (Douglas & Wildavsky, 1983). The risk field is a patchwork open to different schools and perspectives. Psychological risk perception research, for instance, discovered that contextual factors (Breakwell, 2007) shape individual risk estimations and evaluations (Covello, 1983; Renn, 2008). Factors such personal ability to influence risks or familiarity with the incident to name a few provide valuable information about the elements that individuals use to build up their interpretation of risks.
Figure 2. Risk perception theories at individual level (Campbel Institute, 2014)
From a philosophical perspective, there are two main approaches to the study of risk perception, the realist approach and the constructivist approach (Renn, 2008). The first could be described as an attempt "to bring perception as close as possible to the objective risk of an
activity or an event". It implies that up to a specific degree risks can be recognized and
more inclined to displaying a more vigilant and sensitive behavior (Grill-Spector, Henson & Martin, 2006) towards such future probable attacks.
The matters discussed above lead us to our first proposition:
2.2 Security measures may restrict freedom, invade privacy and create discrimination
What is discrimination? According to United Nations, it is the unequal treatment on the basis of race, sex, language or religion. Adopted in 1948, the Universal Declaration of Human Rights, enlarged the list and included color, political affiliation, various opinions, national or social origins, marital or family status, disability or pardoned convictions (United Nation Report on the World Social Situation, 1997). Thus, we can just assume that discrimination arises from the social categorization of individuals (Taylor, Fiske, Etcoff, & Ruderman, 1978), a natural and spontaneous process that occurs in our everyday perception (Macrae, Bodenhausen, Milne, Thorn, & Castelli, 1997). Literature argues that categorization is driven by desires for simplicity (Ford & Stangor, 1992; Macrae, Milne, & Bodenhausen, 1994); for our comfort, we wish, as much as possible, to view individuals within the same category as similar.
How is discrimination related to airport security? As the dangerous possibility of a terrorist on board a plane continues to plague air travel, it clear that the issue of a proper security system within airports is of paramount criticality. However, achieving the right balance between increased safety and avoiding a discriminatory regime towards particular
P1 a: The higher the number of experienced incidents, the lower the acceptability rate
and the higher the perceived level of risk, displaying a more sensitive behaviour towards potential future attacks.
P1 b: If officials in charge of security are given to assess a scenario similar to an
passengers has proved to be a true challenge. The most common side‐effect of enhanced airport security measures includes stereotyping, the stigmatization of an entire ethnic or religious group. For instance, coupling "terrorism" to "Islam" in a nation's collective memory creates the presumption of collective sins. Early theorists such as Allport (1954) suggested that “it seems a safe generalization to say that an ethnic label arouses a stereotype which in turn leads to rejective behavior.” It has ben argued (Allport, 1954; Lippman, 1922) that stereotypes play a central role in social perception. For example, stereotypes are believed to ease the burden of social perceiver in responding to a complex social environment sometimes difficult to assess (Allport, 1954). More contemporarily, Macrae et al., (1994) defined stereotypes as a means to "simplify perception, judgment, and action".
suggests, selects passengers for an additional inspection. Civil liberties groups have cited the Selectee list for violating privacy rights and the potential for racial and ethnic discrimination.
Considering that more and more cases of discriminatory behavior have been reported within the aviation industry, one of this paper's objectives rests in analyzing whether the number of previous incidents experienced by an airport, which could lead to an increased level of security measures, has a negative effect on how ethnic/religious minorities are treated, thereby violating non‐discrimination standards.
The matters discussed above lead us to our second proposition:
2.3 People or technology? Cognitive skills or X-ray? Preferred means of enhancing security
Individual investment behavior is concerned with a person's choices regarding various purchases so as to satisfy one's interest (Nofsinger & Richard, 2002). According to the traditional economic theory, people are believed to be rational individuals who make investment decisions objectively to take advantage of the available opportunities. However, there is a broad literature which argues that their emotional inclinations driven by psychological biases as well as previous experiences define how they perceive the world and how they make decisions. In 2011, Greg B. Davies, Head of Behavioural and Quantitative Investment Philosophy at Barclays Wealth, argued that behavior finance offers considerable long lists of psychological biases to which we are all likely to fall victims to. Buying high and selling low, getting emotionally attached about investing in specific areas or failing to deploy the finances where they are most effective, are just a few examples. However, the bright side P2: There will be a positive correlation between the level of an airport's discriminatory
behavior as reported by passengers, and the number of previous incidents experienced by that airport.*
of this approach explains how experience is the key element that drives humans away from repeating the same mistakes and more carefully assessing the investment decisions and directions (Sagi & Friedland, 2007). Thus, previous experiences prove to be a benchmark for further investment decisions. Hence, the same can be assumed for the financial investment decisions on airports. This paper will analyze whether airports financial investments are directed towards human capital or advanced screening technology and what might explain these decisions.
Airports started implementing current threat detection systems at the beginning of the 1960s, in response to increased hijackings of commercial aircrafts. As shown in Figure 3, in the following decades regulators kept on adding layers of new security measures starting with pre-board screening in the middle of the 60s until the well-known full-body scanners introduced in the first decade of the new century. Similarly, new screening measures were introduced in 2006 to counter potential issues with explosives in the form of liquids, aerosols or gels (Wong & Brooks, 2015). Throughout the entire time, however, human operators continued to play a key role in the detection process by examining X-ray images, resolving metal detector alarms, conducting body scans with metal detection wands, conducting physical searches of baggage, and maintaining order at screening checkpoints (Harris, 2002). According to existing literature, investing in either advanced technology or human operators hold both advantages and disadvantages.
Figure 3. Evolving threats (Wong & Brooks, 2015)
impressive technological achievements. Tests have demonstrated that computer models are capable of making valid personality judgments by using digital records of human behavior. In their research, Youyou, Kosinski, & Stillwell (2015) show that computer-based models are significantly more accurate than humans in a core social-cognitive task: personality judgment. They claim that growth in both the sophistication of the computer models and the amount of digital footprint might result in computers outperforming humans even more decisively. The Realistic Accuracy Model (D. C. Funder, 2012; 1995) reveals a couple of major advantages that computers have over humans. First, machines have the capacity to store a tremendous amount of information, which, in return, is hard for humans to store and access. On airport security, a wide number of factors may contribute to the determination of risk, including information such as passenger type, previous mitigation of risk during the journey or any other known relevant information about the passenger's background, history, behavior. Literature argues that such information is better stored and evaluated by machines rather than human operators. Second, computers use consistent algorithms that increase the judgmental accuracy, while humans are subject to various emotions, hence their results can be biased (Vazire & Carlson, 2011). Based on such evaluations of machinery supremacy over humans, we would expect that to enhance security airports should invest more in advanced technology as it is more reliable than human judgement when it comes to a potential threat.
On the other hand, history has witnessed significant machine failures throughout time. For example, in 2015, an inspection found that Transportation Security Administration full-body scanners failed to detect explosives and weapons 96% of the time (Fishel, Thomas, Levine & Date, 2015). This issue creates a tremendous gap within the American airport security system. Moreover, similar failure cases were reported among European airports: in 2014, Frankfurt airport, Europe's fourth-busiest airport, handling over 60 million passengers per year has imposed manual searches of passenger bags after an undercover European Union inspection uncovered that dangerous weapon and other prohibited items managed to pass through security undetected 50% of the time, proving the inefficiency of the X-Ray scans (Huggler, 2014). The efficiency of advanced technology has therefore been questioned.
the human factor in the design of airport security jobs. Human perceptions have the advantage of being flexible and able to capture many subconscious cues unavailable to machines. Harris (2002), defends his perspective by adding that operators have unique capabilities needed to detect potential threats, such as pattern recognition, abstract reasoning, spatial visualization, and cognitive flexibility and intuition. As a consequence, a necessary skill of human operators which a machine could not replace rests in the ability to identify signals such as anxiety, panic, stress among passengers who could be potential attackers. Although humans face various limitations themselves, a scanner, no matter how advanced, in the absence of cognitive skills could not recognize if a passenger, carrying a weapon is blinking faster than normal or is sweating due to stress.
On the other hand, limitations such as the design of the specific operator tasks, limited absorptive capacity of potentially relevant information, the environment in which the tasks are performed, the selection and training of workers or even the mood in which the operator finds himself on a particular day will be critical to the realization of effective performance or will lead to failing in identifying a risk.
The matters discussed above lead us to our third proposition:
P3 a: Airports will tend to invest in both advanced technology and human operators but
argue that human operators are the centre of successful data analysis and threat identification.
P3 b: Airports will tend to invest in both advanced technology and human operators but
3. METHODOLOGY
Although risk management analysis incorporates data from various airports, none has analysed the difference in the degree of risk tolerance between airports which have already faced a specific attack and those which have not. The level of risk acceptance might prove to be very different between the two, as long as for some airports the incident has previously occurred, while for others it remains just a possible scenario. Moreover, the number of previous attacks might also influence the airport's perceived level of risk. It is interesting to investigate whether a higher number of past events will lead to a higher risk perception than others with fewer past incidents or vice-versa. Further implications of the perceived level of risk can be reflected in the preference for either investing in human capital or advanced technology, depending on which one has proved to be more efficient in identifying a threat. Also, according to the specificity of the occurred attack, a higher rate of discriminatory sentiment could be encountered within the airport.
Thus, the aim of this study is to investigate how officials in charge of security evaluate risk followed by the impact on investment decision-making and passenger discrimination. This chapter will explain in detail how the research was conducted. First, the choice of both quantitative and qualitative research will be discussed. Subsequently, the sampling procedure is described, followed by the means of data collection and analysis.
3.1 Discussing qualitative and quantitative research
people and some phenomena, or that social processes can not be measured by numbers, through statistics. In such situations, the use of quantitative methods might limit the chances of discovering certain aspects of these phenomena (Smith, 2006).
On the other hand, qualitative methods do emphasize the use of numbers or statistics but rely on understanding, interpreting and closely observing the data in its natural setting with a sort of insider view (Ghauri et al., 1995). Qualitative research can be defined as “any research that produces findings not arrived at using statistical procedures or other means of quantification” (Strauss & Corbin, 2008). The focus, therefore, lies on an in-depth understanding of words, opinions, and experiences. Moreover, qualitative methods are more concerned with the individual than the general (Mayring, 2003). Although backed up by a theoretical framework, it is not the theory which should be guiding the study but the data (Taylor & Bogdan, 1998). In qualitative research, the main goal is exploring and understanding a certain case and context rather than obtaining generalizable results (Bryman, 2008).
3.2 Reliability and validity: verification of qualitative and quantitative methods
There is criticism concerning research methods. By far, the major concerns are based on the aspects of validity and reliability (Golafshani, 2003), two different measurement instruments that illustrate the level of trustworthiness and credibility of a study.
On the other hand, with respect to quantitative research, the definitions of reliability and validity must follow two rules: firstly, to ensure reliability, the result should be replicable by using the same analysis method and secondly, the study is considered valid, as long as the means of measurement are accurate and they are actually measuring what they are intended to measure (Golafshani, 2003).
To overcome the limitations of both these research methods the reader should be able to access the concrete observations made, not only summaries or generalizations. Following Silverman (1993), these suggestions were considered in this thesis by voice-recording and transcribing the interviews as well as including direct quotes from these transcripts into the analysing part of the study.
3.3 Selecting the proper research method
The choice of research methodology depends on the nature of the research question. To address the research question of this study, a mixture of qualitative and quantitative methods were used. This includes two methods of data collection: 1) questionnaires with passengers; and 2) semi-structured interviews with officials in charge of security within airports.
3.4 Research Approach
3.5 Cross-Case analysis: an alternative approach
Due to the small sample used in this research, namely two European airports, a cross-case analysis has been chosen as an alternative approach to reveal further relevant findings. According to Bryman and Bell (2007), studying a case study implies an intense and detailed analysis of a single or few cases where the complexity of the nature of the case is studied.
Yin (2003) argues that the case study allows the researcher to analyze complex social interactions to uncover "inseparable variables" that are elements of the phenomena, such as the antecedents in this thesis. Moreover, the essence of a case study lies in discovering why a decision was taken and what were the results (Schramm, 1971). The purpose of the case study is not to generalize the findings to other cases or larger populations (Bryman & Bell, 2007) but to enable the researcher to make a comparison and contrast the findings from two different cases as well as to consider what is common and what is unique across them.
3.6 Data collection
Ghauri et al. (1995) state that collecting data can be either primary or secondary. According to Bryman and Bell (2007), primary data is the information that the researcher gathers on their own, conducting interviews, tests or handing questionnaires. Contrary, secondary data encompasses data such as literature, documents, and articles that are not collected by the researcher themselves but by other researchers and institutions (Bryman & Bell, 2007). In this thesis, both primary and secondary data are collected. After collecting the secondary data from books, scientific articles, and internet sources, the information has been critically analyzed to obtain a better insight of the security system within airports and to construct both the theoretical and the methodological part of the thesis. Primary data was gathered through interviews and questionnaires. The four collection phases took place as follows:
and collecting questionnaires from passengers who frequently use or transit one of the mentioned airports. During phase four semi-structured interviews with representatives of the security departments from the European airports chosen for this study were conducted. These interviews were followed by a round of emails and phone calls for further clarifications.
The first and last set of propositions were addressed through the use of an in-depth interview, a “conversation with a purpose” (Bond & Dexter, 1971). The collected qualitative information was used to get a better insight into the airport's position concerning risk assessment (P1), security investments (P3) and even discriminatory behavior towards passengers, further on compared with passenger's perceptions with respect to this issue (P2), investigated using quantitative research. Interviews are a suitable method because they allow asking open-ended questions to a small sample and exploring individual experiences or opinions regarding the researched phenomenon. In qualitative research, interviews tend to be more flexible. Semi-structured interviews, such as the ones used for this study mostly focus on the interviewee’s opinion and experience, aiming to get rich and in-depth data (Bryman, 2008). Thus, as Pole and Lampard (2002) suggest, they resemble conversations, trying to deal in-depth with the individual case. In a study with a small sample size, like this thesis, the focus is to obtain valuable information which can subsequently be used to comparing the cases. Furthermore, the interview structuring made sure that important theoretical issues were covered in the conversation, which facilitated the analysis according to the research questions. Moreover, the fact that the interview was not completely structured permitted the conversation to develop in a non-constraining way. A full transcription of each interview was produced.
The second set of research questions was addressed by using questionnaires. A questionnaire enables responses to be gathered from large numbers relatively quickly and cost-efficiently (De Vaus, 1991; Fink, 1995). Each questionnaire encompassed a descriptive rating, Likert-type scale which was used to collect quantitative data from 120 passengers within both studied airports (Appendix F). This stage was especially laborious as it is forbidden to hand in questionnaires within the airport facility without a specially-issued
3.7 Sampling
The research shall be limited to the European aviation industry while two European airports will represent the sample. The segment of the population that was selected for this study, in other words, the interviewing sample, is based on a non-probability approach which means that airports have not randomly been chosen (Bryman & Bell, 2007). The airports selected for the research are the results of the willingness to analyze and compare two very much distinctive security systems and their impact on investment decisions and passenger discrimination.
Precisely, the sample is represented by Frankfurt International Airport and Henri Coanda International Airport. These two airports have been particularly chosen as they differ in terms of number and gravity of incidents, security endowments, and passenger traffic. The interviews were conducted with Security Departments representatives from these airports. A timeline of 10 years, starting with March 31st, 2006 until March 31st, 2016 has been chosen for counting the occurred incidents. Officially, five incidents have been reported from Henri Coanda International Airport and 10 for the German airport (Appendix A). www.globalincidentmp.com, an official acknowledged website created to provide the general public, law enforcement, military and governmental organizations with real-time terrorism and security-related incidents around the world was used to collect this data. Also, 120 transiting/departing/arriving passengers within these airports represented the questionnaire sample. This random sampling process was utilized to ensure an extensive coverage of nationalities, races, genders, etc.
3.8 Conducting the interviews
Before conducting the interviews, existing literature was intensively studied. The gained insights were used to construct the open questions in a manner that would ensure the researcher with a deep understanding of the concept of risk and how it is measured. It encompasses a total of 10 questions out of which the first five are given scenarios evaluated on a scale from 1 to 7 while the rest are open questions. The interview script can be seen in Appendix D.
The interviews were conducted between April 6th and April 26th, 2016. Both interviewees were approached through different channels, such as phone call and email. Prior to starting the interview, the participants took the first 5 to 10 minutes to discuss the researcher's background, the reasons for conducting the study, and what was hoped to be achieved. This discussion allowed the researcher to explain the purpose of the study and to discuss issues including confidentiality, voluntary participation, the option to withdraw at any point and details on the university’s approval of my research. At this point, the permission to record the conversation was asked for, under the promise that the transcripts will only be made available for the thesis supervisor. All interviews have been conducted in English and were voice recorded. The meetings took place at the respondents’ offices and lasted between 30 and 45 minutes. As stated before, the interviews were followed by several emails and phone calls, ensuring that the respondents' comments would add to further clarifications.
3.9 Analysis
analysis, the researcher goes beyond what the respondents actually said in order to identify underlying ideologies or ideas that govern what people say.
For the analysis, the first step consisted in transcribing the interviews and carefully reading them to get a proper overview. In this second reading, to capture the essence of each sentence of interest, coding was deployed. In this study, the initial codes were inductive, which means that they originated from the respondents answers (Miles & Huberman, 1994). Thus, the codes emerged exclusively from the data. In the guidelines for conducting thematic analysis made by Braun and Clarke (2006), all data are coded, and codes are gathered into a theme or a pattern. Repetition, a word-based technique was specifically used to generate the themes and uncover further information (Ryan & Bernard, 2003). According to D'Andrade (1995), "words that occur a lot are often seen as being salient in the minds of respondents". The same author (1995) observes that "indeed, anyone who has listened to long stretches of talk, whether generated by a friend, spouse, workmate, informant, or patient, knows how frequently people circle through the same network of ideas." Thus, word frequency or counting the number of times each word occurred highlighted relevant information.
In this thesis, the coding of the material was based on the principles described by Braun and Clarke (2006). After the initial coding, codes were merged into units that would have similar meanings. Further on, in this study, a theme was defined for each unit in a way that each code included could identify with it. All came together in a coding agenda which can be reviewed in Appendix B.
To get a more detailed insight how the transcribed interviews were transferred into analyzable contents for this thesis, other than developing themes, statements, opinions and quotes were extracted from the text and subject to discussion and comparison.
perceived level of discrimination. Pearson Correlations and Independent sample t-tests were deployed.
The questionnaire includes two parts. The first part contained five multiple-choice questions, 3 of which are supplemented by text boxes that provide the opportunity to enter additional information. A further 17 questions allow only for box ticking answers corresponding to each individual's evaluation on a 1 to 7 scale. Among other questions, the questionnaire contained ethnicity related questions (U.S Federal Register, 1997) and religion related questions (Pew Research Centre, 2012). The results of the surveys were analyzed to determine the perceived level of discrimination felt by passengers in the airport security context. Descriptive statistics were generated which included the mean scores.
3.10 Conceptual Model
Figure 4. Conceptual Model
4. RESULTS AND FINDINGS
In this section the results from the conducted interviews and questionnaires will be presented, pointing out the relevant findings. The aim is, according to the research question, to explore how the perceived level of risk affects investment decisions and might induce a discriminatory behavior towards passengers. Firstly, by using quotes and statements extracted from the interviews, as well as the passenger questionnaire analysis, the main propositions of the thesis will be answered and proved if supported or not. These quotes will describe different aspects meant to point out how the interviewees assess risk, what is their view over potential passenger discrimination as well as predilections towards either investing in human operators or advanced technology to enhance security. Besides integrating quotes from the interview transcripts, a comparison between the two distinctive airports will be made, regarding risk assessment and passenger discrimination. Subsequently, Pearson Correlations will be deployed to establish a relationship between the number of events and risk tolerance a well as the number of incidents and perceived level of discrimination as reported by passengers. Additionally, using Independent sample t-test, the average levels of discrimination, in terms of ethnicity and religion between the two airports will be analyzed.
4.1 Perception of risk
4.1.1 Testing proposition P1 a
To test proposition P1 a, several facts relevant to the way the analysis has been conducted are required. Firstly, the interviewees were asked to rate, on a scale from 1 to 7, five distinctive scenarios as displayed by Table 1. It logically follows that as the degree of acceptability increases, the perceived level of risk decreases. Thus, for a better comprehension of the studied variable "perceived level of risk" the scale levels have been reversed as it can be observed.
Table 1. Acceptability/Perceived level of risk assessment based on given scenarios
Table 2. Scenario rating by Frankfurt and Bucharest airports
4.1.1.1 Level of Acceptability for each scenario
Table 3 demonstrates how by correlating the number of incidents and the acceptability level of the studied sample for each one of the five previously mentioned scenarios, the number of incidents positively correlates with the second, fourth and fifth scenario, while it negatively correlates with the first scenario. The overall result is positive, contrary to expectations.
4.1.1.2 Relation between acceptability, risk perception and number of incidents
The following correlation (Table 4) illustrates the relationship between the number of incidents, acceptability, and risk level. There is a low, but positive correlation (r=0.283) between the number of events and the level of acceptability. Subsequently, we can observe a low yet negative correlation between the number of incidents and the perceived level of risk (r=-0.283). Given the small sample of only 2 airports the results are not statistically significant (p>0.05) but reliable for the analyzed sample. Nevertheless, it is not in the purpose of this study to extend the findings to larger populations but to investigate two airport typologies. Concluding, as the number of attacks increases the acceptability level increases too, demonstrating greater tolerance for various incidents while the perceived level of risk decreases, thus rejecting P1 a. In fact, the overall result is exactly the opposite as expected. The higher the number of incidents, the higher the tolerance towards risk airports display, equivalent to a "habituation" behaviour rather than decreasing tolerance and considering the risk higher, thus adopting a more "sensitive" approach towards potential threats.
Table 4.Relationshpi between acceptability, risk perception and number of incidents
4.1.2 Testing proposition P1 b
in order to assess the respondents' evaluation on the same (Table 5) 1 to 7 Likert scale provided.
Table 5. Assessing similarly occurred incidents
As shown in Table 5, scenario S1 took place in Frankfurt International Airport while scenario S4 is similar to an incident experienced by Henri Coanda International Airport (for further detail see Appendix A). The German airport rated the first scenario with "totally unacceptable", the lowest tolerance level. Additionally, its counterpart ranked the fourth scenario with "totally unacceptable". Thus, preposition P1 b is supported.
4.2 Discrimination degree as perceived by passengers
Under the constant threat of terrorist attacks, the need for greater airport security has recently led to significant changes in passenger screening procedures. As Elias (2009) correctly assumes, "balancing individuals’ rights and expectations of privacy with screening effectiveness is a human performance challenge." Although both airports declared to be against passenger discrimination, survey results collected from passengers belonging to various races, creeds and nationalities show differently.
When asked if some ethnic groups are more likely to display terrorist behavior predispositions than others, the Head of S******* S******* and M********* S****** within National Company of Airports Bucharest, Mr. V. declared: "I totally disagree with this
statement(...) In this moment, we can not speak about a profile of the terrorist. So, if in the past we could speak about some ethnic groups (...) we are unable now to do so and it's a very wrong judgement if we continue to think in this way, as we are not able to say that there are some specific ethnic groups more than others predisposed for terrorist/criminal behaviour". When asked to explain this position Mr. V. stated as follows: " (...) Please keep in mind the fact that the authors of the last terrorist attacks in the US, in Belgium, and in other countries are people from Europe; are people well educated, (...) specialists in chemistry, in technologies, in aviation, some are even pilots. They have families, they have children, and they run businesses in the Western Europe, thus I totally disagree with this statement".
On the same question, his counterpart from Frankfurt International Airport, Mr. K. argued: "I am a strong opponent of saying -if you belong to a specific ethnic group you pose
a higher risk." When asked to elaborate on his answer the Vice President Corporate Safety
and Security M********* and Q******* added: "Unfortunately, based on previous
incidents there are some indications that some not ethnic groups but more religious groups are more often involved in security related incidents. So, if I am brought honest I would slightly agree with this statement".
accessible, not the same thing can be said about Frankfurt International Airport. Racial discrimination charges and complaints against Frankfurt Airport are various in nature. For instance, in 2013, a turban-wearing Sikh, a German of Indian origin declared that he has been shamingly screened and petted for no apparent reason and that his turban was rudely and continuously touched (which is forbidden according to Sikh tradition). His "harrowing experience" was forwarded to the officials in charge at the German Embassy in New Delhi (Singh, 2013). Another incident occurred in 2009. An Indian man, traveling to San Francisco and transiting Frankfurt International Airport was rudely approached by an official at the boarding gate. He suggested that the Indian passenger must get a new passport as the current one is "archaeological" and no machine code could scan it. Following up he added that in India that passport must be "brand new". From the incident, it can be construed that not only did he not have much respect for India but also for Indians and the act can be categorized as racist (Oak, 2009).
In what follows an analysis based on the responses of 120 collected questionnaires has been developed to identify potential discriminatory behavior perceived by passengers when departing/arriving/transiting Bucharest and Frankfurt airports. As stated before, a total of five incidents were associated with the Romanian airport and ten incidents for its German counterpart.
The Pearson Correlation developed using the available data previously presented illustrates a moderate, positive correlation (r=0.429) between the number of incidents and the perceived level of discrimination as evidenced by Table 6. Moreover, the results are statistically significant as p<0.01. In other words, as the number of attacks increases so does the perceived discrimination degree as witnessed by passengers, confirming P2.
4.2.1 Conducting Independent sample t-tests for analysing the perceived discrimination level
The discrimination topic is a very complex issue, often denied, which definitely requires more research. This study suggests a closer look at the different discrimination levels in the studied airports thus firstly developing an Independent sample t-test to investigate this issue. Secondly, the assumption based on which specific ethnic and religious passenger categories feel more discriminated than others will be considerated by using the same methodology.
The 120 respondents were asked to rate their perception regarding discrimination within the two studied airports on a 1 to 7 Likert scale where: 1=Completely disagree, 2=Mostly disagree, 3=Slightly disagree, 4=Neither agree nor disagree, 5= Slightly Agree, 6=Mostly agree, 7=Completely agree (Appendix F). Following these results, as illustrated in Table 7, the total discrimination value for Bucharest Airport is 1962 (based on adding up the responses, noted from 1 to 7 from all 60 passengers-Appendix F) and 3176 for the German airport (based on the answers of 60 passengers-Appendix F). Thus, the average discrimination per passenger is equivalent to a value of 32.7 in the Romanian airport and 52.93 for its German counterpart. The values are further used to develop the Independent sample t-test. Airports Number of respondents Average discrimination/pers on Discrimination (total) Bucharest International Airport 60 32.70 1962.00 Frankfurt International Airport 60 52.93 3176.00
Table 8. Results of the Independent sample t-test on discrimination as perceived by passengers in both Frankfurt and Bucharest airports
Using the data presented in Table 8, an Independent sample t-test was developed to further test the assumption based on which the higher the number of experienced incidents, the higher the perceived level of discrimination as reported by passengers in both studied airports. The Bucharest passenger group (N=60) is associated with an average perceived discrimination degree of M=32.70 (SD=17.98). By comparison, the Frankfurt respondents (N=60) were associated with a higher average discrimination level M=52.93 (SD=24.45).
4.2.2 Discrimination in terms of ethnicity and religion
For a more in-depth analysis on the issue, Independent Sample t-tests were deployed recording clear evidence that there are positive and significant levels of discrimination within both airports, in terms of passenger ethnicity and religion, particularly within Frankfurt International Airport. For a more accurate representation of results, the two major ethnicities that have been studied due to a significant majority of respondents have been "White/Caucasian" and "Asian" while "Christian" and Islam ("Muslims") have been analysed in terms of religion from the same consideration.
4.2.2.1 Discrimination perception of major ethnic categories
Table 9 illustrates the number of respondents from each studied ethnic category. Based on the same 1 to 7 Likert scale the total discrimination perceived by the sample of 33 White/Caucasians within Bucharest Airport scores a value of 757, resulting in an average discrimination of 22.93. Similarly, the sample of 6 Asians build up a discrimination value of 199 which divided by 6 (the number of respondents) results in an average discrimination perception of 33.16. The same calculation procedure follows for Frankfurt Airport. With a total of 33 White/Caucasian respondents, adding up to a total discrimination value of 1405, the average discrimination rate equals to 42.57. Meanwhile, 5 Asians scored a total of 273 discrimination points, resulting in an average of 54.6 discrimination per Asian.
As a supplementary remark, concerning ethnicity, Frankfurt International Airport records higher average passenger discrimination levels for all passenger categories. Arabs score the highest values, followed by Black heritage people and Indians.
Ethnicity - Number of respondents White/Caucasian Indian (Indian Subcontinent) Pacific Islander Hispanic
or Latino Asian Arab Black Romani
Bucharest 33 2 1 1 6 4 3 1
Frankfurt 33 7 3 7 5 2 1 0
Ethnicity - Average Discrimination per person White/Caucasian Indian (Indian Subcontinent) Pacific Islander Hispanic
or Latino Asian Arab Black Romani
Bucharest 22.93 40 36 30 33.16 61.75 37.66 76
Frankfurt 42.57 82 42 50.28 54.6 104.5 95 0
4.2.2.1.1 Independent Sample t-test on discrimination as perceived by White/Caucasian passengers in Bucharest and Frankfurt Airport
Table 10. Results of the Independent Sample t-test on discrimination as perceived by White/Caucasian passengers in Bucharest and Frankfurt airports
4.2.2.1.2 Independent Sample t-test on discrimination as perceived by Asian passengers in Bucharest and Frankfurt Airport
Table 11. Results of the Independent Sample t-test on discrimination as perceived by Asian passengers in Bucharest and Frankfurt Airport
4.2.2.2 Discrimination perception of major religious categories Religion - Number of respondents
Christian Hindu Muslim Atheist Jewish Buddhist
Bucharest 28 2 14 4 6 6
Frankfurt 37 7 8 1 2 5
Religion - Average Discrimination per person
Christian Hindu Muslim Atheist Jewish Buddhist
Bucharest 25.21 40 51.71 25 25.5 33.16
Frankfurt 41.21 82 87.37 27 39 54.6
Table 12. Number of Christian & Muslim passengers and average discrimination per passenger
Table 12 illustrates the number of respondents from each studied religious category. A remarkable observation, once again, is that Frankfurt International Airport scores higher
average discrimination levels for each one of the religious groupss, Muslims in particular.
4.2.2.2.1 Independent Sample t-test on discrimination as perceived by Christian passengers in Bucharest and Frankfurt Airport
Table 13. Results of the Independent Sample t-test on discrimination as perceived by Christian passengers in Bucharest and Frankfurt airports
4.2.2.2.2 Independent Sample t-test on discrimination as perceived by Christian passengers in Bucharest and Frankfurt Airport
Table 14. Results of the Independent Sample t-test on discrimination as perceived by Muslim passengers in Bucharest and Frankfurt airports
The Muslim passenger group (N=28) in Bucharest Airport scores an average perceived discrimination level of M=51.71 (SD=5.06) as evidenced by Table 14. Frankfurt's Muslims (N=8) record a much higher average discrimination level M=87.37 (SD=6.49). Results demonstrate how, on average, Muslim passengers in Frankfurt Airport feel remarkably more discriminated than those from Bucharest airport. The difference is significant t(15.1) = 4.32 with p<0.01. Once again, it was expected that Muslim passengers on the German airport to experience higher levels of discrimination than Muslims on Henri Coanda International Airport.
4.3 Investment decision-making
Both of the interviewees agreed on the fact that a combination of both technology and human operators is needed to ensure the highest possible level of security.
"Aviation security means a combination of equipment, procedures, and human resource."
(Mr. V., Head of Security S******* and M********* S******, National Company Airports Bucharest)
" I am not a person who says one or the other. I say a combination of both is needed. What makes the good performance is the ideal combination between technical detection and human resources. "
(Mr. K., Vice President Corporate Safety and Security M********* and Q******, Frankfurt International Airport)
When it comes to successful data analysis and threat identification the respondents had opposite opinions. Mr. V. considers that no matter what a machine does or does not detect, it will always be a human operator's final interpretation and results that will end a screening process. Meanwhile, Mr. K. argues that if a system declares a person or screened object "clear", the human operator is no longer needed as the machine itself has finalized the process and establish the final result.
"Even if we have the newest technology, the very expensive technology, the last decision will be put in place by the human factor. So the people must be well trained, well recruited, very very well prepared to conduct screening (...). So, we are investing continuously in the training of the operators, in recruiting the best people with native abilities who could further develop them in order to take the right decision in a very short segment of time."
(Mr. V., Head of Security S******* and M********* S******, National Company Airports Bucharest)
detection, again, such a system is more reliable than a person doing a hand search (...) with good technology you can make the lives of our staff much easier and to make them work more efficient which I believe it leads to better motivation and performance of the humans (...) First, we have to invest in technology but technology can only work with a proper human being using this technology (...) When using automatic detection as long as the machine says "it's clear" the operator is no longer in the game."
(Mr. K., Vice President Corporate Safety and Security M********* and Q******, Frankfurt International Airport)
The major issue which leads to this difference in opinions is the different perceptions of reliability in terms of human resources or advanced technology. Mr. V. states that both operators and machines can record failures and both can fail in identifying a potential risk. Nevertheless, he argues:
"The equipment could fail but in combination with the human resource, we can not tolerate having a failure. "
Debating on this statement, it can be concluded that he emphasizes human operators as more reliable and capable of coping with potential machinery failure. Following the interview, during confirmation emails and phone calls, Mr. V. added " if a machine is broken
the person can do a manual and visual search of the luggage. But if the person is sick the machine on itself can't do the task of an operator itself. Humans are the core of the detection and screening process."
In support of his predilection to invest in operators, Mr. V. highlights an aspect where a human specialist performs better than advanced technology. He emphasizes the human capabilities by giving an example: "A machine can identify a cell phone and label it as a "cell
phone"; however, put in 2 pieces will the machine still see it as a cell phone or as a battery and a cover separately? meanwhile, the human understanding can still connect the two pieces and know it is a cell phone". Mr. V. points out a unique characteristic which advanced
technology lacks: cognitive skills: "Human operators can identify whether an individual is
Mr. K. explains the advantages of human operators by adding "The advantage is that
a human operator is more flexible and can identify what a machine might miss" while also
naming various limitations operators face: "Human beings get tired and distracted, have a
bad day or feel sick; there are so many things can influence negatively the performance".
Although he points out a major disadvantage machines can not overcome "The machine can
only find what it has been designed to find, taught before. If you calibrate a machine to detect a gun it detects a gun...but if you do not calibrate it to detect a small gun it will not detect a small gun". Nevertheless, he concludes by adding "Technology is more reliable and easy to control."
It is thus clear the direction of future security investments that will be chosen by these airports. Preposition P3 a has been confirmed for Henri Coanda International Airport and it is not supported for its counterpart. Subsequently, Frankfurt International Airport will direct its finances towards acquiring advanced technology primarily, supporting proposition P3 b.
4.4 Word Repetition analysis
THEMES CODES Bucharest International Airport Word frequency Frankfurt International Airport Word frequency Potential attackers Suspect(s) 12 Suspect(s) 1
Ethnic groups 4 Religious groups 1
Terrorist(s) 9 Terrorist(s) 6 Terrorist attack aftermath Fatalities 0 Facilities 1 Injuries 2 Injuries 2 Explosion 2 Explosion 4
Security system Security procedure Security regulations Security area Security authorities 30 Security procedure Security regulations Security area Security authorities 20
Potential risks Bomb 7 Bomb 6
Other threats 15 Other threats 5
Prohibited items Weapons 0 Weapons 2
Guns 0 Guns 5 Technological endowments Advanced technology Equipment Screening systems Machines Detection systems 7 Advanced technology Equipment Screening systems Machines Detection systems 22 Human resources
Human operators 14 Human operators 13
Agents Agents
Specialists Specialists
Table 15. Word repetition analysis
Since perceptions and habits determine how people will act, they will shape the nature of the response actions when confronted with various unlawful incidents (Wachinger & Renn, 2010). If people underestimate the intensity of the threat, they may not take the necessary cautionary actions to protect themselves. Or vice versa: if they overestimate a specific attack they might spend far too many resources on reducing only minimal risks. Also, depending on the perceived nature of the attack/incident, they may misjudge the appropriateness of specific protective behavior or options for mitigation.
The following review based on word frequencies identified will highlight several differences between risk perception and investment-decision making between the two airports. After filtering the initial codes (Appendix B), the remaining codes have been organized into seven major themes. Interestingly, the frequency of specific word usage differs highly between the two airports pointing out different typologies and behaviors.
Bucharest International Airport displays a permanent active position towards any potential unlawful occurrence, as declared by Mr. V. "We are permanently very focused on
any suspicious occurrence" (...) Our crisis management team is always ready to react". This
been in-depth debated during the interview. "Agents conduct visual screening, followed by a
casual questioning if an individual shows signs of suspicious behaviour (...) If the doubt persists, formal questioning is conducted" (Mr. V.). The word "bomb" was heard nine times
throughout the interview, showing a particularly sensitive spot for the airport which has experienced four false bomb alarms during the past ten years out of the five officially reported incidents (Appendix A). Not having confronted with any incidents encompassing the use of weapons, firearms, and guns, these items have not been mentioned. Thus, how literature suggests, experiences influences both the perceived risk and the likely behavior (Wachinger & Renn, 2010).
Meanwhile, Frankfurt International Airport displays, overall, what literature calls "habituation" behavior, described as "a decrease or cease to respond to a stimulus after repeated presentations." The higher the number of occurred incidents, the less sensitive Frankfurt International Airport seems, as suggested by the modest use of context-related words. Interestingly, in support of the theory claiming that previous experiences impact current perceptions (Duncker, 1939) the word usage during the interview records the repetition of the word "weapon," "firearm" and "gun" (Table 15). These prohibited items have been involved in incidents that took place at the airport (see Appendix), a possible reason for which they have been brought into the discussion. Human fatalities have also been mentioned, the airport having experienced a dramatic incident in 2011 when 2 US Airmen were killed (Appendix A). The habituation behaviour is clearly illustrated by Frankfurt Airport, underestimating the potential dangers posed by a passenger boarding with a knife. Mr. K. declares "People are being attacked with knives on the streets as well (...) This small,
everyday item does not pose a major risk (...) If you have a person who is able to do martial arts he can pose a similar risk without any other dangerous item". Continuing, it is
interesting to remark the fact that "ethnic groups" has been replaced with "religious groups" and considered to be more prone to conduct terrorist attacks. This statement comes in support of the P2 findings, namely that specific religious groups feel a higher level of discrimination than other religious minorities within Frankfurt Airport, Muslims in particular.
5. DISCUSSIONS
This chapter will elaborate on the results from the analysis and compare them to the established propositions and research questions. It will also elaborate on other findings that were discovered during the study.
5.1 How is risk associated to attacks perceived?
The variable "Perceived level of risk" was designed to measure the tolerance degree on various incidents. It was believed from the theory that previous experiences build on the personal perception of risk in the direction of airports becoming more sensitive (Grill-Spector, Henson, & Martin, 2006) towards potential future attacks as a response to an increase in the number of incidents. Accordingly, the P1 a proposition states that the higher the number of experienced incidents, the lower the acceptability rate and the higher the perceived level of risk. As shown in the Results section the proposition is not supported. The result is highly unexpected, as, given the numerous current threats faced by the aviation industry, a "habituated" behavior (Campbel Institute, 2014) as a response to an airport frequently being exposed to various incidents should be avoided in favor of a more "sensitive" one.
Several explanations exist to support the findings:
