Legitimacy, Performance, and Procedural Justice

Recent studies suggest that the legitimacy of police actions and the acknowledgment of the procedural justice of police practices are essential parts of enhancing police effectiveness (Reference Skogan and FrydlNational Research Council 2004; Tyler Reference Tyler2004a, Reference Tyler, Skogan and Frydlb; Reference Tyler and HuoTyler & Huo 2002; Reference Tyler and WakslakTyler & Wakslak 2004). Reference Skogan and FrydlThe National Research Council (2004: 291) defines legitimacy in the context of policing as “the judgments that ordinary citizens make about the rightfulness of police conduct and the organizations that employ and supervise them.”

Legitimacy is a critical variable in policing and security procedures—such as security screening at airports—for several reasons. First, the more trust that citizens place in agents of authority, the more authority they are willing to invest in them (Reference Skogan and FrydlNational Research Council 2004). This authority is central to the activities of the police and other agencies that are responsible for ensuring the security and safety of communities. Second, citizens who trust agents of authority, such as the police, are more willing to ask them for help and to assist them in identifying suspects or offenders (Reference Skogan and FrydlNational Research Council 2004). Third, citizens who are involuntarily stopped by the police are more likely to comply with police demands if they feel that they are treated fairly (Reference TylerTyler 1997; Reference Tyler and HuoTyler & Huo 2002). The opposite has also been found to be true: people who feel that they are treated disrespectfully exhibit resistance, consequently making it more difficult for the authorities to carry out their responsibilities (Reference Mastrofski, Snipes and SupinaMastrofski, Snipes, & Supina 1996; Reference White, Cox, Basehart, Barker and CarterWhite, Cox, & Basehart 1991).

Tom Tyler's studies emphasize the two main variables that affect legitimacy: performance and procedural justice (Reference TylerTyler 2001, Reference Tyler2004a). In his work, he examines the attitudes of the public toward the police and the courts and concludes that considerations of both performance and procedural justice in fighting crime affect legitimacy perceptions. Evaluations of performance, including the ability of the police to catch lawbreakers and achievements in fighting crime (Reference Sunshine and TylerSunshine & Tyler 2003), were found to play a significant, though weaker, role than that of procedural justice in predicting legitimacy perceptions (Reference TylerTyler 2001, Reference Tyler, Skogan and Frydl2004b).

A crucial variable that affects legitimacy is procedural justice—that is, citizens’ perceptions of the fairness of the procedures by which rules are created and deployed (Reference Skogan and FrydlNational Research Council 2004). The procedures that lead to a certain outcome are also important considerations when assessing perceived fairness. Studies indicate that peoples’ subjective experiences of legal authorities are influenced by the degree to which they evaluate the process as fair and impartial, rather than by the outcome itself (Reference Leventhal, Thibaut, Spense and CarlsonLeventhal 1976; Reference Thibaut and WalkerThibaut & Walker 1975; Reference TylerTyler 2001; Reference Tyler, Smith, Gilbert, Fiske and LindzeyTyler & Smith 1997; Reference TylerTyler et al. 1997).

Procedural justice has several dimensions: process control (the ability to have a say in the procedures); decision control (the ability to influence the outcome); and the consistency, neutrality, and accuracy of information provided by the authorities (Reference Leventhal, Thibaut, Spense and CarlsonLeventhal 1976; Reference Thibaut and WalkerThibaut & Walker 1975; Reference TylerTyler et al. 1997; Reference SindhavSindhav et al. 2006). In addition to influencing perceptions of specific encounters, procedural justice has been found to have an effect on evaluations of police organizations overall (see review by Reference TylerTyler 2001). Moreover, this argument appears to be true not only for the police, but also for other criminal justice agencies and even private institutions (Reference Skogan and FrydlNational Research Council 2004; Reference Seiders and BerrySeiders & Berry 1998).

Some scholars argue that minority groups’ sense that they have been profiled by law enforcement agencies is also a key variable in understanding police-minority relations (Reference Lundman and KaufmanLundman & Kaufman 2003; Reference Tyler and WakslakTyler & Wakslak 2004; Reference WeitzerWeitzer & Tuch 2002), especially among minority social groups with a history of tension with such agencies (Reference Hasisi and WeitzerHasisi & Weitzer 2007; Reference WeitzerWeitzer & Tuch 2002, Reference Weitzer2006). Indeed, recent research points to the importance of psychological factors in affecting people's interpretations of their interactions with law enforcement agencies (Reference Tyler and WakslakTyler & Wakslak 2004). When citizens think that the police are profiling them, the police have less authority in their eyes. Reference Tyler and WakslakTyler and Wakslak (2004) have found that when people make profiling attributions, they become more resistant to accepting police decisions. Their results indicate that those who made a profiling attribution were less willing to defer to authorities, and those who had experienced high-quality interpersonal treatment were less likely to feel that they had been profiled (Reference Tyler and WakslakTyler & Wakslak 2004: 259).

While the legitimacy processes that underlie airport security are not necessarily similar to those that underlie policing, we think that the comparison is a reasonable one. Indeed, the process of airport security checks has similar components to the interaction between citizens and the police. In this regard Reference SkolnickSkolnick (1966) argues that “… the policeman's role contains two principal variables, danger and authority … the element of danger seems to make the policeman especially attentive to signs indicating a potential for violence and lawbreaking” (Reference SkolnickSkolnick 1966: 44). This is also true in the case of airport security checks (Reference Dugan, LaFree and PiqueroDugan, LaFree, & Piquero 2005; Reference Persico and ToddPersico & Todd 2005).

A very important question concerns the impact of the legitimacy of the airport security process. It is important because the literature on policing, along with qualitative observations at the airport and interviews with security personnel, teaches us that passengers with a negative attitude toward the security process tend to come into conflict with security staff, which in turn makes the process harder to carry out and may even complicate it. Irate passengers tend to be held back and asked to go through even more thorough security checks. These conflicts create much bitterness among the passengers and might give them the feeling that they are being persecuted and oppressed. It is important to note that these conflicts also harm the functioning of the airport security personnel, as they take up time and resources, taint the personnel's relationships with the passengers, and make the security process harder.Footnote ⁴

Recent years have seen a growth in the literature in the field of procedural justice and terrorism. Reference Tyler, Stephen and AzizTyler, Schulhofer, and Huq (2010), for instance, analyze the circumstances under which members of the Muslim-American community voluntarily cooperate with police efforts to combat terrorism. They interviewed Muslim Americans in New York City between March and June 2009. Their most prominent finding is that the procedural justice of police activities is the primary factor shaping legitimacy and cooperation with the police. Reference Jonathan and WeisburdJonathan and Weisburd (2010) study shows how majority communities view the potential costs of policing terrorism. It has been assumed that the community threatened by terrorism will support harsh counterterrorism responses and will be less concerned with possible unintended negative outcomes. However, similar to the findings of Reference Tyler, Stephen and AzizTyler, Schulhofer, and Huq (2010), Jonathan and Weisburd's results suggest that “what the community wants the police to do” (180) in the face of an immediate threat may be more complicated than merely providing forceful and rapid responses, since the community is also well aware of possible negative outcomes of policing terrorism. They argue that in the face of security threats the public does not want the police to focus solely on providing forceful responses while abandoning its classic duties and fair processes.

Israeli Arabs and Airport Security Checks

Israel's airport security procedures were first established in 1968 after the hijacking of an El Al aircraft on July 23 of that year. Three members of the Popular Front for the Liberation of Palestine (PFLP) hijacked an El Al airplane en route from Rome to Tel Aviv and forced the pilot to land in Algiers. This episode drove Israel to develop terrorist profiles that have been employed ever since, and to establish, for the first time, a special department of aviation security. Airport security procedures in Israel are roughly made up of four circles of security: early detection outside the airport zone, inside the airport zone, inside the terminal, and inside the aircraft.Footnote ⁵ In the terminal, the third circle, the airport staff identifies high-risk travelers based on at least three sources. The first source is the screening of passenger lists before passengers even arrive at the airport. This screening is based mainly on intelligence sources, which create watch lists. The second source is passenger screening managed by a data mining system known as the Computer Assisted Passenger Pre-screening System (CAPPS) (United States General Accounting Office 2004). CAPPS collects data from external sources—mainly airlines and travel agencies—about passengers’ flight habits, how they paid for their plane tickets, their flight itineraries, travel records, whether a car was rented, whether passengers are flying alone, their meal preferences, and other data that can be inferred from the flight tickets. The third and most critical source is special questioning. This involves a standard 30-second questioning protocol, a passport check, luggage screening, and the use of metal detectors.Footnote ⁶

Some countries base their national security policies on ethnic affiliation. This is often the case because of transborder ethnic ties of minority groups with external state enemies (Reference EnloeEnloe 1980; Reference Fearon and LaitinFearon & Laitin 2003; Reference GurrGurr 1993; Reference HorowitzHorowitz 1985, Reference Horowitz2001; Reference King and Melvin1999/2000). Moreover, it is much simpler to base homeland security policy on collectives, such as ethnic groups, rather than on individuals. Homeland security agencies are thus able to save resources that otherwise would have to be dedicated to collecting intelligence on individual citizens. Finally, ethnic groups (especially those living in divided societies) often have stable attitudes with regard to their position toward state apparatuses, and so these orientations are seen to be a solid base for assessing the potential of threat to homeland security (Gurr Reference GurrKing & Melvin 1993; Reference HorowitzHorowitz 1985, Reference Horowitz2001).

The political identity of the Israeli Arab minority includes two major components that have important ramifications for their security profile. The first relates to their overall relationship to the state. Arab Israelis (with the exception of specific communities, such as the Israeli Druze) did not generally support (and tended to oppose) the establishment of the State of Israel, and have remained uncomfortable with the country's Jewish national identity (Reference GhanemGhanem 2001; Reference RouhanaRouhana 1997; Reference SmoohaSmooha 1992). As Palestinian Arabs, this minority also has an ethnonational affiliation to Israel's neighboring Arab countries, and in particular to the Palestinians beyond Israel's borders, who are involved in an ongoing political conflict with Israel regarding the solution of the Palestinian problem and the place of Israel in the Middle East.

This political identity appears to have led to more intense security screening of Arabs by the IAA, though it does not publicly acknowledge that its profiling procedures have any ethnic or national component. A report conducted by the Arab Association for Human Rights (HRA) in Israel states that Israeli Arabs have complaints about the severe security checks to which they are subjected when they enter or leave the country (The Arab Association for Human Rights 2006). The report notes that Israeli Arab passengers emphasized the feelings of humiliation and insult that the security checks caused them. This feeling of humiliation was especially powerful among Arabs who were traveling with a group of Israeli Jews, and who were singled out for a different screening procedure. Furthermore, Israeli Arab passengers reported that this feeling was intensified by the disparaging attitude of the security officers. Some passengers reported rude and offensive behavior, and said they felt that they were being treated as second-class citizens (The Arab Association for Human Rights 2006).

The report also emphasizes the fact that having special screening procedures carried out in full view of the other passengers heightened the Israeli Arab travelers’ sense of humiliation considerably, because they had been publicly “marked” as more likely to be terrorists (The Arab Association for Human Rights 2006). The report argues that these passengers therefore had to deal with the accusatory gazes of Israeli Jewish passengers, as well as with the shame and embarrassment of having their personal and intimate possessions scrutinized in plain view of other passengers and the suspicious and somewhat blunt attitude of the security personnel. However, there is currently no empirical evidence regarding Israeli Arab passengers’ experiences at the airport, and this is the major contribution of the current study.

As we noted earlier, airport security procedures in Israel are comprised of four levels of intervention. This study focuses on the third level (intervention in the terminal), which includes identification of high-risk passengers, basic questioning of the passenger, and in-depth inspections when needed (body search, additional search of luggage, intensive questioning, and so on).

Research Questions

The purpose of this study is to understand the variables that influence passengers’ legitimacy perceptions regarding the security checks at Israel's major international airport. Following the literature on police-citizen relations, one might assume that passengers’ attitudes toward the security checks would be influenced by their ascribed identities (i.e., race, ethnicity, nationality, and so on), such that passengers from disadvantaged social groups would express negative attitudes. Accordingly, our first research question is as follows: What role does ethnonational identity (Jew versus Arab) play in predicting passengers’ legitimacy perceptions of the airport security checks in Israel? Our hypothesis is that the attitudes of Arab passengers toward airport security will be significantly more negative than those of Jewish passengers.

Although ethnonational identity may well be a central issue, we should not assume that it is the only factor that influences legitimacy perceptions of airport security measures among Arabs and Jews. There are also marked sociodemographic distinctions between Arabs and Jews, such as age, income, education level, gender, and marital status (Reference HyderHyder 2008). For instance, we know that the Israeli Arab minority is much younger and poorer on average than the Jewish majority. Data also indicates that the Israeli Arab population has lower educational achievements than Israeli Jews (Reference EtiWeisblay 2006).Footnote ⁷ Bearing in mind the patriarchal structure of Arab society in Israel, we should also consider the effect of gender and marital status on perceptions of the legitimacy of airport security. This consideration informs our second question: To what extent are passengers’ trust perceptions influenced by sociodemographic variables, such as gender, age, marital status, income, and education level?

In this study we wish to move beyond ascribed identities (ethnicity and sociodemographic variables) when analyzing passengers’ legitimacy perceptions of airport security procedures. Following recent research on the importance of procedural justice in evaluations of legitimacy, we thought it important to assess perceptions of the fairness of such procedures. Accordingly, our third research question is as follows: How do the passengers’ beliefs that they were treated differently from other passengers during the security check (profiling) and felt humiliated and intimidated by the security checks affect their perceptions of legitimacy toward the airport security checks? Our hypothesis is that those passengers who report negative experiences during the security checks will view the security procedures as less legitimate. Finally, we pay attention to specific characteristics of the security procedures. We measured one element that has been specifically noted in criticisms of Israeli airport procedures: whether or not the passenger's suitcase was opened for additional checks. Our fourth question, therefore, is as follows: How do different security processes influence the passengers’ legitimacy perceptions of airport security procedures? Our hypothesis is that more stringent procedures will lead to lower legitimacy evaluations.

Methodology

In order to evaluate passengers’ perceptions of airport security procedures, we conducted a survey among a random sample of passengers who had finished going through the security procedures at Ben Gurion Airport. We received approval from the IAA to conduct the survey. The Institutional Review Board at the Hebrew University of Jerusalem reviewed the research protocol to ensure compliance with human subject research policies.

The survey was carried out by a research team from the Hebrew University over a period of four weeks during August 2008. The specific method by which the survey was carried out was coordinated with the IAA so as to ensure the random oversampling of Israeli Arabs. In every case we made contact with passengers immediately after they had passed through security screening.

The sample was stratified in order to ensure the adequate representation of Israeli Jews (308) and Israeli Arabs (306), and it totaled 614 individuals.Footnote ⁸ The survey was carried out in one of four languages—Hebrew, Arabic, English, or Russian—according to the respondent's native tongue, or whichever language the passenger felt most comfortable speaking. All of the assistants who carried out the survey spoke at least one of these languages fluently. Passengers who completed the full questionnaire received a coffee voucher as a token of appreciation. By providing this incentive we hoped to reduce refusal rates. The final refusal rate was 60 percent. We could not identify any specific characteristics of those passengers who refused to complete the questionnaire; most replied that they were in a hurry to get to the duty-free shops.

One important assumption of our research is that airport security personnel can readily identify and differentiate Arab from Jewish travelers. This characterization is carried out by drawing on a number of indicators. The first indicator is the passenger's name—Arabs have both first names and family names that are distinct from the names of Jews in Israel. Second, the security personnel ask where the passenger has come from, or where he or she lives. Most of the Arabs in Israel (90 percent) live in majority Arab communities, and so place of residence tends to reveal the passenger's ethnicity. The third indicator is dress—many Arab passengers are from more traditional communities that maintain conservative dress styles, especially among women. The combination of these indicators helps the security personnel to identify the passenger's ethnicity (Arab or Jewish), and thus aids in profiling Arab passengers.

This study focuses on passengers taking outbound flights from Israel, not on passengers arriving in Israel. The reason for this has to do with the way that the security checks are carried out. In Israel, security checks are carried out when people are leaving the country, not when they are entering it. The assumption is that inward passengers have been checked at their points of departure overseas. There are some very rare cases in which security staff come directly to the aircraft and detain certain passengers for further checks. The focus of our study is the security checks among people leaving Israel. This process is extremely structured and orderly and allows for in-depth research observation, quite unlike studying the population of passengers returning to Israel.

Dependent Variable

The purpose of the current study is to learn about passengers’ perceptions of trust toward the airport security inspectors. Based on previous research, the definition of legitimacy included trust in the agency and the official as well as moral identification with the values of the official and the obligation to obey an official even if you think that a particular order is unjustified (see Tyler 2004a). Our dependent variable was operationalized using one statement: “I trust the security inspectors.”Footnote ⁹

The descriptive findings of the dependent variable in Table 1 show that, overall, 75 percent of the passengers trusted the security staff at the airport. Nevertheless, when comparing Israeli Jewish and Israeli Arab passengers, we found a significant difference between these groups. We can see that Israeli Arab passengers held less positive attitudes toward the security checks than Israeli Jewish passengers (64 percent and 85 percent, respectively). A majority of both groups trusted the security staff, but that trust was significantly greater among Israeli Jews.

Table 1. Descriptive Statistics of the Dependent Variable—Trust in the Security Inspectors

Note: Asterisks denote significance levels from analysis of variance: ^*< 0.05; ^**< 0.01; ^***< 0.001.

Independent Variables

There are five types of independent variables in this study. The first is the ethnic identity of the passenger, which is measured by contrasting Jews and Arabs. The second group comprises the sociodemographic variables, including age, income, education level, gender, and marital status (see Appendix 1). We also measured characteristics of the passenger's imminent flight, the third domain of variables, by asking questions about the passengers’ reason for traveling (business, tourism, family visit, pilgrimage, and so on) and about whether or not the passenger was flying alone. We also asked how frequently the passengers fly and asked them to identify the destination of their flights. We included in the background variables a measure estimating the passengers’ justification of the security checks in Israel.Footnote ¹⁰ The fourth type of independent variable concerns the passengers’ evaluations of the security process. This variable is an overall measurement of a passenger's belief that he or she was profiled, humiliated, and/or intimidated during the security checks and represents the component of procedural justice that involves concern about being treated as others are treated. This variable was measured by three statements: “The treatment I received during the security check was different from the treatment other passengers received,” “The security check caused me to feel humiliated,” and “I felt intimidated by the security check.” The respondents were asked to rank their agreement with statements from 1 to 5 (1 = Strongly disagree; 5 = Strongly agree). Combining these measures into one measure resulted in a Chronbach's alpha of.71, indicating a reasonable level of internal reliability for the index.

Table 2 presents the perceptions of airport security performance among passengers. We asked the passengers to what extent the security check contributed to their sense of safety during the flight. A majority of 83 percent of the passengers felt that the security check contributed to their sense of safety during the flight. Furthermore, when comparing Israeli Arabs and Israeli Jewish passengers, we found a small difference (about 80 percent versus 87 percent, respectively). This finding shows that both Arab and Jewish passengers acknowledge the contribution of airport security checks to their sense of safety during the flight.

Table 2. Descriptive Statistics of the Independent Variable—Performance of Airport Security

Note: Asterisks denote significance levels from analysis of variance: ^*< 0.05; ^**< 0.01; ^***< 0.001.

Table 3 presents the passengers’ evaluations of whether they went through biased security procedures at the airport (feeling of neutrality versus profiling, humiliation, and intimidation). Itshows that 23 percent of all passengers felt that the treatment they received during the security check differed from that of other passengers. However, Israeli Arab passengers were considerably more likely (33 percent) than Israeli Jewish passengers (13 percent) to agree with the statement that they were treated differently by the airport's security personnel. While 5.5 percent of all passengers felt that the security checks caused them to feel humiliated, a significant disparity between Jews and Arabs passengers was found. Arab passengers were more likely to report that the security checks caused them to feel humiliated than Jewish passengers were (19.1 percent vs. 5.5 percent, respectively). A similar pattern was also found with regard to feelings of intimidation during the security checks. Table 3 shows that 11.4 percent of all passengers felt intimidated by the security check, and, as with the other measurements of bias in security checks, Arab passengers were significantly more likely to report feeling intimidated by the security checks when compared to Jewish passengers (17.9 percent versus 4.9 percent, respectively).

Table 3. Descriptive Statistics of Perceptions of Biased Security Checks

Note: Asterisks denote significance levels from analysis of variance: ^***< 0.001.

The fifth type of independent variable is our measure of the characteristics of the security process: whether the passenger's or his or her companions’ suitcases were opened for additional checks (yes or no). We assumed that the implementation of additional security checks, such as opening a passenger's suitcase, would negatively affect passengers’ perceptions of the legitimacy of the security checks. Table 4 shows that about 27 percent of the Israeli passengers reported that their suitcases went through additional security checks that required them to be opened and examined. When comparing Arab and Jewish passengers, we can see that about 10 percent of the Jewish passengers’ suitcases were opened, compared with 46 percent of Israeli Arab passengers’ suitcases. These data confirm concerns that Arab Israelis are singled out for especially stringent security procedures. We do not think this is particularly surprising, given recent discussions of this issue in the courts in Israel (Supreme Court Appeal, 4797/07), but this is the first time it has been demonstrated in empirical data.

Table 4. Descriptive Statistics of the Independent Variable (Security Treatment)—Whether Passenger's Suitcase Was Opened for Additional Checks

Note: Asterisks denote significance levels from analysis of variance: ^***< 0.001.

Multivariate Analysis

While our data so far suggest strong differences between Arabs and Jews in terms of their trust toward the airport security inspectors, in order to answer our research questions it was necessary to subject our data to a multivariate statistical model. For example, we already know that Arabs are likely to evaluate the airport security procedures less positively, but what we do not know is whether those differences persist once we have controlled for other variables, including those that represent elements of procedural justice, performance, and treatment. We present our models using a step-by-step approach reflecting our five research hypotheses (see Table 5).

Table 5. OLS Regression Model Predicting Passengers’ Trust Regarding Airport Security Inspectors

Note: Asterisks denote significance: ^*< 0.05; ^**< 0.01; ^***< 0.001.

Model 1 shows that in the absence of other independent variables, ethnic identity played a significant role in determining passengers’ trust toward the airport security inspectors (β = −0.253), whereby Israeli Arab passengers tended to trust security inspectors less than Jewish passengers did. Ethnic identity becomes slightly less important but remains strongly statistically significant (comparing the unstandardized β-values, β = −0.193) when we include the sociodemographic variables, the characteristics of the present flight (frequency of flying), and whether the passenger justifies the security checks. Furthermore, Model 2 shows that unmarried passengers tend to hold more negative views of the security inspectors than married passengers do, while the higher the level of the passenger's education, the lower his or her evaluation of the airport security inspectors. The frequency of passengers’ travel seems to negatively affect their trust in the security inspectors: those passengers who flew once or more during 2008 expressed less trust toward airport security inspectors than passengers who had not taken a flight during 2008. Endorsement of the statement, “The security check is justified considering the reality of Israel's security situation,” seems to positively affect trust in airport security inspectors. Gender, income, and age are not significantly related to trust in this model, and the largest standardized effect is found in relation to ethnic identity.

In the third stage we estimated the role of the performance of airport security personnel in predicting the passengers’ trust in the airport security inspectors. Model 3 indicates that performance plays a major role in predicting trust, and it is the strongest predictor in the model (β = 0.304). Passengers who felt that the security check contributed to their sense of safety during the flight tended to express more trust in the security inspectors.

In Model 4 we added the measure of procedural justice (degree of bias in security checks), which assessed whether the passenger thought he or she had been profiled during the security procedures and whether he or she felt intimidated and humiliated by the security checks. This variable becomes the strongest predictor in the model (β = −0.227, followed by Performance β = −0.217), with passengers who believed that they had been specially identified, humiliated, and intimidated during the security procedures giving much lower trust evaluations. Importantly, with the addition of this measure (procedural justice), the importance of ethnic identity, while still a significant factor, declines by a quarter.

In the full model, Model 5, we added a description of the procedures deployed in the security checks that the respondent had just been through. The findings show that having one's (or one's companions’) suitcase opened for additional checks has strong negative influences on the passenger's trust in the security procedures. We should also mention that in this model, perceptions of biased security checks still have the most salient effect on passengers’ perceptions of legitimacy (β = −0.231). Perhaps most important, Model 5 shows that once the component of having a suitcase being opened is added to the model, ethnic identity ceases to be a significant predictor of trust in the security inspectors.

Discussion

Our study has yielded a number of important findings regarding security screening procedures. To the best of our knowledge, this is the first time that researchers have been able to document differential security screening procedures between Arabs and Jews in Israel. Importantly, despite the differences that we found, both groups in Israel perceived airport security procedures to be generally legitimate. At the same time, while there are significant differences in the legitimacy evaluations of Arabs and Jews, in the context of a multivariate model that includes elements of procedural justice, ethnonational identity loses its significance. Below we discuss these results and consider their implications for security screening procedures more generally. We also note some of the limitations of drawing conclusions from this study.

The fact that Israeli Arabs report experiencing special security procedures at the airport is not surprising. As noted above, there has been a series of legal cases about profiling and the limits of the use of ethnonational identity for screening procedures. Indeed, the Supreme Court of Israel recently asked the IAA and the General Security Service to present the relevant facts and legal justifications of the selective screening of Israeli Arab passengers at Israel's airports (Supreme Court Appeal, 4797/07).

Moreover, following the analysis of passengers’ complaints and the results from the survey, it is clear that passengers believe that such security procedures cause embarrassment and ill feeling among Arab passengers. This is documented in the airport authority's own complaint files. For example, the Arab father of a student attending a European university wrote to the minister of transport to complain of his son's treatment during his security screening at the airport. In his letter, he states,

This complaint was answered by the airport's department of public affairs on April 1, 2007. In its reply, the department rejects all accusations of discrimination and racism raised by the passenger's father and emphasizes that a passenger's religion or gender has nothing to do with airport security procedures, and that the questions directed to passengers are noninvasive and are posed in a way that avoids violating human dignity. The department also reports that it conducted an inquiry among members of the security staff who had been in contact with the passenger, and it came to the conclusion that the security staff followed protocol and had not violated any guidelines. The reply also explains to the passenger's father that every car that enters the airport is stopped for examination and sometimes has to go through additional checks, in keeping with the government's instructions. Finally, it mentions that all airport security staff are instructed to adhere to the airport's ethical code while exercising their duties, and while the passenger's feelings are regrettable, the department hopes the passenger understands the security necessities at the airport and their endeavors to improve their passenger service (Department of Public Affairs, April 1, 2007).

The problem is that it is clear from our data, as well as from recent legal challenges, that Arab passengers are being profiled—and such profiling may lead to a reduction in the effectiveness of screening procedures more generally. One of the more surprising findings in our data is that both Arabs and Jews appear to have generally positive evaluations of the screening procedures. This is perhaps one of the reasons why security procedures at Ben Gurion have been so successful. Citizens appear to understand the necessity of using tough and demanding security procedures in the Israeli context. Ultimately, the highest level of security can only be ensured when everyone cooperates and participates with the security agencies. This is true in airport security as well as in policing more generally (Reference Tyler and HuoTyler & Huo 2002; Reference Tyler and WakslakTyler & Wakslak 2004; Reference LumWeitzer & Tuch 2004).

However, our data suggest that the IAA is in danger of losing the advantage that strong legitimacy evaluations confer. This is reflected in the significant differences between Jews and Arabs in basic legitimacy evaluations when other variables are not taken into account. Commentators often argue that the reasons for such differences are inherent in the more general conflicts between Arabs and Jews and the serious concerns that many Arab citizens have with Israel's Jewish national identity. Our data, however, suggest something quite different, and our conclusions have implications for security procedures in other countries.

Our data show that the nature of Arab passengers’ treatment is key to understanding their lower legitimacy evaluations. Our findings indicate that elements of the airport security process that are related to treatment (i.e., whether suitcases are opened and the perceived time that it takes to accomplish the checks) and the profiling of passengers are more important predictors of legitimacy perceptions toward airport security procedures than ethnonational identity is.Footnote ¹¹ Indeed, when these processes are taken into account, ethnonational identity no longer has a significant impact. In this sense, the IAA's response to the abovementioned passenger's complaint misses the mark. The key to legitimacy evaluations lies in passengers’ perceptions, and our study suggests that the differential treatment of Arab passengers, who generally accept the legitimacy of airport screening procedures, will likely lead to a weakening of legitimacy evaluations in the Israeli Arab community. On this point, our findings are consistent with other studies that have not found a direct influence of ethnicity on police legitimacy, especially when the effect is mediated by indicators of procedural justice (Reference Tyler and HuoTyler & Huo 2002; Reference Tyler and WakslakTyler & Wakslak 2004).

With the exception of our question regarding the opening of suitcases, our data do not necessarily tell us whether different groups are actually being treated differently; they tell us only that different groups perceive that they are being treated differently. Accordingly, we think it is important that future researchers attempt to identify objective measures of the processes underlying airport security both in Israel and in other countries. In this regard, our data suggest the importance of fair procedures, but they often do not allow us to disentangle such procedures from the perceptions of the passengers themselves. Another limitation of the current study is that it does not look at the effect of procedural justice and legitimacy on cooperation or compliance with the law. This article is focused on identifying the predictors of legitimacy; it does not address predicted behavioral effects of the airport security process. In addition, Reference Tyler, Stephen and AzizTyler, Schulhofer, and Huq (2010) suggest that when analyzing public trust, we should distinguish between public and private agencies (policing); this suggestion implies that certain kinds of policing could have a communicative, stigmatic effect, and it may be this effect—not the unfair treatment per se—that has an effect on passenger trust. We could not test this argument, since airport security in Israel is conducted solely by the state. However, we would strongly recommend that similar studies in countries that do make use of private security services distinguish between public and private agencies when analyzing the public's trust of airport security.

We further recognize that unique aspects of the Israeli case somewhat limit the conclusions we can reach in the current study. Some might argue, for instance, that Israel has a long history of terror attacks, that the acuteness of threats to homeland security is significantly unique to Israel, and that this might in turn make Israeli airport security policies irrelevant to those of other countries. Unfortunately, however, over the past decade, and especially following the events of 9/11, most Western countries have increased their focus on homeland security in general and aviation security in particular. These countries face security threats similar to those in Israel, which suggests that the current study can indeed be applied more widely.

Another limitation of the Israeli case is the nature of the Israeli Arab minority as a group that challenges some of the core values of the regime and the fact that this minority has ethnic and national associations with nations (neighboring Arab countries and Palestinians outside of Israel) that are still in an ongoing political conflict with the State of Israel. It is very hard to find similar minorities in Western democracies that simultaneously possess these two characteristics, yet, our argument that procedural justice in airport security can satisfy even this high-risk minority could be true for other minorities as well.

Finally, our analyses are observational, and though we have tried to control for the key possible confounders of ethnonational identity, only a truly randomized experiment with a random allocation of airport security procedures could provide an unambiguous answer to our research questions. The explained variance of the full model reached 31 percent, which is certainly reasonable for a criminological model (see Reference Weisburd and PiqueroWeisburd & Piquero 2008). But the large degree of variance unexplained in these models certainly raises questions regarding whether there may be important biases that we have not taken into account. Nonetheless, given present security concerns we think that such an experimental study is unlikely at this time, and we think that our data provide important insights into the role of procedural justice in security procedures at airports. Although we could identify only one other study on this issue, both that study and ours reinforce findings in policing that suggest that just procedures are the key to increasing legitimacy perceptions among ordinary citizens (Tyler 2004b; Reference Tyler and HuoTyler & Huo 2002; Reference Tyler and WakslakTyler & Wakslak 2004).

Conclusions

Ensuring the safety of air travel systems has become much more prominent since September 11, 2001, and the level of passenger security screening has been raised throughout the world. Nowadays, passengers in every major Western airport are subjected to heightened levels of security screening that not only are inconvenient but also raise important questions about the treatment of members of specific groups that are seen as presenting special security risks. In this context, Israeli security procedures are seen as particularly stringent and thus provide an important site for examining questions about the effects of heightened screening on passengers.

While the majority of Israeli Arabs and Jews recognize the importance of heightened security procedures and have high legitimacy evaluations of the screening process, there are significant differences between Jews and Arabs in their evaluations of its legitimacy. For some Israeli Arabs, the experience of passing through security at Ben Gurion Airport is perceived as unpleasant. The results of our research indicate that this unpleasantness is by and large the result of the experiences of Arab passengers and not a result of their ethnonational identity and its conflicts with the Jewish character of the Israeli state. We showed that once we take into account the perceived quality of the security check, which is often defined as the procedural justice of the process, the differences in legitimacy evaluations between Jews and Arabs are no longer significant.

Just as with policing, our study suggests that when citizens are treated respectfully and impartially in airport screening, their legitimacy perceptions will be high. The underlying components of the Israeli-Arab conflict, often seen as so central in Israeli social life, appear to be outweighed by the simple aspects of how people are treated and how they perceive those processes. This suggests important policy proscriptions for Israel, and indeed for airport screening procedures all over the world. In their efforts to focus on prevention, which often include profiling and embarrassing public procedures, security agencies may actually jeopardize passengers’ trust in airport security. Such security is dependent on the cooperation of citizens, and heightened security procedures focused on particular groups may lower legitimacy evaluations and thus the cooperation of the public. The overall positive evaluations of Israeli Arabs suggest that the IAA has much goodwill to work with. We suspect that, given the risks of airline terrorism, this is true all over the world. Nonetheless, that goodwill may be squandered if security agencies do not pay figure out how to maximize security while maximizing procedural justice.

Our results point to fact that the most sensitive stage in the airport security procedures is the public interaction between the passenger and the security personnel, especially the high visibility of the security checks and the way in which they are carried out. One possible way to reduce the negative side effects of the security process is to limit the visibility of the process by relying more often on technology—for example, biometrics, computerized tomography (CT) scans, and magnetic resonance imaging (MRI). Another area in which effort should be made is in improving security personnel training programs by focusing on the importance of procedures that encourage evaluations of legitimacy among passengers.