Is It the What or the How? The Roles of High-Policing Tactics and Procedural Justice in Predicting Perceptions of Hostile Treatment: The Case of Security Checks at Ben-Gurion Airport, Israel

Abstract

What affects perceptions of hostile treatment by police, characterized by feelings such as humiliation and intimidation? Is it what the police do to the citizen, or is it about how they do it? The important effects of procedural justice are well documented in the policing literature. Yet, it is not clear how high-policing tactics, coupled with procedural justice, affect one's sense of hostile treatment: is it the case that what the police do does not matter as long as they follow the principles of procedural justice, or do some invasive or unpleasant tactics produce negative emotions regardless of the amount of procedural justice displayed by the officer? In the present study we examine this question in the context of security checks at Ben-Gurion Airport, Israel. Using a survey of 1,970 passengers, we find that the behavioral elements of procedural justice are an important antidote, mitigating the negative effects of four “extra” screening measures on the perceived hostility of the checks. At the same time, two security measures retain an independent and significant effect. We discuss the implications of our findings and hypothesize about the characteristics of policing practices that are less sensitive to procedural justice.

Following recent protests and civil disorders in Ferguson, Missouri and other American cities, President Barack Obama established a Task Force on 21st Century Policing. “When any part of the American family does not feel like it is being treated fairly [by law enforcement], that's a problem for all of us,” he stated (President's Task Force on 21st Century Policing 2015: 5). One of the recommendations of the Task Force was that “…law enforcement agencies should adopt procedural justice as the guiding principle for internal and external policies and practices to guide their interactions with rank and file officers and with the citizens they serve” (President's Task Force on 21st Century Policing 2015: 1). Procedural justice refers to the fairness embedded in the processes by which power holders exercise their authority, and involves behaviors such as respectful speech, inviting citizen input, and transparency in decision making. Such behaviors, in turn, were found to promote highly desirable outcomes, including an overall sense of fair treatment, legitimacy, satisfaction, and willingness to comply and cooperate with police officers (e.g., Blader & Tyler 2003a, 2003b; Jonathan-Zamir et al. 2015; Murphy 2014; National Research Council 2004; Schulhofer et al. 2011; Sunshine & Tyler 2003; Tyler 2004, 2009, 2011).

On December 30, 2014, the same month the Executive Order establishing the Task Force was signed, a police officer in New Jersey stopped a vehicle for allegedly driving through a stop sign. The officer politely greeted the occupants of the vehicle, identified himself by name and affiliation, and explained why they were being stopped. When one of the passengers asked a question about the location of the stop sign, the officer patiently explained. He was polite and calm. Yet, at least one of the passengers did not appear to fully comply with the officer's demands, and less than 90 seconds later, Jerame Reid, a 36-year old Black man, was shot dead by the officers who feared that he was reaching for a gun (Sanchez 2015). This event is one (extreme) example of the complexity of police-citizen interactions. The procedural justice model in its most simple form would have predicted that the officer's behavior would lead to citizen compliance and cooperation, which would have made the use of force unnecessary. Yet, like many police-citizen interactions, this encounter clearly involved other factors that led to the unfortunate outcome. Recognizing that the outcomes of procedurally-just treatment vary, important questions can be asked about this variation and the factors that affect it.

In this article we focus on one expected outcome of police-provided procedural justice—the mitigation of negative emotions that may arise in police-citizen interactions: humiliation, intimidation, and a sense of indifferent treatment (referred to herein as perceptions of hostile treatment); and one factor that is expected to elicit such negative emotions: the use of “high-policing” practices. High policing addresses strategic problems at the macro-level, such as national security (rather than local crime and disorder problems), and is characterized by tactics that are less transparent and accountable, and thus prone to violations of human rights and due process (e.g., Bayley & Weisburd 2009; Brodeur 1983, 2010; Brodeur & Dupeyron 2003). We ask if the behavioral elements of procedural justice (including treating citizens with respect, concern, and transparency) “neutralize” the expected negative outcomes of high policing, or, alternatively, if some intrusive tactics produce a sense of hostile treatment independent of the “amount” of procedural justice embedded in the officer's behavior.

We address this question in the context of airport security checks, which, since the terror attack of September 11, 2001, have become both very common and potentially invasive, humiliating and threatening (Hasisi et al. 2012). We use a survey of 1,970 passengers boarding an airplane at Ben-Gurion Airport (Israel) to assess the effects of four “extra” security measures (being asked a large number of questions; being questioned by many security officers; having one's suitcase opened and searched; and being questioned away from the other passengers) on perceptions of hostile treatment (feeling humiliated, intimidated, and being treated with insensitivity), while accounting for the procedural justice displayed by the security officers (as experienced by the passenger).

We begin with a review of the concept of procedural justice, its behavioral elements, and the framework of “high policing.” We then raise our main research question—what are the roles of high-policing tactics, coupled with procedural justice, in producing subjective perceptions of hostile treatment? We review recent debates on the interplay between what the police do and how they do it in the context of police street/car stops. We then continue with a description of our study setting—security checks at airports—and argue that while airport security is a form of “high policing,” it is critical that passengers feel positively about the process. Next, we describe our study procedure, sample, questionnaire, and construction of main variables, and present our analysis, in which we find that although procedural justice is clearly important in mitigating an overall sense of hostile treatment, it does not eliminate the negative effects of two invasive security measures: having the passenger's suitcase opened and searched, and being questioned away from other passengers, were found to strengthen feelings of hostile treatment after controlling for how the passenger felt she was treated. In the Discussion we speculate on the characteristics of policing practices that are less sensitive to procedural justice and suggest policy implications.

The Notion of Procedural Justice

The term “procedural justice” has been drawing much attention in recent policing scholarship. Originating from the seminal work of Thibaut and Walker (1975), and later developed in the context of policing by Tom Tyler and many others, “procedural justice” (also termed “procedural fairness” or “fair treatment”) (Elliot et al. 2011) refers to the fairness embedded in police processes and behaviors. It is not concerned with the desirability of the outcomes delivered by police, but with the process itself, as evaluated by citizens regarding a particular encounter with the police or police treatment more generally (Braga et al. 2014; National Research Council 2004; Schulhofer et al. 2011; Sunshine & Tyler 2003; Tyler 2004, 2009, 2011; Tyler et al. 2014b).

Thibaut and Walker (1975) identified that people are more willing to accept undesirable outcomes when they view the process that led to these outcomes as fair. They proposed that fair procedures give people a sense of control over the process (e.g., by being invited to express their opinions), which, in turn, should make the outcomes fair. Over time, this framework was criticized for ignoring other concerns people may have with procedures. In response, Lind and Tyler (1988) developed the “group value” model of procedural justice, later expanded to hierarchical settings in the “relational model” (Tyler & Lind 1992). At their core is the proposition that groups do not just provide their members with instrumental outcomes, but also form a setting in which people construct their identities (Tajfel & Turner 1979, 1986). Thus, fair processes are important, in large part, because they influence one's judgments about her identity and status. They send the message that the group values the individual, who, in turn, feels good about herself and about being part of the group (Blader & Tyler 2003b; Bradford et al. 2014; Tyler & Blader 2013).

Perceptions of fair treatment were found to have two important benefits for police. The first is expressed in specific encounters between police officers and members of the public. When the citizens involved feel that they have been treated in a procedurally-just manner, they are more likely to behave in desirable ways, including complying with police requests, accepting the outcome of the encounter even if undesirable, and providing information and assistance to the officers (Mastrofski et al. 1996; McCluskey 2003; Paternoster et al. 1997; Tyler & Huo 2002). Second, general assessments of police procedural justice were found to be strongly correlated with evaluations of police legitimacy (e.g., Elliot et al. 2011; Gau et al. 2012; Kochel et al. 2013; Mazerolle et al. 2013; Murphy et al. 2008; Schulhofer et al. 2011; Sunshine & Tyler 2003; Tyler 2001, 2004, 2009, 2011; Tyler & Fagan 2008; Tyler & Wakslak 2004; Tyler et al. 2014b). Legitimacy assessments, in turn, were found to predict important outcomes in terms of police success, including willingness to cooperate and comply with the police, provide information, assist in solving crime, empower the police, and even obey the law more generally (Jackson et al. 2012, 2013; Murphy 2014; Sunshine & Tyler 2003; Tyler 2006; Tyler & Fagan 2008; Tyler & Huo 2002; Tyler et al. 2007, 2014b).

The strong relationship between procedural justice and police legitimacy is the main finding emerging from the literature, consistently replicated in different contexts and countries, including Israel, our study site (Factor et al. 2014; Jonathan-Zamir & Weisburd 2013). Nevertheless, it should be noted that some studies find variation in the relative importance of procedural justice versus instrumental considerations in predicting police legitimacy (e.g., Brockner et al. 2001; Hinds & Murphy 2007; Tankebe 2009b), while others criticize the measurement and operational definitions of key terms within this framework (Gau 2011; Maguire & Johnson 2010; Reisig et al. 2007; Tankebe 2013). Further elaboration is beyond the scope of the present review; however, it should be emphasized that despite the ongoing debate, it is rarely disputed that perceiving police as exercising their authority using fair processes is an important, positive outcome, for both practical and normative reasons (e.g., Jonathan-Zamir et al. 2015; Lum & Nagin forthcoming; Tankebe 2009a).

The Behavioral Elements of Procedural Justice

What makes individuals feel that they were treated fairly? Based on early procedural justice research (Lind & Tyler 1988; Thibaut & Walker 1975; Tyler et al. 1996), Blader and Tyler (2003a, 2003b) identify two main components to procedural justice: the quality of the decision-making procedures and the quality of interpersonal treatment. Information related to decision making presumably signals as to the fairness of the outcomes of the encounter, and includes both participation (or voice)—enabling the citizens involved to express their opinions, and taking this input into account when making decisions, and neutrality—making decisions in a neutral, unbiased fashion. Neutrality is often expressed in transparent decision making, because explaining the decisions provides power holders with the opportunity to communicate that they were using objective, unbiased criteria. The type of treatment received sends a message about the “social atmosphere of the group or situation” (Blader & Tyler 2003a: 748), and fair treatment suggests that the group recognizes the status of the individual and values her. This function includes dignity and respect—treatment with politeness and dignity, while acknowledging and respecting the rights of those involved, and displays of trustworthy motives—showing that the police are truly concerned with the well-being and quality of life of the citizens involved and/or the larger public (Blader & Tyler 2003a, 2003b; Jonathan-Zamir et al. 2015; Murphy 2014; National Research Council 2004; Schulhofer et al. 2011; Sunshine & Tyler 2003; Tyler 2004, 2009, 2011).

The literature suggests that these four behavioral elements are consistent predictors of subjective, overall feeling of fairness, independent of the race, class, or ideology of the individual (Sunshine & Tyler 2003; Tyler 1988, 1994, 2003; Tyler & Huo 2002). At the same time, variations have also been noted. For example, authorities’ compliance with ethical principles and their ability to correct errors were also found to affect fairness. Moreover, the effects of the four behavioral elements were found to depend on the situation in which citizens encounter authorities: dispute management versus assistance with problems (Lind & Tyler 1988; Lind et al. 1997; Tyler 1988). At the same time, the policing literature has done little to illuminate such variation, or identify factors that may work alongside procedural justice in promoting (or impending) its expected desirable outcomes. In the present study, we ask: what is the role of what the police do (in addition to how they do it) in predicting perceptions of hostile treatment, perceived as humiliating, intimidating, and insensitive?

“High-Policing” Practices and Perceptions of Hostile Treatment

In the present study we focus on one component of what the police do that may produce perceptions of hostile treatment—“high-policing” practices. Brodeur (1983, 2010) explains that in contrast to “low policing,” which is focused on “classic” local crime and disorder problems, the goal of high policing is to protect the “state” or “national security” (even if this aim coincides with protecting citizens, as is the case with counterterrorism). Developing from this unique orientation, the tactics of high policing are also different from everyday policing practices. They are characterized by the collection of data and it analysis for the purpose of creating intelligence and threat assessments, and frequently involve various surveillance methods. It is therefore not surprising that high-policing tactics are often cloaked in secrecy and deception (Brodeur 2010), and are thus less accountable, and have been associated with violations of human rights, due process, and procedural justice (Bayley & Weisburd 2009; Brodeur 2010).

Zooming in on the practical expressions of high policing, we propose that police-citizen encounters that develop from high-policing agendas, such as demanding security checks at airports (see below) or intensive questioning and searching as part of counterterrorism operations, are likely to be experienced as particularly hostile—threatening, humiliating, and insensitive. We base this proposition on two main arguments, both developing from the characteristics of high policing. First, we can expect high-policing encounters to be particularly intrusive, both because in high policing the collection of intelligence is an end in and of itself (in contrast to low policing, where information is gathered for the purpose for building criminal cases; see Brodeur 2010); and because high policing often aims to create what Brodeur (2010: 230) calls a “chilling effect”—temporarily providing information (or disinformation) in order to create intimidation. Second, because of the covert nature of high policing, the citizens involved often do not know why they are being singled out, questioned and/or searched, and may (correctly or not) attribute their selection to profiling, prejudice, or discrimination by law enforcement. Minorities who experience ongoing discrimination in multiple areas of life are particularly prone to such interpretations, because the interaction reminds them, or draws attention to, the continuous inequity they feel (see the “expressive harm” hypothesis) (Risse & Zeckhauser 2004). Nevertheless, we ask if such practices have an independent effect on perceptions of hostile treatment, or, alternatively, if they impact feelings of humiliation, intimidation, and insensitive treatment through the behavioral elements of procedural justice. In other words, is it the case that when “high-policing encounters” are carried out in a procedurally-just fashion, they do not elicit negative emotions, as sometimes suggested in the literature (e.g., Schulhofer et al. 2011; Tyler & Wakslak 2004), or, alternatively, do some policing practices produce negative evaluations regardless of how they are carried out?

As reviewed above, the literature suggests that high-policing practices stand in contrast to procedural justice. Yet, few studies demonstrate that even when perusing a high-policing agenda such as counterterrorism, treating citizens with procedural justice can be of great value (e.g., Huq et al. 2011; Tyler et al. 2010). At the same time, we are unaware of empirical studies that have directly tested the effects of high-policing tactics, coupled with procedural justice, on the expected outcomes of the later. Thus, in addition to illuminating the effects of factors other than procedural justice on its expected outcomes, our question has direct policy implications: is it the what or the how? In order to mitigate the negative emotions that may arise in some police-citizen encounters, do police only need to learn how to apply the principles of procedural justice to unpleasant tactics, or do they need to rethink the use of some of these practices altogether? Before turning to the specific context of our study—airport security checks—we review recent discussions on our main research question. These debates were carried out in the context of “stop, question, and frisk.”

Are Perceptions of the Treatment About What the Police Do or About How They Do It? The Case of “Stop, Question, and Frisk”

Questions about the interplay between what the police do and how they do it were recently raised as part of an important and timely debate on police street/car stops (including “stop, question, and frisk” and “Terry stops”; see Epp et al. 2014; Gau 2013; Gau & Brunson 2010; Meares 2014; Tyler et al. 2014a). Epp et al. (2014) assert that advocates of this policing practice encourage police to carry out the stops in a highly respectful manner, assuming (based on the theory of procedural justice) that being stopped would not bother people as long as they are treated “fairly.” At the same time, they find that African Americans in Kansas City judged the acceptability of the stop based not only on politeness, but also on the type of stop—investigatory versus traffic—where the former was perceived as illegitimate and undermined acceptability. As pointed out by Meares (2014), they only considered the dignity/respect element of procedural justice. Nevertheless, their results suggest that treating individuals with procedural justice may not always be enough.

Focusing on traffic stops, Gau (2013) finds that requesting a consent to search the motorist's vehicle significantly undermined the legitimacy of the stop, even after controlling for procedural justice. Thus, Gau (2013: 772) concludes that “…procedural justice is not enough to wash out the deleterious impact of consent search requests, so it is important that officers recognize that they cannot nullify the negative effects of consent requests by being polite and respectful.” Gau also finds that requesting consent to search and various reasons for the stop undermine feelings of fairness, but did not control for how the stop was carried out (the behavioral elements of procedural justice).

Tyler et al. (2014a) recently found that for young men in New York City, the level of exposure to policing (number of street/car stops in the past year and assessment of police intrusiveness) did not have a significant effect on legitimacy once three aspects of “fairness” were controlled for: procedural justice, the fairness of the outcomes, and the legality of the stop. In contrast to Gau (2013), they conclude that “it was not street stops per se, or even the intrusions that they make into people's lives, but whether people evaluate police actions as involving fair interpersonal treatment and appropriate justification” (Tyler et al. 2014a: 775). At the same time, when predicting both encounter-specific and broad procedural justice, increasing number of stops had a significant and negative effect: with each additional stop, respondents were less likely to experience the stop, and police processes overall, as fair. Similar to Gau (2013), the behavioral elements of procedural justice were used in these models to define procedural justice, not as independent variables in a model predicting an overall sense of the treatment. Nevertheless, these findings suggest that when trying to understand citizens’ evaluations of police treatment, we should consider the practices the police use in addition to how they use them. In our study, we directly examine the effects of high-policing practices on perceptions of hostile treatment while accounting for the procedural justice embedded in the officers’ behavior. We do so in the context of airport security.

The Context of Airport Security

Since the terror attack of September 11, 2001, security checks have become a salient and demanding component of passengers’ airport experience (Sindhav et al. 2006). What is more—they are extremely frequent. Lum et al. (2015) report that in the United States, in 2011, airplanes were boarded over 700 million times. In Israel, our study site, which included a little over 8 million residents in 2014 (CBS Israel 2015), airplanes were boarded almost 7 million times that year. Each of these boardings was preceded by a series of security checks. Given the prevalence of airport security checks and the burden they put on passengers, and given the argument that airport security is in essence a form of policing practice (Hasisi & Weisburd 2011; Lum et al. 2015), it is not surprising that airport security has become an important topic of inquiry within the framework of legitimacy and procedural justice, often with reference to inequity, ethnic profiling, and differential treatment of Arab or Muslim passengers (Hasisi & Weisburd 2011; Hasisi et al. 2012; Lum et al. 2015; Sindhav et al. 2006).

Airport Security as a Form of “High Policing”

We propose that airport security is a form of high policing, and may thus elicit feelings of hostile treatment. First, in line with the characteristics of high policing reviewed earlier, the goal of airport security is to protect national security, as it is expressed in the prevention of aviation terrorism. Specifically in Israel, the primary responsibility of the Security Division of the Israel Airport Authority (IAA) is to identify, prevent, and handle terrorist threats that target departing flights or the airport itself. Second, although security personnel are employees of the IAA, their guidelines and methods of operation, including risk assessments and the intelligence on which they are based, are solely determined and provided by the Israel Security Agency (ISA—the “Shabak”), which is clearly a high-policing agency tasked with protecting internal security.¹

Additionally, in line with other characteristics of “high policing,” airport security procedures are cloaked in secrecy and sometimes involve intimidation. In Israel, in 2011, the Minister of Defense signed a certificate of immunity preventing the guidelines and procedures for conducting airport security checks to be presented as evidence in court [HCJ 4797/07 The Association for Civil Rights in Israel v. Israeli Aviation Authority (March 10, 2015)]. They are thus concealed from the public by law. Additionally, the process itself is not transparent to passengers. For example, in the past, following preliminary basic questioning (see below), passengers at Ben-Gurion Airport received a sticker from the security officer. The color of the sticker indicated (in all likelihood) the level of risk they posed. The goal of the sticker, the meaning of its color (which changed from time to time), and the reason for being assigned to a particular risk category, were concealed. Further, as detailed in HCJ 4797/07 The Association for Civil Rights in Israel v. Israeli Aviation Authority, once classified as “high risk,” the passenger is asked to undergo several stages of additional screening, often with prolonged periods of wait between them. These may include sequential questioning by different security officers in different locations of the airport, a search of one's belongings in public, or a thorough body search in an isolated room. The entire process is typically not accompanied by explanations concerning its purpose, stages, or expected duration. From the perspective of the passenger, it often involves feelings of frustration, anger, humiliation, and intimidation, as well as nerve-racking waiting and uncertainty about whether or not she would successful complete the process in time to board her flight.

It should be noted that a new security system, which should minimize passengers’ inconvenience, was introduced at Ben-Gurion Airport in 2015. In the new system, all luggage is screened automatically underground, and only in suspicious circumstances a thorough manual search is carried out, away from public view. In addition to improving security, the system portrays a picture of equal treatment (at least on the surface), and is expected to eliminate the humiliation of having one's belongings searched in public. While this may indeed be the case, this system also includes obvious characteristics of high policing, including searching one's personal belongings without her presence and concealing the criteria for manual searches (HCJ 4797/07 The Association for Civil Rights in Israel v. Israeli Aviation Authority).

Passengers’ Perceptions of Airport Security Checks

Despite the inherent difficulties, we argue that it is critical that passengers do not board their flight feeling that the security screening process they experienced was humiliating, intimidating, or insensitive to their needs and concerns. First, like police and other public service providers in democratic societies, airport security authorities operate in a sociopolitical environment that requires balancing security with fairness (Amir & Einstein 2001; HCJ 4797/07 The Association for Civil Rights in Israel v. Israeli Aviation Authority; Lum et al. 2015). Second, based on the vast literature on legitimacy and procedural justice, it is expected that perceiving the security checks positively would lead to numerous desirable outcomes, including passengers’ willingness to comply and cooperate with security personnel, provide them with information and obey their directives (Murphy 2014; Tyler et al. 2014b). Finally, passengers’ feelings are important for pure financial reasons: Sindhav et al. (2006) identified a relationship between four dimensions of fairness (distributive, procedural, interpersonal, and informational) and overall satisfaction with the service experience at the airport. Like other competitive markets, unhappy customers may seek alternatives, which is clearly a highly undesirable outcome from the standpoint of the aviation industry.

Returning to our main question, although research is scarce, there is evidence to suggest that some airport security measures may undermine the fairness or legitimacy of the process, even if the principles of procedural justice are followed. Hasisi and Weisburd (2011) carried out a survey of over 600 passengers boarding an airplane at Ben-Gurion Airport immediately after passing through security. Their analysis showed that opening the passenger's suitcase and searching its content had a significant and negative effect on trust in the security inspectors, after controlling for the passenger's ethnicity (Jewish/Arab; a critical variable in this context) and perceptions regarding the fairness of the checks. This finding was replicated by Hasisi et al. (2012) in an analysis that included foreign passengers and predicted overall satisfaction with the security process. The effect was particularly strong for Arab passengers, a finding attributed to the “expressive harm” hypothesis (Risse & Zeckhauser 2004).

Notably, in both models the perceived fairness of the checks had the strongest effect on trust/satisfaction, which suggests that the core elements of the legitimacy model behave in the setting of airport security as they do in other policing contexts (see above). At the same time, these findings again suggest that although procedural justice is clearly critical, the tactics being used should also be considered when assessing the emotional outcome of the encounter, because some policing (or airport security) practices appear to elicit negative emotions (or undermine positive ones) even when executed in a procedurally-just fashion.

The Study

Study Context

Similar to Hasisi and Weisburd (2011) and Hasisi et al. (2012), we have carried out a survey of passengers boarding an airplane at Ben-Gurion Airport. Ben-Gurion is Israel's primary international airport, often referred to as the gate to the country. In 2014, over 14 million passengers entered or exited Israel through this airport, in over 100,000 flights (IAA 2015). Security at Ben-Gurion Airport is carried out in four circles of intervention (see review by Hasisi et al. 2012). In the present study, we focus exclusively on the third—security checks at the terminal, before check-in. As evident from publicly available sources (e.g., Duek 2013; Hasisi et al. 2012), these checks begin with a passport examination by a security officer and very brief questioning (roughly 30 seconds), including questions such as “where are you flying to?”; “where did you come to the airport from?”; “did you pack alone?”; “were you asked to deliver anything?” Following this short, often neutral conversation, passengers are typically directed to the check-in counters to continue the boarding process.

However, for passengers classified as “high risk,” the security inspection is only beginning. As detailed in HCJ 4797/07 The Association for Civil Rights in Israel v. Israeli Aviation Authority mentioned earlier, some passengers (mostly Arabs) may be subjected to extensive questioning by several security officers, one after the other, first in proximity to the main line but then (possibly) in different offices at the terminal. Their suitcase is opened and searched in sight of other passengers. Personal belongings are individually checked and placed on a table, exposed to all. Suitcases are screened, emptied, and screened again (as indicated above, this visible procedure has recently changed; however, these practices were still in use during our survey period). Some passengers are asked to undergo a thorough body search in a private room. The process may take several hours, often including long waiting periods between its stages. From the passenger's perspective, feelings of anger, frustration, uncertainty, humiliation, intimidation, and concerns about missing one's flight, were frequently noted. It is thus clear how these “extra” high-policing measures may elicit a sense of hostile treatment. But what role, if any, do the behavioral elements of procedural justice play in mitigating emotions such as humiliation and intimidation? In our analysis below, we test the effects of these “extra” measures, and the behavioral elements of procedural justice as experienced by the passenger, on evaluations of hostile treatment.

The Survey and Variables

Our survey was carried out in two waves (July–August, 2013 and October, 2014).² Following previous research in this area (Hasisi & Weisburd 2011), we have randomly sampled passengers from three population groups: Israeli Jews, Israeli Arabs, and foreign passengers. Our sampling was stratified to ensure a large enough sample from each group to allow for subsequent analyses within the groups. Passengers were randomly approach by our first research team shortly after passing through security (while in line for passport control), were asked to participate in a survey about their airport experience, and were offered a voucher for coffee and cake at one of the duty-free cafés to encourage participation. Passengers who agreed were given a special ticket, with which they approached our second research team located in the duty-free area, where they filled out the questionnaire and received the voucher. Questionnaires were available in four languages—Hebrew, English, Arabic, and Russian. Similar to Hasisi and Weisburd (2011), we estimate the response rate at about 40 percent. Passengers who refused to participate did not appear systematically different in their visual characteristics (sex, age, ethnicity, overall mood) from those who agreed. Our final sample consists of 1,970 passengers, of whom 660 are Jewish Israelis, 687 Israeli Arabs, and 623 foreigners.³ See Appendix for other sample characteristics.

The questionnaire included a total of 42 items. In the first section, respondents were asked to rank their agreement with statements on a scale of 1 to 5 (1 = do not agree; 5 = completely agree). These statements tap the main themes measured in previous analyses of police legitimacy, including trust, procedural justice, effectiveness and willingness to cooperate (e.g., Gau 2011; Jonathan-Zamir & Weisburd 2013; Reisig et al. 2007; Sunshine & Tyler 2003; Tyler & Wakslak 2004), but with reference to the security authorities at Ben-Gurion Airport. In the second section, passengers were asked about the “technical” features of the security checks they were subjected to today, such as the number of questions they were asked and if their suitcase was opened and searched. These questions were designed based on publicly available descriptions (such as HCJ 4797/07 The Association for Civil Rights in Israel v. Israeli Aviation Authority mentioned above), with the intent of capturing the extent to which the passenger was subjected to the additional screening measures. At the same time, highly sensitive questions that may cause unease (e.g., “were you subjected to a full body search?”) were avoided. The final section inquired about sociodemographic and flight characteristics (such as travel reason and number of flights taken in the past year).

Our analysis includes two scales: the dependent variable—the perceived hostility of the security screening process; and one of the two main independent variables of interest—the behavioral elements of procedural justice. The construction of the scales was based on their conceptual meanings as reflected in the procedural justice literature and in publicly available testimonies of high risk passengers regarding their feelings about the security checks (see above). The discriminant validity and internal consistency of the scales were confirmed using factor analysis and measures of Cronbach's Alpha. Table 1 reports the specific survey items that make up each scale, the factor analysis, Cronbach's Alpha values, and the descriptive statistics of the scales.

Table 1. Factor Analysis Differentiating Procedural Justice and Perceptions of a Hostile Security Screening Process, as Experienced by the Passenger

^aExtraction method: Principal Component Analysis; Rotation method: Varimax with Kaiser Normalization; Kaiser-Meyer-Olkin measure of sampling adequacy = 0.89; N = 1694. Only factor loadings > 0.40 are displayed.

^bIn the procedural justice factor, the parentheses indicate the component of procedural justice captured by the item.

There are no items reflecting participation in our procedural justice index, because it does not appear to translate to the context of airport security. First, it displays little variability, because as part of the routine 30-seconds questioning, all passengers are asked to indicate where they are flying to, who packed their bags and so forth. Second, in the context of police patrol, “participation” is typically expressed in behaviors such as the officer asking the citizen about the reasons for her behavior (e.g., “Why were you speeding? Were you late for work or something?”), her side in a dispute, reason for calling the police, and so forth (see Jonathan-Zamir et al. 2015). We did not find suitable parallels in the context of airport security. We further discuss the implications of our operationalization of procedural justice in the Discussion section below.

With regard to our dependent variable, we should note that evaluations of the hostility of the process were relatively low, indicating overall satisfaction: on a scale ranging from 4 to 20, over 17 percent had the score of 4 (indicating the lowest level of hostility), and 45.7 percent had no more than 7. Only 7 percent of passengers were in the upper quartile of the distribution, with the score of 15 or above. This finding is in line with Hasisi and Weisburd (2011), who found that the majority of passengers in their sample expressed trust in the security inspectors. At the same time, as reviewed above, we expected feelings such as humiliation and intimidation to intensify if the passenger was subjected to one or more of four “extra” security procedures: being asked more than 10 questions during the security checks (“yes” = 9.5 percent; N = 1,848); being questioned by more than three security officers while waiting to check in (“yes” = 5.3 percent; N = 1,893)⁴; having the passenger's (or his/her companion's) suitcase opened and searched (“yes” = 23.9 percent; N = 1,943); and whether part of the passenger's security checks took place in a different area, away from the other passengers (“yes” = 14 percent; N = 1,947).⁵

Most of the passengers in our sample (67.2 percent) were not subjected to any of these extra measures. 19.2 percent were subjected to one extra measure; 9.2 percent to two, 3.7 percent to three, and only 0.7 percent were subjected to all four. Notably, in line with prior literature on profiling in airport security (e.g., Hasisi et al. 2012), our data indicate that passengers were not subjected to these additional checks at random. As evident from Table 2, passengers who were subjected to at least one extra measure tended to be Israeli-Arab or foreign passengers (as opposed to Israeli Jews). They were also likely to be unmarried, young, male, more religious, fly alone, and for reasons other than tourism. Additional screening was also more frequent during our first survey period. Thus, as detailed below, we control for these and other factors in our final model.

Table 2. Characteristics of Passengers Who Were Subjected to At Least One “Extra” Security Measure

^aThe overall sample consists of 1970 individuals; for each variable the valid N is reported.

^bAll differences are significant at the 0.001 level.

Similar to Hasisi and Weisburd (2011), our control variables include the characteristics of the specific flight passengers were about to board: their reason for travel; whether or not they are flying alone; and the number of flights they have already taken in the past year (we suspected that “frequent flyers” may view the security checks differently—either more negatively due to an “accumulation effect,” or more positively due to habituation). We also inquired about sociodemographic characteristics that are typically measured in attitude surveys, such as sex, age, and education. Importantly, we control for the population group the passenger belongs to (Israeli Jewish; Israeli Arab; foreign nationality), particularly because past research indicates that Israeli Arabs are often subjected to heightened security screening and view the security authorities less positively. Finally, we control for the survey wave and for whether or not the passenger felt that given the current security situation, the security checks are justified (see Hasisi & Weisburd 2011; National Research Council 2004). See Appendix for descriptive statistics of the control variables.

Analytic Strategy

As illustrated in Figure 1, we treat the entire security screening process at the airport as a single “package” the passenger receives, which in the end leaves her with an overall “conclusion” regarding the hostility of the process, as expressed in feelings of humiliation, intimidation, and insensitive treatment. Importantly, the content of this “package” varies from passenger to passenger. It is made up of the “technicalities,” that is what was done to the passenger, including the basic checks (which do not vary across passengers) and the “extra” security measures (to which only some passengers are subjected, to varying degrees); and the procedural justice element, which indicates how the passenger felt she was treated overall, with reference to both the basic and “extra” checks. Clearly, the security screening process involves multiple stages that may vary in the “amount” of procedural justice the passenger receives. In this analysis, we make no pretense to identify the unique level of procedural justice embedded in each stage, but rather capture an overall assessment concerning the behavioral elements of procedural justice the passenger was exposed to in the security screening process as a whole (the how). This assessment, along with the “technical” aspects of the checks (the what), are used to predict eventual feelings about the hostility of the process.

Figure 1. Analytic Strategy.

We use a series of Ordinary Least Squares (OLS) regression models to identify the independent effects of the “technical” features of the screening process, and procedural justice, on negative emotions elicited by the security checks. First, we test whether being subjected to any of the four “extra” security measures produces perceptions of a hostile process. Because being subjected to additional screening is confounded with several sociodemographic and flight characteristics, in the second model we add the control variables. Finally, in the third model we add the behavioral elements of procedural justice, and test whether any negative effects found in the first and second models disappear, or are at least moderated, once the way the security screening process was carried out (in the eyes of the passenger) is controlled for.

Findings

We present our findings in Table 3. It reveals that all three models are statistically significant, and explain 10 percent (model I); 27 percent (model II) and 40 percent (model III) of the variance in passengers’ evaluations of the hostility of the security screening process. All R ² changes are statistically significant at the 0.001 level. In model I, tolerance levels were larger than 0.78 for all variables. In models II and III, they were larger than 0.56, suggesting that multicollinearity is not a problem in these models (O'Brien 2007). Model I shows that, as expected, each of the four extra screening measures has a statistically significant and positive effect on assessments of a hostile process. Because of the confounding between being subjected to additional screening and several background characteristics, the effects of all four security measures are smaller in model II, and two security measures—being asked more than 10 questions during the security checks and being questioned by more than 3 security officers—lost their significance. This model also shows that several of the control variables significantly affect perceptions of a hostile process, particularly beliefs about the security checks being justified and the passengers’ ethnicity.

Table 3. Predicting the Overall Hostility of the Security Checks (Humiliation; Intimidation; Insensitive Treatment)

Notes: ^*p ≤ 0.05; ^**p ≤ 0.01; ^***p ≤ 0.001.

Model III directly addresses our research question—what are the roles of both the additional screening measures (what was done to the passenger), and procedural justice (how the passenger felt she was treated), in predicting assessments of a hostile process? As can be expected from the vast literature on procedural justice, it is clearly critical: once included, the variance explained by the model rises by about 50 percent (from 27 to 40 percent). Moreover, procedural justice shows the strongest effect on hostility (as indicated by the beta value of −0.47). Finally, model III shows that procedural justice is important in mitigating the negative effects of all four extra security measures, as reflected in smaller regression coefficients: holding all other variables constant, in model II a passenger subjected to all four extra measures would score 2.8 points more on the “hostility scale” (range: 4–20). Once procedural justice is accounted for (model III), the magnitude of this effect drops by over 50 percent (to 1.3). At the same time, procedural justice was not enough to extinguish the negative effects of two security procedures. Opening and searching the passenger's suitcase, and questioning the passenger in an isolated area, show a statistically significant effect on hostility, even if security personnel are perceived to be exercising their authority without bias, with dignity and respect, and while displaying care and concern.

Finally, although not our main focus, it is noteworthy that several control variables are associated with evaluations of a hostile security screening process, regardless of the “technical” features of the checks or the fairness embedded in the process. Especially noteworthy are Arab passengers, who inherently judge the security process as more hostile, and passengers who believe that given the present security situation the security checks are justified, who tend to view the security checks more positively.

Discussion

Our findings do not contradict popular arguments concerning the importance of procedural justice in mitigating feelings such as humiliation, intimidation, and a sense of indifference in police treatment, but suggest that in addition to how, what police do also matters. In line with past research (e.g., Sunshine & Tyler 2003; Tyler & Huo 2002), in our model the behavioral elements of procedural justice were the primary antidote to an overall sense of hostile treatment. At the same time, we also find that searching the passenger's suitcase, and checking her in a different area of the airport, significantly affects her sense of a hostile treatment, independent of the amount of procedural justice experienced. Thus, our findings provide an important example demonstrating the need to consider the tactics used by police when examining the emotional outcomes of police treatment. The context of airport security provides a valuable setting for addressing this question because of the natural variation in the security procedures passengers are subjected to.

Should our findings be interpreted as “bad news” for policy makers, police, or other power holders striving to improve their public image through procedural justice? We think not. In order to be viewed favorably, police should implement the principles of procedural justice, because they have a particularly strong moderating effect on feelings such as intimidation, humiliation, and perceptions of insensitive treatment, and they lessen the negative effects of intrusive or unpleasant practices to some extent. As recently argued, they are also the morally right way for police to behave in democratic societies, where policing is done by consent (Lum & Nagin forthcoming; Tankebe 2009a). At the same time, our findings suggest that policy makers should not expect the principles of procedural justice to be the “ultimate cure” for any intrusive, humiliating, threatening, or otherwise unpleasant policing practice. The nature of the practice also matters, and for some tactics—procedural justice will simply not be enough to overcome their negative effect.

What are these tactics? In the specific context of security checks at Ben-Gurion Airport, they were opening and searching the passenger's suitcase and questioning her away from the other passengers. Unlike the other two practices considered in our analysis (being asked a large number of question and being questioned by relatively many security officers), these procedures do not only constitute “more screening,” but are qualitatively different from the “ordinary” screening process. A passenger may be asked a large number of questions by several security officers while still waiting in the main queue with all other passengers. Both to the passenger and to bystanders, this somewhat more intensive questioning may not appear substantially different from the “standard” procedure. At the same time, having one's suitcase opened and searched in public, and being asked to step away from the main queue and accompany security officers to a different location, deviate substantively from the security routine most passengers undergo. These two security measures also appear to elicit more uncertainty and intimidation (“where are they taking me? why? what will happen next? will I make it to my flight?”) and a sense of embarrassment as a result of being singled out in public and/or having one's intimate belongings hand-searched and displayed. These unique characteristics may explain the partial resilience of these screening measures to procedural justice.

But beyond our specific context, what can our findings tell us about the characteristics of policing tactics that elicit negative emotions even if the officers do everything “right”? At this stage, we can only speculate. One possibility is that our findings simply reflect partial measurement of procedural justice. We have already noted that we did not measure the participation component of procedural justice because we concluded that it is not applicable to the context of airport security. However, it could be that it takes a particular form that is yet to be identified—an important aim for future research. Moreover, as reviewed earlier, scholars have identified factors other than the four known behavioral elements of procedural justice that may affect overall fairness (Lind & Tyler 1988; Lind et al. 1997; Tyler 1988). Nevertheless, our findings demonstrate that procedural justice—at least as commonly defined and measured to date—does not fully account for overall evaluations of the process.

Another possibility is that some high-policing tactics entail element/s exogenous to procedural justice that affects one's overall evaluation of the treatment. We suggest two possibilities as to what they may be. The first is a “hassle factor” (as coined by Sindhav et al. 2006: 324). It may be that some policing practices are annoying, disturbing, or irritating beyond a certain reasonable threshold, and are thus more “resilient” to procedural justice. Returning to the framework of “high” versus “low policing,” we have proposed earlier that policing practices carried out for the purpose of serving a high-policing agenda may be perceived as more intrusive and intimidating, and may thus produce a greater sense of disturbance, annoyance, and overall “hassle.” However, even low policing practices that exceed a certain reasonable threshold may accumulate and produce a sense of hassle. This proposition is supported by the findings of Tyler et al. (2014a), who identified that the more police stops citizens were subjected to, the less they tended to perceive the present stop, and police processes more generally, as fair. We suspect that feelings of annoyance and disturbance gradually accumulated with each additional stops, eventually undermining perceptions of fairness.

The second possibility concerns evaluations of distributive justice, that is, “the fairness and equity of the police delivery of services to persons across social and demographic groups” (Tyler & Fagan 2008: 239). It may be that when subjected to high (or low) policing practices, some individuals feel that the mere fact that they were selected to have their personal belongings searched, for example, is unfair, regardless of how they were treated. This proposition was at the core of the appeal to the Israeli Supreme Court mentioned earlier. It is also in line with the “expressive harm” hypothesis (Risse & Zeckhauser 2004), linked to the setting of airport security profiling by Hasisi et al. (2012). Notably, distributive justice received some attention in this context (Tyler et al. 2014a), and in our survey, the item The security officers treated me like every other passenger, which we have used to measure neutrality, could be viewed as partially tapping distributive justice. But given our findings, we suggest that it is addressed more directly in future studies predicting citizens’ assessments of interactions with police.

Before concluding, the uniqueness of our study setting and other potential limitations and suggestions for future research should be discussed. First, although the context of airport security has been studied as a policing one (Hasisi & Weisburd 2011; Lum et al. 2015), and we propose that it is a unique opportunity to study the outcomes of high-policing practices, it may be that it is somewhat different from street policing, both “high” and “low.” For example, it could be argued that unlike many types of police-citizen interactions, being subjected to airport security is ultimately a matter of choice. Passengers may decide at any point that they do not wish to undergo harassing or invasive security checks and simply leave the airport, or, recognizing that at least some level of screening is inevitable, choose a different form of transportation from the outset. We argue, however, that in modern reality air travel is often the only feasible way to get to one's destination, thus making interactions with airport security almost as inevitable as interactions with police. Nevertheless, we encourage future research to examine the interplay between what the police do and how they do it in the context of everyday police patrol.

Second, although Israel has been recognized as an important and fruitful setting for the study of police (Jonathan-Zamir et al. 2014; Perry & Jonathan-Zamir 2014; Weisburd et al. 2009), airport security is carried out somewhat differently in Israel than in the United States, for example (Lum et al. 2015). Further, public expectations regarding procedural justice may affect the findings in analyses such as ours. For example, low expectations for receiving procedurally-just treatment from authorities may mean that procedural justice plays a smaller role in forming overall assessments of the process. In turn, expectations about “acceptable” levels of police-provided procedural justice depend on cultural norms, and may thus vary across societies and ethnicities (Brockner et al. 2001; Hinds & Murphy 2007; Tankebe 2009b). In this context, we should note that although we control for ethnicity in our analysis, it was not our main focus, and thus we did not examine interaction effects. It may be that separate models for Jewish, Arab, and foreign passengers would have led to somewhat different findings. Thus, in addition to the how and the what, the who may also matter, and should be explored in future research.

We should also note that our findings only reflect the views of passengers who agreed to participate in our survey. Nevertheless, as noted earlier, our procedure and findings parallel earlier studies on airport security in Israel. Moreover, our response rate is similar to that reported in prior studies, and we could not detect systematic differences between passengers who agreed to participate and those who declined. Thus, we maintain that our study provides an important example demonstrating that what the police do, not just how they do it, should be considered in discussions of fair policing. Our findings particularly illuminate the important effects of high-policing tactics, some of which may produce a sense of hostile treatment even if the principles of procedural justice are followed.

Conclusions

We began this article by noting recent attention to the proposition of policing through procedural justice, but have also raised questions about its outcomes in the complex world of police-citizen interactions. We argue that various factors may work alongside procedural justice, contributing to or weakening its expected positive outcomes, and in this article have focused on the role of “high policing” tactics. Our analysis of the effects of “extra” airport security measures on negative emotions, such as humiliation and intimidation, supports this proposition. It is not disputed that procedural justice is critical to feeling positively about the process. It was the strongest “cure” in our model to negative emotions that may arise as a result of the security checks. At the same time, two of the four “extra” security measures we examined affected passengers’ overall evaluations of hostility after procedural justice was accounted for. Thus, procedural justice, at least as commonly defined and operationalized today, does not fully account for individuals’ evaluation of the process, and other factors, such as the nature of the policing practice being used, should also be considered. In terms of policy, the police and other power holders can expect important benefits from procedural justice. At the same time, it is not the case that they can engage in any intrusive or unpleasant practice, and be perceived positively, just as long as they follow the principles of procedural justice. Some policing tactics, which we suspect are particularly harassing or suggest inequity in their distribution, will likely produce negative emotions even if they are executed in a procedurally-just fashion.