Abstract
Much research has discussed the “police personality.” Yet, it is still unclear what particular traits make up the police personality—or whether it exists at all. This can be partially attributed to the limited availability of data collected within individual police departments. Using a nationally representative sample of adults (Add Health), the current study examines whether the Big Five personality traits, temperamental characteristics, and empirically informed covariates are related to being a police officer. Results indicate that, compared to the general population, individuals in law enforcement score significantly lower on openness to experience. Police officers are also more likely than non-officers to have experienced divorce, served in the military, lean politically conservative, and be male. These findings provide partial support for the notion that there are distinct factors that define the police personality.
In the law enforcement profession, certain traits are assumed to coalesce in individuals forming a distinct “police personality.” This personality has been defined as a value orientation specific to law enforcement officers. It consists of interrelated personality characteristics including bravery, loyalty, authoritarianism, cynicism, suspiciousness, physical courage, self-assertiveness, aggression, distrustfulness, secretiveness, conservatism, dogmatism, and alienation (Bennett & Greenstein, 1975; Lefkowitz, 1975; Twersky-Glasner, 2005; Van Maanen, 1973). The police personality has also been described as a high need for exhibitionism, the need to do new and different things, argumentativeness, assertiveness, and dominance (Mahanta & Kathpalia, 1984). Skolnick (1966) noted that the police working personality is formed by the officer’s role as an authority figure—emerging partly from the pressures associated with being in law enforcement—and consists of cynicism, suspiciousness, external isolation, internal solidarity, and conservatism. The common theme describing the police personality is the idea that law enforcement officers are a distinct group that can be identified based on a certain set of personality characteristics. As Van Maanen (1973) noted, “police are almost always depicted as a homogeneous occupational grouping somehow quite different from most other men” (p. 408).
Before discussing the police personality, it is important to have a general concept of personality and personality traits. The term personality refers to the combination of characteristics of thoughts, feelings, and behaviors forming an individual’s distinct character and is thought to remain relatively stable throughout the life course. Monte (1999), for instance, defines a personality trait as a biological, psychological, and sociological mixture that predisposes an individual towards a specific kind of action under certain circumstances. Some argue that these traits are primarily genetic and set in place early in life (perhaps before birth) and remain stable over the life course, others suggest that these traits are more malleable and are shaped predominantly by socialization, and some scholars claim personality traits are formed by some combination of genetics and environment—and the interaction and/or interplay between the two (Caspi, 1998; DeYoung, 2010; Pomerantz & Thompson, 2008). Pertaining to the police personality, scholars have similarly proposed different “models” or pathways to the formation of the police personality.
The Police Personality
Some of the first studies to discuss the police personality focused on the “working police personality” and attempted to reveal the origins of the police personality (Bennett, 1984; Bennett & Greenstein, 1975; Skolnick, 1966). Recall that Skolnick (1966) argued that the police officer’s working personality consist of cynicism, suspiciousness, external isolation, internal solidarity, and conservatism—and was formed by their role as an authority and the pressures associated with the job. Bennett and Greenstein (1975), likewise, discuss the police personality as a value orientation specific to law enforcement officers that make them unique from those they police and found that police science majors value systems were significantly different from experienced officers, but not other college majors. This led them to conclude that something occurs during the recruitment or training process that makes the value system of officers markedly different from the normal population. Bennett (1984) also found evidence that police academy training seemed to change recruits’ values to be more in line with that of experienced officers, but this effect was attenuated after individuals spent time on the street.
To clarify, there are four general views on the origins of the police personality. First, the predispositional model argues the police personality consists of traits the individual possessed prior to entering law enforcement. Second, the selection model suggests police applicants and candidates have the same common personality traits as the general population, but because of screening procedures only certain personality types are admitted into the academy and then the police force (Check & Klein, 1977). Third, the socialization model suggests the police personality is an artifact of the formal and informal processes of police work (Bennett & Greenstein, 1975). In other words, the police personality is a function of the police subculture and learned in the academy and on the job. At first, this temperament or disposition may only appear while on the job but soon bleeds over into their everyday life and becomes part of their personality. Fourth, the null model proposes that there may be no police personality and those in law enforcement have the same general personality traits and characteristics as the larger population.
While the first three models (predispositional, socialization, and selection) suggest a different mechanism at work in the etiology of the police personality, all assume that individuals in the law enforcement profession have personalities that somehow differ from that of the non-officers. It should also be noted that these models are not mutually exclusive, and that more than one may aid in explaining the underpinnings of the police personality. For example, individuals with certain traits may have a greater likelihood of joining a police force (predispositional hypothesis); during training and hiring, individuals with specific personality traits and predispositions are weeded-out (the selection hypothesis); and while on the job an officer is socialized to act, or think, in a certain way (the socialization hypothesis). It is possible—and likely—that all of these processes take part in shaping an officer’s personality, temperament, and/or behavior. If so, police officers may differ in personality and temperament from the general population. This is an empirical question that warrants further investigation. Indeed, although a number of studies have examined the police personality (Balch, 1972; Carpenter & Raza, 1987; Check & Klein, 1977; Garbarino et al., 2012; Twersky-Glasner, 2005), scholars have yet to explore whether certain personality characteristics are unique to law enforcement officers compared to the general population. Importantly, all the models have found support in the literature with studies supporting no particular theory, beckoning for further research in this area (see, e.g., Bennett & Greenstein, 1975; Hogan & Kurtines, 1975; Rokeach et al., 1971; Van Maanen, 1973).
It is evident that although findings are mixed in terms of which model better explains the origins of the police personality, results tend to support the notion that police officers have a distinct set of personality characteristics that distinguish them from the general population (Carpenter & Raza, 1987; Harper et al., 1999; Matarazzo et al., 1964). Yet, questions still remain as to what exactly the police personality is, what it consists of, if it is generalizable, and how it should be measured. Given that prior findings pertaining to the formation of the police personality have varied from study-to-study and sample-to-sample, the evidence for a singular explanation for the police personality is somewhat weak. It may be that there is a complex process occurring in which certain traits are found to cluster in law enforcement officers. In order to better understand the police personality and how it is formed, scholars must first identify what characteristics—if any—distinguish officers’ personalities from that of the larger population. The most well-known, recognizable, and researched personality characteristics are the “Big Five.”
The “Big Five” and the Police Personality
One of the most commonly used personality measures is the “Big Five” or the “Five Factor Model” (Schmitt et al., 2007). Indeed, it has been utilized in many studies across a variety of disciplines to examine myriad research questions (Gerber et al., 2011; Hurtz & Donovan, 2000; Loehlin et al., 1998). Psychologists, economists, political scientists, and behavioral geneticists, for instance, examined the heritability of the Big Five factors (Loehlin et al., 1998), how the personality traits are related to job performance (Hurtz & Donovan, 2000), and if the Big Five predict political affiliation (Gerber et al., 2011). Moreover, the Big Five has been shown to be both reliable and valid using various instruments and cross-culturally (Schmitt et al., 2007). The model captures five broad domains of personality: extraversion, neuroticism, agreeableness, conscientiousness, and openness to experience. Briefly, extraversion is comparable with pursuing excitement and novelty, and taps into qualities like sociability, outgoingness, talkativeness, and assertiveness (Roccas et al., 2002). Neuroticism includes characteristics related to anxiety and emotional liability; and neurotic individuals are typified as having anxiety, insecurity, depression, and anger (Fenster & Locke, 1973). Agreeableness is attuned with the motivational goals of conformity and traditional values. Individuals scoring high on agreeableness are characterized as good-natured, gentle, modest, compliant, and cooperative (Black, 2000). Conscientiousness includes characteristics such as responsibility, carefulness, thoroughness, scrupulousness, and having self-control. Openness to experience is analogous with universalism and the motivational goals of self-direction and refers to intelligence, imaginativeness, sensitiveness, and open-mindedness (Barrick & Mount, 1991). Judge et al. (2002) note there is a consensus the Big Five can be used to illustrate the most salient aspects of an individual’s personality and that the model has been found to be generalizable across measures, cultures, and sources of ratings.
Although not examining the full Big Five personality traits, some studies have looked at one or multiple facets of the Big Five and the police personality. Gudjonsson and Adlam (1983), for instance, investigated the personality differences between four different British law enforcement groups. Compared to normative groups of a similar age, recruits scored higher on extraversion, impulsiveness, and venturesomeness (i.e., openness to experience). Garbarino and colleagues (2012) examined the personality profiles of Italian special force police officers. Compared to the general population and career soldiers, officers were found to be more emotionally stable and moderately more agreeable, extraverted, open to experience, and conscientious. Collectively, these articles both point to the notion that police recruits and officers may be more extraverted, agreeable, open to experience, and conscientious.
Additional studies, however, did not find evidence of a distinct “police personality” (Fenster & Locke, 1973; Mahanta & Kathpalia, 1984). For example, neuroticism was not found to be a major personality trait of New York City officers (Fenster & Locke, 1973). The authors state that the results are limited given their design and that future studies using nationally representative populations are needed in order to generalize these results. Mahanta and Kathpalia (1984) conducted a study of the personality characteristics of police, judicial, and correctional officers using measures of psychoticism, neuroticism, extraversion, and dissimulation. Results revealed there were no mean differences between the three groups of law enforcement, as well as those in law enforcement and a control group of parents. Thus, there was no evidence of a police personality with police scoring generally the same as those from the general population in measures of neuroticism and extraversion.
Few studies have used the Big Five personality traits in their entirety when examining the police personality (Garbarino et al., 2012; Salters-Pedneault et al., 2010). Salters-Pedneault et al. (2010), for instance, compared the personality and psychological profiles of police recruits, firefighter recruits, and a normative sample. Police and firefighter recruits were found to have similar mean scores on neuroticism, extraversion, openness to experience, and agreeableness, with police recruits scoring higher on conscientiousness. When compared to the general population, police and firefighter candidates were more extraverted. Though the sample was limited and these findings are difficult to generalize because it included firefighters in the sample and a single Italian special forces unit, the study suggests that police and firefighters recruits’ personality characteristics were not systematically different from the larger population.
Scholars have also explored the value of using the Big Five in predicting police performance and its utility in officer selection. Cortina and colleagues (1992), for instance, found that while the Big Five was a valid measure of police performance, it did not add incremental validity over the Civil Service exam. Extraversion, neuroticism, and conscientiousness were, however, significant predictors of turnover. Similarly, Black (2000) found that higher-order traits of neuroticism and extraversion were predictive of positive job performance for law enforcement officers with conscientiousness having the strongest relationship with performance, while openness to experience and agreeableness had no effect. Barrick and Mount (1991) performed a meta-analysis examining the ability of the Big Five to predict job performance and found that police performance was related to all personality factors except openness to experience—however, those scoring high on openness to experience were most willing to engage in learning experiences and benefit from training programs. On the other hand, Sanders (2008) examined whether the Big five personality traits predicted police officer performance and concluded that the Big Five personality characteristics were not predictive of good or poor officer performance. Sanders notes that this null finding may be partly attributed to the difficulty in measuring officer performance and lack of variation in supervisor performance ratings (i.e., most officers were rated as above average). Though these articles were not examining the police personality directly, these studies demonstrate that the Big Five model can be useful in investigating different aspects of policing.
Overall, the studies demonstrate that evidence for the “police personality” is mixed with some studies finding no differences in the Big Five traits and other studies revealing that—at least for some aspects of the Big Five—there were significant personality differences between law enforcement officers and other groups. Importantly, previous studies have been limited in generalizability due to their small sample size or inclusion of recruits in the sample. However, these studies do suggest that the Big Five model can be useful in exploring the personality differences between law enforcement and those working in other occupations or the general population.
The Current Study
Law enforcement officers have been described as having a distinct set of personality characteristics that distinguish them from the general population. For example, officers have been described as aggressive, dominate, assertive, suspicious, and conservative (Mahanta & Kathpalia, 1984; Skolnick, 1966). Research on the police personality, however, is varied. Research on the origins of the police personality, for example, finds evidence for the predispositional model (Rhead et al., 1968; Rokeach et al., 1971), the socialization model (Van Maanen, 1973), a selection effect (Hogan & Kurtines, 1975), and the null hypothesis (Balch, 1972; Bennett & Greenstein, 1975). The evidence points to the idea that it is probably a combination of the models that aid in forming, defining, and reinforcing the police personality. Given that findings pertaining to the formation of the police personality have varied from study-to-study and sample-to-sample, the evidence for a singular explanation for the police personality is somewhat weak. It may be that there is a complex process taking place in which certain traits are found to cluster in law enforcement officers. Regardless of how it is formed or where it comes from, all these models—except for the null hypothesis—assume that police officers are somehow different from the general population.
As previously stated, evidence is mixed as to whether police officers have a distinct set of personality characteristics different from that of the general population. This is due, in part, to the lack of control variables and generalizability issues in prior studies. Balch (1972), for instance, noted that studies have found Portland police are similar to college students, Denver police were less authoritarian than the general public, and recruits in New York were more authoritarian than the larger population. With this in mind, it is necessary for research to use data from nationally representative samples. Considering that the aforementioned criticisms came in the 1970s, and few studies have addressed these claims, it is necessary for researchers to use national data and a more methodologically rigorous analysis controlling for multiple covariates. Many of the previous studies were conducted 40 years ago and were unable to control for gender because the samples did not include a sufficient number of women, or none at all. Provided that more women are now serving in law enforcement, it is advantageous to control for the effect for gender, as it may show some personality differences to be spurious (e.g., it may be that males are more authoritarian).
To be sure, the Big Five personality traits were chosen because many scholars have utilized the measures when examining the police personality (Garbarino et al., 2012; Salters-Pedneault et al., 2010) and they have been shown to be reliable and valid indicators of relatively stable personality traits (Baldasaro et al., 2013; Judge et al., 2002; Schmitt et al., 2007). Covariates such as stress (Beutler et al., 1988), depression (Lorr & Strack, 1994), anxiousness (Balch, 1972; Evans et al., 1992), anger/hostility (Evans et al., 1992; Garbarino et al., 2012), divorce (Kappeler et al., 2000; McCoy & Aamodt, 2010), college education (Fenster & Locke, 1973), prior military service (Farkas & Manning, 1997; Pasillas et al., 2006), and political ideology (Balch, 1972, 1977; Skolnick, 1966, 1977) were chosen because they were found to be related to law enforcement and the police personality.
For example, research has shown that after only two years on the job, candidates experienced an increase in somatic symptoms, stress, anxiety, and alcohol vulnerability (Beutler et al., 1988). Lorr and Strack (1994) utilized multiple measures of depression when examining the personality profiles of law enforcement candidates. Research has been inconsistent in terms of anger/hostility, with some finding that police personnel had higher scores of anger (Evans et al., 1992) and others concluding that officers had lower scores on anger-hostility compared to the general population (Garbarino et al., 2012). The relationship between divorce and being in law enforcement is also mixed with some claiming that police officers have higher divorce rates than the general population (Kappeler et al., 2000), yet others have found that the divorce rate for law enforcement personnel is lower than that of the general population (McCoy & Aamodt, 2010). Fenster and Locke (1973) found personality differences between college-educated and noncollege-educated officers and civilians. Since it has been argued that police work parallels that of the military (Farkas & Manning, 1997) and because others have included it as a covariate when examining the police personality (Pasillas et al., 2006), military service will be included in the analysis. Finally, officers have been described as politically conservative (Balch, 1972, 1977; Coleman & Gorman, 1982; Skolnick, 1966, 1977) and so political ideology will be included as a covariate. The current study will add to the existing literature by using a nationally representative sample of adults to examine whether personality and temperamental differences exist between law enforcement and the general population while controlling for a host of theoretically informed covariates, leading to a more nuanced examination of the police personality. As such, this study’s aim is to add to the police personality literature by exploring whether individuals in that profession have unique personality characteristics.
Methods
Data
The current study uses data drawn from Wave 4 of the National Longitudinal Study of Adolescent to Adult Health (Add Health; Harris, 2009; Udry, 2003). The Add Health is a five-wave study of a nationally representative sample of American adolescents in grades 7 through 12 during the 1994–1995 school year. The first wave of data was collected through two different surveys: an in-school survey and an in-home survey. Initial data collection began at the school-level, originally starting with a sample frame of 26,666 schools. Prior to sampling, the schools were stratified by size, type, urbanization, region, and percent white. A sample of 132 schools were selected and every student attending those schools were administered a 45-minute self-report questionnaire about their background, peers, school life, school work and activities, health-related behaviors and their parent’s background. This resulted in a sample of over 90,000 adolescents for the in-school questionnaire. A subsample of adolescents was then selected to take part in the wave 1 in-home component of the study. Additional information was obtained on the adolescent’s behaviors, tobacco use, alcohol use, drug use, social relationships, and personality traits. During the first wave, 17,670 primary caregivers (generally the respondent’s mother) were also interviewed.
The second wave of data was collected roughly one year later and drew on the same sample of adolescents who were interviewed at wave 1, with some exceptions (N = 14,738). In 2001 through 2002 a third wave of data was collected when most of the participants were young adults. Since respondents were now in adulthood, the surveys were updated to include more age-appropriate questions. In total, 15,197 participants were re-interviewed during the third wave. The fourth wave was obtained during 2007–2008, from approximately 80% of the original wave 1 respondents resulting in a sample size of 15,701 respondents. At this time, most of the respondents were between 24 and 32 years old. Questions tapped items such as personality factors, temperament, child-rearing history, educational achievement, work history, and marriage history. Due to the oversampling during Wave 1, some individuals had a higher probability of being included in the sample than others (i.e., not randomly selected). For this reason, survey weights will be used in all analyses.
Measures
Dependent Variable
Law Enforcement
The dependent variable was taken from a single question asked during the wave 4 interview. Specifically, the question asked, “Now I’d like to record a description of your (current/most recent) job. When you see the list of categories, please tell me which best describes what you (do/did) at your (current/most recent) job.” The occupations were coded following the Standard Occupational Classification (SOC) system to classify respondents’ current, or most recent, paying job that was at least 10 hours per week, excluding military service. During the interview, participants were guided though the four SOC levels: major group, minor group, broad occupation, and detailed occupation until the appropriate job title was identified.
The variable was coded in order to capture those who were police officers. Thus, only those respondents who reported being: police officers, sheriff’s patrol officers, transit and railroad police, detectives, managers of police and detectives, law enforcement workers, criminal investigators, and first-line supervisors. The variable was dichotomously coded with 0 = non-law enforcement and 1 = law enforcement.
Key Independent Variables
The Big Five personality traits were measured via the Mini-IPIP (for a detailed description see Baldasaro et al., 2013). The Mini-IPIP is a short form inventory based on the 50-item International Personality Item Pool five-factor model. The Mini-IPIP has been shown to have exceptional content coverage, high test-retest correlations, and criterion validity (Donnellan et al., 2006). Baldasaro and colleagues (2013) found support for the Mini-IPIP five-factor model based on confirmatory factor analysis and exploratory factor analysis with exploratory factor analysis revealing that items loading on one primary factor and any cross-loading being small. Thus, the five-factor structure of the Mini-IPIP inventory used in the Add Health has received support within the literature.
Extraversion
The extraversion scale was measured via four items taken from the Mini-IPIP. For each item, respondents were asked how much they agreed with each statement: “I am the life of the party,” (reverse coded) “I don’t talk a lot,” “I talk to a lot of different people at parties,” (reverse coded) and “I keep in the background.” Responses were summed so that higher scores indicate higher extraversion (α = .69).
Neuroticism
The neuroticism scale was created by summing four items taken from the Mini-IPIP. For each item, respondents were asked how much they agreed with each statement: “I have frequent mood swings,” (reverse coded) “I am relaxed most of the time,” “I get upset easily,” (reverse coded) and “I seldom feel blue.” Responses were summed so that higher scores indicate higher neuroticism (α = .58).
Agreeableness
The agreeableness scale was generated by summing four items taken from the Mini-IPIP. For each item, respondents were asked how much they agreed with each statement: “I sympathize with others’ feelings,” (reverse coded) “I’m not interested in other people’s problems,” “I feel others’ emotions,” (reverse coded) and “I am not really interested in others.” Responses were summed so that higher scores indicate higher agreeableness (α = .69).
Conscientiousness
The conscientiousness scale was measured via four items taken from the Mini-IPIP. For each item, respondents were asked how much they agreed with each statement: “I get chores done right away,” (reverse coded) “I often forget to put things back in their proper place,” “I like order,” (reverse coded) and “I make a mess of things.” Responses were summed so that higher scores indicate higher conscientiousness (α = .65).
Openness to Experience
The openness to experience scale was generated by summing four items taken from the Mini-IPIP. For each item, respondents were asked how much they agreed with each statement: “I have a vivid imagination,” (reverse coded) “I am not interested in abstract ideas,” “I have difficulty understanding abstract ideas,” and “I do not have a good imagination.” Responses were summed so that higher scores indicate higher openness to experience (α = .65).
Covariates
Stress
Perceived stress was measured using a validated 4-item shortened version of the Cohen’s Perceived Stress Scale (Cohen et al., 1983; Cohen & Williamson, 1988; Dowd et al., 2014; Karam et al., 2012; Warttig et al., 2013). All items were coded on a five-point scale so that higher scores reflected more perceived stress. The scale was generated by summing the four items tapping a respondent’s perceived stress over the last 30 days (α = .72).
Depression
Depression was measured via a five-item scale. Specifically, respondents were asked to think about the past seven days and indicate how often each of the following things was true during the past seven days: “you were bothered by things that usually don’t bother you; ” “you could not shake off the blues, even with help from your family and friends; ” “you felt you were just as good as other people; ” “you had trouble keeping your mind on what you were doing; ” “you felt depressed; ” “you felt sad.” Each item was coded such that 0 = never or rarely, 1 = sometimes, 2 = a lot of the time, 3 = most of the time. To create the scale, respondent’s answers were summed together with higher scores representing more depressive symptoms (α = .80).
Anxiousness
Four items were used to generate the anxiousness scale. Specifically, respondents were asked how much they agreed with the following statements: “I worry about things,” “I am not easily bothered by things,” (reverse coded) “I get stressed out easily,” and “I do not worry about things that have already happened” (reverse coded). To generate the scale, all items were summed together with higher scores reflecting higher anxiousness (α = .70).
Anger/Hostility
Respondents were asked four items taken from the anger facet of the NEO PI-R (Costa & McCrae, 1992). Respondents were asked how much they agreed with the following statements: “I get angry easily,” “I rarely get irritated,” (reverse coded) “I lose my temper,” and “I keep my cool” (reverse coded). To generate the scale, all items were summed together with higher scores indicating higher anger/hostility (α = .77).
Divorced
Divorce was measured by one item asking respondents how their marriage to their previous partner ended. The item was coded so that 0 = non-divorce (i.e., they were still married, they had never been married, their spouse died, or they did not know), 1 = divorce or annulment.
College Degree
Respondents were asked to indicate the highest level of education that they had achieved to date. The item was coded so that 0 = 8 th grade or less, some high school, high school graduate, some vocational/technical training, completed vocational/technical training, some college, 1 = completed college (bachelor’s degree) or higher.
Military Service
Respondents were asked to indicate whether they had ever been in the military. The variable was dichotomously coded such that 0 = no, 1 = yes.
Politically Conservative
Respondents were asked to identify their political affiliation. Specifically, respondents were asked, “In terms of politics, do you consider yourself very conservative, conservative, middle-of-the-road, liberal, or very liberal?” The item was coded so that 1 = very liberal, 2 = liberal, 3 = middle-of-the-road, 4 = conservative, 5 = very conservative.
Control Variables
Age
A count variable indexed the respondent’s age. Responses ranged from 24 to 34.
Male
A dichotomous variable indexed the participant’s sex (0 = female, 1 = male).
White
A binary variable indexed the respondent’s race (0 = non-white, 1 = white).
Black
A dummy variable indicating the participant’s race (0 = non-black, 1 = black).
Hispanic
Descriptive Statistics (n = 13,145).
Note. SD = Standard Deviation.
Analytic Plan
The analysis for the current study will proceeded in a series of four steps. The first step is to estimate a logistic regression model that examines the effect of the Big Five personality traits (i.e., extraversion, neuroticism, agreeableness, conscientiousness, openness to experience) on being law enforcement. The second step involves estimating another logistic regression model. This model, however, includes temperamental variables (i.e., stress, depression, anxiousness, anger/hostility) along with the Big Five personality traits. The third step is to estimate a logistic regression model that examines the effect of the Big Five personality traits, temperamental variables, and empirically informed covariates (i.e., divorce, college education, prior military service, political views) on being a police officer. The fourth step is to re-estimate the model including the Big Five personality traits, temperamental variables, covariates, and control variables (i.e., age, sex, and race). The covariates will be entered into the regression equations in a stepwise fashion, allowing for the observation of changes in the relationship between the Big Five personality traits and being in law enforcement as a function of the covariance shared with the different covariates. It is also important to note that—to aid in interpretation and comparisons between models—the sample size was held constant for each model such that only those respondents who had complete data on every variable in the final model (i.e., model 4, which includes all variables) were included in the previous models. In other words, listwise deletion will be utilized in the current study. Despite dropping these cases, a robust sample size remained. This resulted in a sample size of over 13,000 individuals, which provided sufficient statistical power (Barnes et al., 2020). Importantly, there were no systematic missing responses (i.e., the missing values were randomly distributed).
Results
Predictors of Being in Law Enforcement (n = 13,145).
*p < .05; Note: OR = Odds Ratio; SE = Linearized Standard Error.
Model 1 analyzes the effect of the Big Five personality traits on being in law enforcement before any additional variables are added to the model. Two personality variables are statistically significant. Specifically, the coefficient for neuroticism is negative and significant with results indicating that a one unit increase in neuroticism is related to a 12% decrease in the odds of being in law enforcement, adjusting for the other personality covariates (IRR = .88, p < .05). Additionally, the coefficient for agreeableness was negative and significant, with the odds ratio revealing that a one unit increase in agreeableness decreases the odds of being in law enforcement by 9% (IRR = .91, p < .05).
Model 2 presents the results of the logistic regression of law enforcement on the Big Five personality traits and additional temperamental variables. As can be seen, once these variables are added to the model, the relationship between the personality characteristics and being in law enforcement is attenuated and rendered insignificant. Moreover, none of the covariates in the model are significant predictors of being in law enforcement.
Model 3 reveals the results from the third logistic regression model analyzing the effect of the Big Five personality traits, temperamental characteristics, and theoretically informed covariates on being a police officer. Again, none of the Big Five traits or temperamental characteristics are significant. Prior military service is, however, significant with results indicating having served in the military was related to a 244% increase in the likelihood of being a police officer (IRR = 3.44, p < .05). Results also reveal that for every one unit increase in being politically conservative, the odds of being in law enforcement increases by 38% (IRR = 1.38, p < .05).
Model 4 presents the full model examining the effects of the Big Five personality traits, temperamental factors, covariates, and control variables on being in law enforcement. As can be seen, once the full model is analyzed, only one of the Big Five personality traits is significant. Specifically, for every one unit increase in openness to experience, the odds of being a police officer decrease by 7% (IRR = .93, p < .05). Being divorced was related to a 54% increase in the likelihood of being a police officer. Prior military service increases the odds of being in law enforcement by 151% (IRR = 2.51, p < .05). Conservatism remains significant, with a one unit increase in being politically conservative increasing the odds of being a police officer by 32% (IRR = 1.32, p < .05). Male is the only significant control variable and is associated with a 542% increase in the odds of being in law enforcement (IRR = 6.42, p < .05).
Discussion
Police have been described as a homogeneous group with distinct personalities consisting of characteristics such as authoritarianism, extraversion, conservatism, exhibitionism, suspiciousness, and aggressiveness (Lefkowitz, 1975; Van Maanen, 1973). Prior research has found that, compared to the general population, police recruits and applicants were more extraverted, less depressed, more impulsive, and more open to experience (Carpenter & Raza, 1987; Gudjonsson & Adlam, 1983; Salters-Pedneault et al., 2010). Additional studies have found that police scored higher in neuroticism and extraversion than the general population (Mahanta & Kathpalia, 1984). To date, however, no study has used a nationally representative sample of individuals to explore the personality traits and temperamental characteristics (i.e., stress, depression, anger) associated with individuals in law enforcement. Moreover, no prior research has included demographic control variables (i.e., age, race, and sex) as well as additional theoretically and empirically informed covariates (i.e., divorce, education, military service, political ideology), which may show the police personality to be spurious. With this in mind, the current study set out to examine the personality and temperamental predictors of being a police officer.
Results from the current study reveal that officers scored significantly lower on openness to experience (recall that openness to experience refers to universalism, intelligence, imaginativeness, sensitiveness, and open-mindedness). Respondents who reported having been divorced, had prior military service, who leaned more conservative, and males were also significantly more likely to be in law enforcement—with males having a 542% increase in the odds of being in law enforcement. This pattern of findings may suggest that certain factors coalesce to form a distinct police personality—such as a domineering or authoritative personality type (see, e.g., Skolnick, 1966; Twersky-Glasner, 2005; Van Maanen, 1973). Given that prior literature has found that stress, depression, anger, and anxiety were associated with the police personality, they were included in the model along with the Big Five personality traits (Pasillas et al., 2006). No temperamental variables were related to being in law enforcement in any model, which was unexpected but in line with prior findings (Gudjonsson & Adlam, 1983; Lawrence, 1984).
This pattern of findings may be due to the length of time officers have been on the job. Though personality traits have been shown to remain relatively stable throughout the life course, the “socialization model” suggests that once an individual is introduced into police subculture, their attitudes (and perhaps personality) change (Bennett & Greenstein, 1975). Since the mean age in the sample was 28, it is possible that once these individuals are acclimated to the subculture their attitudes and characteristics may change. Therefore, the lack of findings for covariates such as stress, depression, and anger might be due to officers not being on the job for an extended amount of time (e.g., they have less exposure to the police subculture). It may be that personality and temperamental differences would become more pronounced once officers have been on the job longer. Future studies should investigate whether officers with more experience on the job differ from newer recruits on personality characteristics.
While the current study filled an existing gap within the literature, certain points must be acknowledged. It is possible that an officer’s behavior is not representative of the individual but a characteristic of the situation. The current study assumes that the traits measured are relatively stable throughout the life course and are not situational. Additionaly, the current study was exploratory and utilized only one wave of data and, thus, could only show correlations. For instance, males were found to have much higher odds of being in law enforcement. However, it is unknown if this is simply due to many more males applying for the jobs or to the rigorous training and physical requirements that may make males more likely to stay in such a profession. Future studies should employ more rigorous quasi-experimental analyses—such as the one used by Garbarino et al. (2012)—to examine if the Big Five personality traits predict being in law enforcement. Future studies should also examine whether one’s social network predicts being in law enforcement. Since the Add Health only collected social network data during adolescence and the Big Five personality traits were not measured until adulthood, the current study was unable to examine this.
Additionally, the data were collected in 2009 and it is possible that the selection process for officers was different at that time than others. While this is possible, there are reasons to believe that this would not greatly impact the results. First, the data came from Add Health Wave 4 data collected in 2009—which is not too long ago. Hirschi, for example, argued that while the world has changed in ways we find important—such as technology—research typically finds that whatever changes have occurred have little relevance for theory or policy (Hirschi, as quoted in Laub, 2011, p. 321). Second, the police personality and police subculture appear to be relatively stable and resistant to change (Herbert, 1998; Malmin, 2012; Salters-Pedneault et al., 2010; Skolnick, 1977; Van Maanen, 1973). Third, results from the current study are in line with others finding that police personality consisted of factors such as low openness to experience, prior military service, and being more conservative (Balch, 1972; Coleman & Gorman, 1982; Gudjonsson & Adlam, 1983; Pasillas et al., 2006; Seklecki & Paynich, 2007; Skolnick, 1977).
It should also be noted that while findings from the current study are generalizable to the United States, they may not be applicable to police officers in other countries. There are reasons this too may not be a major concern. Mainly, the predicators included in this study have been found to be associated with the police personality in other countries—including Britain, Italy, Scotland, and South Africa—and shared characteristics of the police personality have been found cross-culturally (Garbarino et al., 2012; Gudjonsson & Adlam, 1983; Harper et al., 1999; Pienaar et al., 2007). Moreover, a study using data collected from 56 different nations revealed that the Big Five was highly replicable across all cultural regions around the world, the factor scales had high levels of internal reliability as well as cross-cultural and cross-instrument validity, and the Big Five has been found to be reliably measured across diverse human cultures (Schmitt et al., 2007). With this being said, the study was limited to the Big Five personality measures and, although these measures have been shown to be reliable and valid (Baldasaro et al., 2013; Schmitt et al., 2007), there may be other traits or characteristics not captured in the Mini-IPIP that were not controlled for that could comprise the “police personality.”
Notwithstanding the aforementioned limitations, the current study is the first to utilize a nationally representative sample to examine whether the Big Five personality traits, temperament characteristics, and empirically informed covariates are related to being a police officer. Whether it is because certain individuals are more attracted to law enforcement (i.e., the predispositional model), certain personality types are more likely to be admitted law enforcement (i.e., the selection model), or that officers are socialized into the police culture through formal and informal processes of police work (i.e., the socialization model), results from the current study provide partial evidence for a distinct police personality—with officers being less open to new experiences and more likely than non-officers to have prior military service, a more conservative political ideology, and be male. Findings from the current study also demonstrate the importance of including additional covariates beyond the Big Five, as the results changed once other variables related to the police personality were included in the model. Indeed, the model which included only the Big Five personality traits revealed that officers rated significantly lower on neuroticism and agreeableness. However, once empirically relevant variables were included, officers were only found to differ only on openness to experience.
Putting findings from the current study in a broader context, the results suggest a greater import should be placed on casting a wider net when hiring to ensure a diverse pool of candidates in terms personality and political ideology—with a strong emphasis on the recruitment of females and individuals from the civilian population. The finding that males were 542% more likely to be in law enforcement identifies a major point of concern and something that departments should prioritize when recruiting. Considering the finding that officers had a 54% increase in the odds divorce, departments may want to make marriage counseling available—and the finding that officers scored lower on openness to experience may indicate that officers may be reluctant to try services such as counseling or psychotherapy, which should be addressed as well. Additionally, finding that race was not a significant predictor of being an officer suggest that departments are doing well when it comes to reducing the discriminatory hiring practices based on race/ethnicity. Once hired, however, making the workplace culture (or the police subculture) less masculine and militaristic might aid in the retention of female officers and those with characteristics less in line with the traditional police working personality (Farkas & Manning, 1997; Seklecki & Paynich, 2007).
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research uses data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Special acknowledgment is due Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Information on how to obtain the Add Health data files is available on the Add Health website (
). No direct support was received from grant P01-HD31921 for this analysis.