Incivility: The Interaction Of Gender And Race In One-on-one Police-civilian Interactions

The proliferation of body-worn cameras (BWCs) offers a unique opportunity to study police-citizen interactions. While traditional research places emphasis on actions taken by police officers, there remains a large gap concerning the behaviors of citizens towards police officers. This is problematic given the substantial body of research documenting police work as high in emotional labor. To help fill this gap, the present study documents incivility, as experienced by police officers from citizens, using a sample of 156 police-citizen interactions. Key measures of incivility include interruptions, profanity, slurs, and deception. Given the interactional nature of the observations, measures of incivility are captured for the officer and citizen. In addition to documenting incivility, additional research questions include to what extent incivility varies by gender or race. Results of this research could help inform the development of training and wellness programs.

Introduction

There is a vast amount of research in agreement that police work is high in emotional labor (Lee & Ashforth, 1996; Leiter, 1993; Maslach & Jackson, 1984). Emotional labor translates to the management of one’s feelings, in order to display one’s self in accordance with situational demands (Hochschild, 1983). There is also an extensive amount of interdisciplinary research that concludes that emotional labor leads to workplace burnout, emotional and physiological distress (Leiter et al. 1998, Leiter & Maslach 2004, Maslach & Leiter 1997). Interdisciplinary research additionally highlights workplace incivility as a primary cause of emotional labor (Cortina et. Al. 2001) – suggesting that small, seemingly insignificant interpersonal injustices can have a greater effect than the layperson may think.

With this in mind, it is crucial to note that there are no empirical studies that examine incivility as experienced by police officers with direct observational data. This is problematic given the substantial body of research that says police work is high in emotional labor.

Additionally, there is a wide body of research that studies the interaction between policing and civilian characteristics like gender and race. While there are many studies pertaining to civilian demographic characteristics, there are not many conclusive findings. Most would agree with the finding that men are more likely to respond to emotions with criminal behavior (Broidy & Agnew, 1997). Despite decades of research, the area of race is still problematic. Most researchers agree that there are in fact disparities but cannot agree on where these disparities come from. Some say that minority citizens (primarily African Americans) have a history of being profiled and mistreated by police in America (Sue et al., 2008). While there is quite a sizeable number of studies that support police interactions differing in terms of a civilian’s race, there are some that say otherwise. These studies usually do in fact show racial disparities, but they also have confounding variables. One 1984 study that looked into the effect of race on arrests found that although there was an initial race effect, there were other contextual factors (in this study poverty) that led to arrest rates differing and concluded that race has no effect on arrest decisions (Smith et. Al., 1984).

Using data generated from unredacted BWC videos, the current study aims to fill the gap around policing and incivility while also documenting how civilian characteristics affect the occurrence of incivility. In addition to documenting the rate of occurrence, this study is also designed to analyze the effects of civilian gender and race on the totality of incivilities that may occur in one-on-one interactions.

Methods

This study uses unredacted BWC footage from 156 separate police-citizen interactions. These interactions took place over a three-year time period from 2017-2019. This sample of 156 separate interactions was strategically extracted from a much larger dataset, as the selected interactions take place with only one officer and one civilian present. 118 of the interactions involved white civilians and 38 involved minority civilians. 103 of the interactions involved male civilians and 53 involved female civilians. A sample of one on one interactions was chosen to avoid the complexity that other people might add and to lay a foundation for future research to build upon.

• Proliferation of BWC
• CSI lab info
• Measures

Dependent Variable

The outcome variable of this study is the rate of incivility as experienced by the officer. Due to the give-and-take nature of one on one police-citizen interactions, the rate of incivility as experienced by civilians is also measured. Past research states that incivility is low-intensity deviant behavior (Pearson et. Al. 2001) that may connote breaches of etiquette, professional misconduct, and moral decay (Carter, 1998; Gladwell, 1996; Johnson, 1988; Martin, 1996; Roberts, 1985). In this study, incivility can be measured by using several different variables coded for in the lab. Namely profanity, slurs, interruptions (as experienced by both civilians and officers), and deception (as experienced by officers). Accuracy is central to the lab’s analysis; there is a heavy focus on generating reliable measures and dimensions of incivility.

The use of profanity is coded for when a profane word is directed at either the civilian or officer. It is important to note that profanity is not coded for every time a profane word is used, only when directed profanity is used. A slur is coded for when the use of a degrading/insulting statement is directed at the civilian or officer. Once again it is important to note that a slur is not coded for unless it is directed at the civilian or officer. Slurs usually pertain to one’s identity (race/gender/sexuality) or character. An interruption is coded for when either the civilian or officer speaks over the other party, stopping or disturbing their speech. This dimension of incivility is easier to identify than the previous two. Any time one party speaks before waiting for the other to be done is coded as an interruption. Deception is the only one-way dimension of incivility. It is also the only dimension that is coded as occurring or not occurring. It is coded when the civilian is proven to have lied or given false information to the officer. It is important to note that deception is not coded for when a coder might think that a civilian is lying, or when an officer suggests that the civilian might be lying on BWC video. Deception is only coded for when false information is proven to have been given.

Independent Variable

The independent variables of this study are civilian characteristics and contextual factors. The civilian characteristics that were accounted for are gender and race. For this study, the race was broken down into two sub-categories ‘white’ and ‘other’ in order to account for the homogeneous population where the policing interactions took place and for statistical analysis. This study takes place in a small university town with less than 35,000 residents, most of which are white.

Contextual factors in this study include whether the interaction occurred during the day or night. This code is based on whether there is daylight or darkness during the interaction. Another contextual factor in this study is the level of verbal intensity levels as perceived by the coder. If ‘0’ was coded, then the situation was deemed not intense (a passerby would walk right past an interaction deemed not intense). If ‘1’ was coded, then the situation would be deemed intense (an intense interaction would cause a passerby to take notice). These judgments were made by trained undergraduate research assistants.

While demographic characteristics of citizens are recorded, data on the officer’s gender and race are not readily available to the CSI lab. Most officers in the present police agency are male and white.

Results

Officer Experienced Incivility

In this study, 35.89% of one-on-one stops that a police officer makes will involve at least one dimension of incivility, it is important to note that (21.4%) of interactions involved more than one form of incivility. Considering the totality of interactions (156) 23.7% of officer-directed incivilities in this study were suspect interruptions, 12.8% were profanities directed at officers, and 8.3% were suspect deceptions. There were 0 slurs directed at officers in this study.

When breaking down incivility as varied by gender, men in this study were more likely to interrupt an officer (56.75%) and to use directed profanity (75%) than females (25% and 43.25% respectively). Women in this study were more likely to deceive an officer (54%) than a man was (46%). Of the civilians that interrupted, men averaged 4.67 interruptions compared to women who averaged 3.87 interruptions per interaction. Of the civilians that cursed at police officers, men averaged 10.13 profanities per interaction compared to women who averaged 1.4 profane words. Deception, as previously mentioned was the only dimension of incivility that the CSI lab codes as ‘occurring/not occurring’ so there is no data on the average number of lies told by civilians.

When considering incivility as varied by race, the results in this study show that white civilians are more likely to be uncivil than minority citizens in each dimension of incivility. White civilians were more likely to interrupt (78.37%), use directed profanity (75%), and deceive an officer (76.92%) than minority civilians (21.63%, 25%, and 23.07% respectively). Of the civilians that interrupted, white civilians averaged 3.93 interruptions compared to minority civilians averaging 5.75 interruptions per interaction. Of the civilians that used profanity directed at officers, white civilians averaged 6.67 curse words compared to minority civilians who averaged 11.8. There is no data for the average number of deceptions made by civilians.

The findings pertaining to the interaction of race and gender on incivility, as it is concerned with officer-directed incivility, are intriguing, there is no specific group (white men, white women, minority men, minority women) that show a higher likelihood of incivility. Each dimension has a different group that leads to incivility. Minority females have the highest average of interruptions at 8.5, followed by minority males (4.83), white males (4.6), and finally white females (3.2). Minority males in this study used the most profanity averaging 18.67 profane words followed by white males (8), minority females (1.5), and white females (1.33). The average number of deceptions cannot be calculated, however, there was a total of 5 counts of deception for both white males and white females, 2 counts of deception for minority females, and only 1 count of deception for minority males.

Civilian Experienced Incivility – better heading?

In this study, 30.76% of one-on-one interactions had at least one dimension of incivility directed at the civilian. It is important to note that 10.14% of interactions had more than one dimension of civilian experienced incivility present. Considering the totality of interactions (156), 22.43% involved officers interrupting a civilian, and 11.54% of interactions involved an officer using directed profanity at a civilian. There were 0 slurs directed at civilians in this study.

While officers in this study were less likely overall to use incivility, they had a much higher likelihood of being uncivil to men rather than women. 71.43% of officer interruptions took place during interactions with men compared to 28.57% of interruptions in interactions with females. 88.88% of profanity directed at civilians was directed at males compared to 11.11% profanities that were directed at females. Of the interactions that involved officer interruptions, male civilians were interrupted an average of 2.96 times compared to an average of 4 times that officers interrupted females.

When considering officer incivility as varied by race, this study shows that officers are more likely to be uncivil to white civilians than minority civilians. 82.85% of officer interruptions involved a white civilian compared to 17.15% of interruptions that involved minority civilians. When officers used directed profanity, 83.33% of profanities were directed at white civilians and 16.67% were directed at minority citizens. Of the interactions that involved an officer interrupting a white civilian, officers averaged 3.24 interruptions compared to an average of 3.33 interruptions when interacting with a minority citizen. Of the interactions that involved profanity directed at a civilian, officers averaged 2.33 profane words per white civilian and 3 profane words per minority civilian.

The findings pertaining to the interaction of race and gender on incivility, as it is concerned with civilian experienced incivility, are very interesting in comparison to officer-directed incivilities. Once again there is no specific group in this study that shows a higher likelihood of experiencing incivility. Both dimensions of incivility have a different group that is more likely to experience uncivil acts from officers. When interruptions occurred, officers averaged to interrupt minority females the most (5), followed by white females (3.75), white males (3.04), and lastly minority males (3.04). When officers used profanity directed at civilians, they averaged using the most profane words when interacting with minority males (3), followed by white males (2.53), white females (1), and finally minority females at 0.

This is interesting in relation to the interaction of gender and race in officer-directed incivility. The results from this study show that the group that leads in incivility – in either form: officer directed or civilian directed incivilities – is also the group that leads in the opposite form. For instance, minority females in this study averaged to interrupt the most out of all of the groups, and they are also the group that is interrupted the most. Minority males in this study averaged using the most directed profanity, and they were also the group that received the highest average of directed profanities.

Chi-Squared Analysis

When the data regarding % chance of officer-directed incivility was run through chi-squared tests, there was no statistical significance for incivility in any dimension. The p-value for suspect interruptions was p= .2394. The p-value for suspect profanity was p=.3710. The p-value for suspect deception was p=.6115. When the data pertaining to the total count of civilian incivilities per interaction was run through chi-squared analysis, there was once again no statistical significance found in any dimension (deception could not be tested). The p-value regarding the total count of interruptions was p=.7673 and the p-value regarding the total count of directed profanities was p=.7473.

When the data on % chance of civilian-directed incivility was put into chi-squared tests, there was again no statistical significance for incivility in any dimension. The p-value for officer interruptions was p=.7766. The p-value pertaining to officers’ use of directed profanity was p=.5028. When the data that accounts for the total count of officer incivilities per interaction was run through chi-squared tests, there was no statistical significance in either dimension. The p-value for the total count of interruptions was .1238 and the p-value for the total count of directed profanity was .4629.

When the data from this study was run through chi squared tests, the p values suggest that any differences that are found in the descriptive statistics above could have been credited to chance.

T-Test Analysis

When the data from this study pertaining to officer-directed incivility was plugged into t-tests, there was no statistical significance for incivility (in any dimension) for either race or gender. The p-value for the interaction of gender and civilian interruptions is p=.6491 and p=.6574 for the interaction between race and interruptions. The p-value regarding the interaction of gender and civilian use of directed profanity was p=.0921 and p=.5981 for the interaction between race and profanity. The p-value for the interaction of gender and civilian deception was p=.1625 and p=.9103 for the interaction between race and deception.

When the data from this study pertaining to civilian-directed incivility was plugged into t-tests, the interaction between gender and officer use of profanity was the only statistically significant finding (p=.0033). The p-value for the interaction between gender and officer interruptions is p=.9194 and p=.4002 for the interaction between race and officer interruptions. The p-value for the interaction between gender and officer’s use of directed profanity was p=.0033 and p=.7414 for the interaction between race and use of directed profanity.

When the data from this study was run through t-tests, the p-value for all but one of the tests showed no statistical difference. The only statistically significant interaction (p=.0033) between any quantitative variable and the dichotomous variable was the interaction of gender and directed profanity. This finding means that police officer are more likely to use directed profanities at male civilians.

DISCUSSION//DRAFTS UNFINISHED IDEAS

ADD CONTEXTUAL FACTORS INTO RESULTS…

Results show of 156 one-on-one interactions, 35.89% of those interactions involved at least one dimension of officer-directed incivility. Interestingly, incivility experienced by civilians has a similar frequency at 30.76%. This reciprocal (ish) nature of communication is consistent with Lakin and Chatrad’s work (2003) on “the chameleon effect” which found this phenomenon to be causally associated with interpersonal rapport.

While the frequency of officer and citizen incivility roughly mirror each other, the number of times incivility occurs in a given interaction differ greatly. Civilians have nearly the same odds of using directed profanity and interruptions as officers do, however, civilians averaged x7.95 profanities and x4.3 interruptions compared to officers x2.44 and x3.25 respectively. DOUBLE-CHECK THESE NUMBERS

One interesting finding was that minority citizens’ rates of incivility are higher than their white counterparts (x11.8 profanities x5.75 interruptions compared to x6.67 x3.93 respectively). LITERATURE ABOUT MINORITY PERCEPTION OF POLICE…DISLIKE/DON’T RESPECT/FEEL LIKE THEY WILL BE TREATED UNFAIRLY….This is interesting because officer rates of incivility remain relatively even (x3 minority directed profanity and x3.33 minority directed interruptions compared to x3.24 and x2.33 to their white counterparts). While incivility is usually directed at males, it is promising to see no variance in officer incivility based on race. GO MORE INTO DIFFERENCES BASED ON CHARACTERISTICS

RATE DIFFERENCES BETWEEN OFFICERS/CIVILIANS (EMOTIONAL LABOR/SUPPRESSION) OFFICERS GENERALLY USE LESS INCIVILITY…. AT LEAST THE DIMENSIONS WE CODE FOR IN THE LAB.

WHILE THERE IS NO ONE GROUP THAT SHOWS A HIGHER LIKELIHOOD OF INCIVILITY, MINORITY CIVILIANS IN THIS STUDY HAVE THE HIGHEST AVERAGE OF INCIVILITY…..