Making the Grade? Public Evaluation of Police Performance in Anchorage

Brad Myrstol

Myrstol, Brad A. (Summer 2005). "Making the Grade? Public Evaluation of Police Performance in Anchorage." Alaska Justice Forum 22(2): 5-10. A randomly selected sample of 2,485 adult residents of the Municipality of Anchorage participated in the Anchorage Community Survey, a telephone survey conducted by the Justice Center over a five-month period in late 2004 and early 2005. This article presents data from the eleven questions in the survey which pertained to public perceptions of and experiences with the Anchorage Police Department.

Since the mid-1980s the idea of community-oriented policing has captured the imaginations of police administrators and ordinary Americans alike. For some perspective on how far reaching community-oriented policing has become, consider data collected as part of the Law Enforcement Management and Administrative Statistics (LEMAS) program. According to a survey of over 3,000 state and local law enforcement agencies conducted in 2000, over 60 percent of all local police and sheriffs departments employed full-time sworn officers whose primary responsibility was community policing activities. In Alaska, a total of 98 different agencies and organizations have been granted in excess of $41 million in federal Community-Oriented Policing Service (COPS) monies (see prior Forum articles for various discussions of community policing in Alaska).

A driving force behind the community policing movement is a desire to reduce the level of alienation between police and citizens and the lack of police accountability, both believed to have been by-products of the professional model of policing that came to prominence in the Progressive Era of the early twentieth century. While the professional model of policing comprises many characteristics, a distinguishing feature—perhaps the distinguishing feature—is the explicit separation of the police from the public. As an orienting philosophy, the professional model’s deliberate construction of a boundary between the police and the public derives from the concerns of Progressive Era reformers about machine politics and systemic corruption among police and other government entities. Officers were given jobs based on political patronage, with hiring decisions based on an officer’s political or ideological allegiance. To overcome this political corruption, reformers instituted the civil service system in which employment decisions were to be based upon objective standards and merit rather than the political affiliation of an applicant or the individual whim of a supervisor.

An unintended, and ironic, consequence of the professional model was a dramatic reduction in police accountability, since police—now protected by civil service provisions—could operate relatively free from citizen oversight and control. Technological innovations, especially the advent of the automobile, and the rise of a management culture—specifically, the adoption of business management models stressing efficiency—have only exacerbated police-public alienation as police have been encouraged to adopt a reactive rather than proactive posture and thus withdraw even further from the public sphere into patrol cars and specialized units. Instead of getting to know members of the public and learning of problems through impromptu and informal interactions, officers gain knowledge through responding to calls for service, which tend to be negative encounters, or by glancing through the windows of a cruiser.

To address this police-public alienation and restore lost institutional legitimacy, scholars and police administrators alike have advanced a community-oriented approach that serves to make the boundary erected between police and the public more permeable, though not completely porous, since police in most jurisdictions are still considered civil servants. At the heart of the community policing paradigm is a commitment to increased police-public interaction and accountability to the public.

For nearly a decade, the Anchorage Police Department has engaged in a variety of activities falling under the rubric of community policing. Most notable among these efforts have been the Community Action Policing Team project implemented in the Mountain View community council area in the mid-1990s and the Cops in Schools program. The department’s commitment to the underlying philosophy of community-oriented policing can also be seen in less public, non-operational ways —for example, in the way it has actively sought community feedback on its performance and in its efforts to seed an organizational culture sensitive to the needs and desires of a growing and increasingly diverse city. One element identified in the department’s strategic plan formulated after an internal review in 2003 was the need to improve public perception of the department.

Recognizing the need for reliable data with which to determine the public’s perceptions of the performance of APD, the Justice Center at the University of Alaska Anchorage dedicated an entire section of its Anchorage Community Survey, 2005 to empirically measuring residents’ evaluation of APD’s performance. Included in the Anchorage Community Survey were eleven items relating to perceptions and experiences with APD. Six of these items asked residents to rate APD on the following dimensions of performance:

  • Responding quickly to calls for help or assistance;
  • Not using excessive force;
  • Being helpful and friendly;
  • Treating people fairly;
  • Investigating and/or solving crimes; and
  • Preventing crime.

Respondents were also asked how much confidence they had in the department as well as how accessible they felt the department to be.

In order to gauge the extent to which personal experience with the police influences individual attitudes towards them, the Anchorage Community Survey also included items measuring both direct and indirect experiences with APD officers:

  • Direct: “Have you, yourself, come into contact with an Anchorage police officer for any reason in the past 12 months?”; and
  • Indirect: “Has anyone you know well, such as a family member or close friend, come into contact with an Anchorage police officer for any reason in the past 12 months?”

Finally, those who indicated having come into direct contact with an Anchorage police officer were asked to evaluate that experience:

  • “Thinking about your most recent experience with an APD officer, would you, yourself, characterize the officer’s behavior as competent? By ‘competent’ I mean the officer handled things in a manner you thought was appropriate for the situation?”

The first question addressed in this article is: How do Anchorage residents perceive the Anchorage police department’s performance? Percent distributions and mean scores for each item are presented in Table 1. These results show that, in general, Anchorage residents have a positive view of Anchorage police performance. This finding corresponds to an extensive research literature on public perceptions of the police which shows that the police enjoy widespread support among the general population. It also squares with previous Justice Center research examining public attitudes toward APD (see “A Further Perspective on Satisfaction with Policing” by Matthew Giblin in the Fall 2003 Forum).

Table 1. Survey Results: Public Evaluation of Police Performance in Anchorage

Officer helpfulness and friendliness garnered the best rating (1.93), followed by their use of force (2.10), the fairness with which they treat people (2.13), and response time (2.24).

Ratings of police performance with regard to crime-control activities were lower, however. Only a slight majority of respondents rated APD as “excellent” or “pretty good” for investigating and/or solving crimes (average rating = 2.41), and a minority of respondents gave APD’s crime prevention a favorable rating (average score = 2.68).

In terms of access, nearly all respondents (88.6%; average rating = 1.77) reported that they felt APD was more accessible than not, with more than one-third stating that the services of APD were “very accessible” to them. For the final assessment of APD performance, respondents were asked how much confidence they had in the Anchorage Police Department. Just over 60 percent of those who responded had “a great deal” or “quite a lot” of confidence, while more than a third stated they had only “some,” “very little,” or no confidence in APD.

When asked about their experiences with APD, better than six out of ten respondents reported that they had had personal experience with an APD officer in the 12 months preceding their participation in the survey (47.8% direct experience; 43.3% indirect experience). The direct experience figure corresponds almost exactly with previous Justice Center research, which reported a figure of 49.3 percent. Thus, it would appear that contact with an APD officer is not uncommon for Anchorage residents.

To summarize, the answer to the first question—How do Anchorage residents perceive the Anchorage police department’s performance?—is, in the most general of terms, “positively.”

We now move on to our next question: Do the ratings provided by Anchorage residents vary significantly according to the characteristics of respondents?

The answer to this question is a bit more complicated. In order to simplify the presentation of findings somewhat, this part of the analysis is limited to the first six APD performance measures: response time; use of excessive force; helpfulness/friendliness; fairness; criminal investigation; and crime prevention activities. The individual characteristics of respondents examined in this section are:

  • Age;
  • Race/ethnicity;
  • Hispanic background/origin;
  • Gender;
  • Educational attainment;
  • Current work status;
  • Residential tenure;
  • Household size; and
  • 2003 household income.

Average scores for the six performance measures are presented for each demographic variable in Table 2.

To assess whether or not there were statistically significant differences in respondents’ ratings, the average scores for each category or level of a variable was compared to every other category or level. For example, for the variable age, the average score of 18-24 year-olds for each performance measure was compared with those of 25-34 year-olds, 35-44 year-olds, 45-54 year-olds, 55-64 year-olds, as well as those age 65 and older; then the average score of 25-34 year-olds for each performance measure was compared to those of 35-44 year-olds, and so on. A total of 54 statistical models of this sort were constructed—one for each individual characteristic and performance measure combination. The results of this analysis are presented in Table 3.

Table 3. Statistical Significance of Between-Group Differences (ANOVA) by Respondent Demographic Characteristics

Table 3 illustrates that evaluations of police performance do vary significantly by demographic characteristics. The characteristics found to display the most consistent variation across performance measures are, in descending order of importance: current work status (significant differences for every performance measure); age (significant differences for 5 of 6 performance measures); educational attainment (significant differences for 5 of 6 performance measures); race/ethnicity (significant differences for 5 of 6 performance measures); gender (significant differences for 5 of 6 performance measures); household size (significant differences for 3 of 6 performance measures); residential tenure (significant differences for 2 of 6 performance measures); Hispanic background/origin (significant differences for 2 of 6 performance measures); and household income (significant differences for 2 of 6 performance measures).

The APD performance measures inducing the greatest variation across respondents were, again in descending order: treating people fairly; helpfulness and friendliness; use of force; response time; investigating/solving crimes; and finally, crime prevention.

More detailed analyses show that these significant differences were very specific in nature. For example, it was not simply that work status mattered, but that it mattered in very particular ways. Understanding these particulars is critical to our understanding of citizens’ perceptions of police (and presumably other legal institutions as well). So our next question is how, precisely, did each of the demographic variables matter?

Age. Our analyses revealed that the most consistent differences in respondents’ evaluations of police performance existed between those aged 65 and over and everyone else. Furthermore, while senior citizens differentiated themselves from those in all other age groups, the largest gap was found between them and respondents in the 18-24 age group. Another pattern detected in the data was that in no case did an older group offer a more negative evaluation of the police than a younger age group; that is, evaluations of police performance were monotonic—without exception, they got better as respondents got older.

Race/ethnicity. The statistically significant differences found between racial/ethnic categories were limited to White/Caucasian–Alaska Native/American Indian and White/Caucasian–Black/African American contrasts. In all cases, where significant differences were found, respondents who reported their race/ethnicity as White/Caucasian provided a significantly more favorable evaluation of APD performance than their Alaska Native/American Indian and Black/African American counterparts. With respect to the White/Caucasian–Alaska Native/American Indian contrast, significant differences existed for evaluations of response time, use of force, helpfulness/friendliness, and treating people fairly. Significant differences for the White/Caucasian - Black/African American were limited to respondent evaluations of use of force and treating people fairly.

Educational attainment. The effects of formal education on evaluations of police performance were more mixed than for many of the other demographic comparisons. In general, however, it can be said that respondents with more formal education offered more favorable views of police performance. In particular, those with advanced educations—graduate or professional degrees—consistently gave more favorable opinions of APD’s work. Those with associate’s degrees, however, tended to be more critical than both those with less education and those with more. An acute difference was seen when those with associate’s degrees were compared with those with a bachelor’s degree or higher.

Current work status. Most powerful among all the demographic factors examined was a respondent’s current work status. This was the only demographic characteristic that displayed significant between-group variation for every performance measure. At the heart of these differences was whether or not a respondent was retired. Retirees were found to exhibit dramatically different views of APD’s performance than all other work categorizations. The direction of these differences was such that retirees always provided a more positive evaluation —regardless of the performance measure. Other group differences did emerge, however. In particular, differences were detected between those respondents who reported being in the work force (both employed and unemployed) and those who were not in the work force (i.e., homemakers and full-time students). Thus, three distinct work status categories emerged with respect to the evaluation of police performance: those currently in the work force; those retired from the work force; and those who were not currently in the work force.
Household income. In general, household income was found to demonstrate little variation for the evaluation of the Anchorage police department. The only two statistically significant differences detected existed between the same two groups—those whose households earned between $25,000-$49,999 and those with reported household incomes of $80,000 or more. Both differences were for use of force and treating people fairly.

Household size. Statistically significant differences between household size categories existed for three police performance measures: response time; helpfulness; and crime prevention. All of the statistically significant differences found were between single resident households and those with 5 or 6 residents.
Residential tenure. Whether or not a respondent had lived in the current residence for 5 years before the survey was found to demonstrate only marginal variation in evaluations of APD performance. Differences between those who had lived in their residence for at least 5 years and those who had lived in their residence for less than five years existed for only two performance measures: helpfulness/friendliness and treating people fairly. In both instances, those who reported living in their current residence for at least five years provided significantly more favorable evaluations of APD.

Gender. While respondent’s gender was found to have a strong influence on evaluations of police performance, the gender differences found did not indicate a clear gender gap in evaluations of police performance. It was not the case that women consistently offered more favorable evaluations of Anchorage police than men, or vice versa; rather, gender appeared to influence evaluations of police in very specific ways. Women offered more favorable evaluations with respect to: helpfulness/friendliness and investigating/solving crimes; they provided more negative evaluations than men for: use of force; treating people fairly; and preventing crime.

Hispanic origin/background. A person’s Hispanic background or origin produced only two significant differences in evaluations of APD’s work performance. As a group, Hispanic respondents reported lower ratings of APD with respect to helpfulness/friendliness and treating people fairly than did non-Hispanics.

To summarize the analysis to this point, the pattern of responses observed in these data was shaped at least in part by the demographic characteristics of those who participated in the study. Notably, the influence of these individual-level characteristics was not uniform: some factors appeared to be consistently influential, while others seemed to have little bearing on citizen evaluations of police performance.

Up to this point, the variation attributable to each of these demographic factors has been detected only when each of them was examined in isolation from all others. But, of course, we know that in the real world individuals bring their entire biography into social situations, including social science interviews. While it seems obvious, it is important to remember that people are more than any single categorization—not simply “Alaska Native/American Indian” or “female” or “between the ages of 25 and 34.” To understand the relative impact of each demographic factor on evaluations of police performance, it is important to examine each of them within the context of all the others.

In order to get a better assessment of the relative significance of each demographic factor, an ordinary least-squares regression (OLS) was used. The goal was to predict the values of the six measures of APD performance. The impact of each demographic variable on respondent evaluations of police performance was examined while controlling for (i.e., holding constant) the impact of all the others.

The OLS analyses revealed three demographic characteristics to be consistent predictors of public evaluations of police performance: work status; educational attainment; and race/ethnicity. As with the bi-variate analyses reported earlier, current work status was the most consistent and powerful predictor of evaluation of police performance. The effect of work status was quite specific—retirees were significantly more likely than those in other categories to render a positive evaluation of APD job performance.

Table 4. OLS Regression Results

Respondents’ educational attainment was also found to be a highly significant predictor of their evaluations, though the specific effects of education varied more than for current work status. Those with at least a bachelor’s degree were more likely than those with less formal education to give a positive evaluation of APD performance for four of the six measures: use of force; helpfulness and friendliness; treating people fairly; and crime prevention. In contrast, respondents with an associate’s degree were more likely than those with more education, as well as those with less education, to provide a more negative evaluation of police performance for response time and investigative activities.

The third demographic characteristic found to demonstrate consistent predictive utility for the public evaluation of police performance was race/ethnicity: whether a person was White/Caucasian was significant in predicting the evaluation of APD’s performance. Specifically, those respondents who reported their race/ethnicity as White or Caucasian were significantly more likely than non-Whites to offer a more positive evaluation for five of the six dimensions examined. (Race was not a significant factor for evaluations of crime prevention activities.)

Two demographic variables were found to have inconsistent effects. Age and gender were each found to influence respondents’ evaluations of police performance for three of the six measures examined. Age influenced public perceptions of response time; helpfulness and friendliness; and treating people fairly. Respondents aged 65 and older were significantly more likely to offer a positive evaluation of APD’s response time than those younger, while those in the 25-to-34 age group offered more negative views of police helpfulness and friendliness and fairness than those in other age groups. Holding all other demographic categories constant, significant differences emerged between men and women on helpfulness and friendliness, investigative activities and crime prevention; men offered more negative evaluations than women for all three measures.

Three variables had significant predictive power for only one performance measure. The length of time a person had lived in the current residence was significant for determining public perceptions for police use of force. When other demographic characteristics were taken into account, whether or not a person had lived in the current residence 5 years preceding the survey was found to be a significant predictor of their evaluation of police use of force. Specifically, those who had lived in their current residence for at least 5 years were more likely to give a negative evaluation. With respect to household size, single residents were more likely to offer a positive evaluation of police crime prevention activities than larger households. Household income was also found to be significant for only one performance measure. A respondent’s household income was a significant predictor of evaluation of police use of force. In particular, those who reported a household income of at least $80,000 were more likely than those whose households earned less to provide a positive evaluation.

The final demographic characteristics was Hispanic background/origin. This factor was not found to be a significant predictor of public evaluations of police performance—for any measure.

To summarize this section, the most powerful predictors of Anchorage residents’ evaluations of police performance were whether a respondent:

  • Was retired;
  • Had attained a bachelor’s or graduate degree; and
  • Was White/Caucasian.

Other factors found to be significant were:

  • Age; and
  • Gender.

Demographic characteristics found to have little or no predictive power were:

  • Hispanic background/origin;
  • Residential tenure;
  • Household size; and
  • Annual household income.

I began this article with a brief overview of community policing. The gist of that discussion was that at the heart of the community policing movement is a commitment on the part of police organizations to reintegrate themselves with the communities they police and of which they are members themselves. A big part of the effort at reintegration is to increase departments’ accountability. One way for a department to do this is to ask residents for feedback on the job it is doing. Sometimes the results provide departments and their officers a nice pat on the back, and sometimes public feedback gives police agencies a clear indication of what aspects of performance need improvement.

The data presented in this article do both of these things. It is clear that in general Anchorage residents are supportive of the job APD is doing with respect to responding quickly to calls for service, not using excessive force, being helpful and friendly, treating people fairly, investigating/solving crimes, and preventing crime. However, these data also clearly show that residents’ evaluations of their police department are contingent; there is not a consensus of views. That there is substantial variation in opinion according to a person’s demographic and social characteristics affords APD the opportunity to identify not only those who feel well-served by the department, but also to take note of those who reveal some disconnectedness from it.

Brad Myrstol is a research associate with the Justice Center.