ATTITUDES TOWARD THE POLICE: MEASUREMENT ISSUES

 

The use of surveys as a tool to assess police-citizen interactions dates back to the 1970s in the United States and the United Kingdom. Although early surveys established that most people are supportive of the police and satisfied with the way they perform their duties, they also identified segments of the population (especially blacks, the young, and males) who are less satisfied than others (see, for example, Lundman 1974; Thomas and Hyman 1977). Today, citizen surveys are used for a variety of purposes, including tracking the prevalence and nature of citizens’ encounters with the police among different ethnic groups (Ewin 1996), evaluating community policing initiatives (see Skogan and Hartnett 1997), and cross-national comparisons of police-citizen interactions (see Davis et al. 2004). Increasingly, citizen surveys have also been used by local police administrators to gauge community opinion of the police, diagnose problems, and measure the success of new initiatives. Surveys can also be used by municipal government administrators or watch-dog agencies to assess police performance and hold the police accountable for how they interact with the public.

Police researchers have come to define two types of citizen surveys: community surveys and contact surveys (Maguire 2004). The former refers to surveys that attempt to gauge the attitude of the public toward the police. In effect, it is a way of “taking the temperature” of the citizenry in a way that is less biased and more precise than assessing public opinions through stories and editorials in the media or other outlets of public discontent. Community surveys can help progressive police commanders to know when they are on track and when they are in trouble with the entire community or segments of the community defined by race, class, or other demographics.

Contact surveys are administered to people who have initiated contact with the police (voluntary contacts) or who have been stopped or detained by the police (involuntary contacts). Instead of measuring general attitudes, which is what community surveys measure, contact surveys attempt to measure the quality of specific encounters. Although the number of complaints against the police may serve as an indicator of excessive police abuse, contact surveys can assess the full spectrum of police-citizen interactions, from highly positive to highly negative. Police commanders can use contact surveys to determine if some districts are handling public interactions better than others, and take appropriate corrective action.

Issues with Community Surveys

According to the Bureau of Justice Statistics’ LEMUS survey, community surveys have come into wide use: Roughly one in three police departments conducts community surveys each year (Weisel 1999). To produce an unbiased estimate of community sentiment regarding the police, community surveys must attempt to sample a representative group of people from the community. The best way to achieve a representative sample is to draw households randomly from census data or telephone lists, and then interview those selected over the phone or in person. Because these methods are relatively costly, many departments that conduct community surveys on a regular basis use the less costly methods of mail surveys or online surveys through their websites. But there are problems inherent with both methods. Online surveys are completed by those who have access to a computer and who were motivated to seek out the department’s website where they found the survey. Mail surveys are likely to have a low response rate. Both methods, while inexpensive, raise the question of whether the households that answer the survey are representative of the community at large.

Regardless of how the surveys were done, the results of community surveys may be difficult to interpret. When citizen satisfaction is assessed using global questions, the conceptual foundation of the resulting measure is vague or fuzzy (Worrall 1999; Wells et al. 2005). Recent research asking people to explain the reasons for their global assessment of the police demonstrated that some respondents base their attitudes on specific encounters, whereas others base their attitudes on general impressions of their local department, presumably formed through the media or conversations with acquaintances (Frank et al. 2005). Thus, the meaning of global satisfaction measures is unclear, and may depend as much on what is in the local media as on departmental priorities and policies or conduct of officers.

Issues with Contact Surveys

Respondents in contact surveys may be drawn from either the police files of persons who have had specific types of encounters with police officers or from respondents to community surveys who indicate that they have had a recent encounter with a police officer. The former source is workable if there is a high proportion of citizens who have had recent police encounters. This is likely to be the case in urban, but not in suburban or rural areas. When using police files as a sampling resource, researchers must decide which subpopulations of experience with the police they are interested in. Voluntary contact samples may be drawn from crime reports, from reports of domestic violence incidents, from persons who walk into stationhouses requesting assistance, or from callers to crime or drug hotlines. Involuntary contact samples may be drawn from summonses, arrests, traffic stops, or stops of people on the street. These sources are likely to vary in accessibility (whether the information is available in a computer database or other central repository) and in the completeness of contact information (phone number or address). Researchers must balance ease of access and quality of information against the purposes of the research project to determine which method is best for them.

Using Citizen Surveys in Cross-Jurisdictional Comparisons

Increasingly police experts and researchers have argued for developing methods to compare the performance of law enforcement agencies as a way to promote greater police accountability to the public (Maguire and Uchida 2000). In England and South Africa, where police forces are national and the movement for comparative measures is strong, citizen surveys have been incorporated as an integral part of performance measurement schemes.

However, it has been well documented that citizen surveys, especially those that measure global community opinion rather than reactions to specific encounters, show substantial variation across demographic groups defined by gender, age, and race (see, for example, Skogan 2005). Therefore, if citizen surveys are to be used to measure comparative performance of police agencies, results ought to be adjusted for demographic differences in the communities being compared, as suggested by Sherman’s work on evidence-based policing (Sherman 1998).

Issues involved in comparing survey results across jurisdictions in different countries are even trickier as cultural differences come into play. In surveys using common items in New York and St. Petersburg, Russia, Davis and colleagues (2004) noted that a much larger proportion of Russian respondents registered “Don’t Know” on questions related to police misconduct. Subsequent investigation of the source of the difference revealed that Russians were less willing than Americans to express an opinion about taking bribes, physical abuse, and so forth, if they had not had firsthand experience with that form of misconduct.

Next post:

Previous post: