Research Summary: Police Performance and Surveys

PDF (96 KB)

Some key police performance metrics must be gathered through public opinion surveys. Little reliable information of this type is currently collected. Improving the validity and frequency of collected data would enhance police performance measurement.

Police undertake a broad array of work. Other than the traditionally-assigned tasks of pursuing, arresting and charging criminals, preventing crime from occurring, and dealing with traffic-related offences and accidents, police are further expected to resolve various conflicts in their communities, reduce or prevent social disorder, and construct and maintain community relations. It is important to understand that the measurement of police performance is a complicated task that has multiple dimensions. There is no single measure that will be even remotely close to measuring the performance of police in all their tasks.

Both direct and indirect measures need to be accounted for when attempting to measure police performance. Direct measures of police performance commonly used include crime rates, number of arrests and fines issued, clearance rates, and call for service response time. Some indirect measures of police performance include public opinion surveys, direct observations of social behaviour, situational studies and independent testing.

The measurement of police performance through public opinion polling is of two kinds: 1) general questions on satisfaction with police; and 2) specific questions on police performance. The general questions on satisfaction with police asked on surveys are supposed to be the simplest and quickest way to measure the overall level of satisfaction of citizens with their police service. These questions provide a quick indicator for the overall support for police among the citizens; carry implications for the support police receive from the public in accomplishing their mandate; and infer whether the decrease in the perceived legitimacy or effectiveness of the police could potentially lead to non-compliance with the authority of the police and increased crime rates.

However, generalized questions on satisfaction with police tend to be too general to tease out specific information on what it is that citizens favour about their police services. More importantly, it is impossible to tease out the reasons explaining why the citizens are unhappy with their police. Other methodological issues associated with generalized questions include unstandardized, inconsistent, and at times poor choice of wording on questions and response categories, which make comparisons across time and place impossible. There seems to be confusion when it comes to the meaning of terms “favourable views of,” “confidence in,” and “trust of” the police; these terms tend to be used interchangeably on surveys, when in fact these terms measure different public sentiments.

When it comes to specific questions on police performance, there are many examples of the types of questions that are being asked of police in contemporary public opinion surveys in Canada and the Western World. Unfortunately, these questions are used in a more or less arbitrary manner without being properly tested for validity and reliability. Ground work on the meaning of the questions, whether they measure what they are supposed to measure (reliability), and how well do they measure it (validity), is largely absent in the literature.

There are three notable studies that attempt to conceptualize and test specific measures of police performance measured through public opinion polling: 1) Maguire and Johnson (2010) in the U.S. who derived questions from theoretical constructs and tested them against Mastrofski's six dimensions of policing as a service industry; 2) Coleman (2012) in Canada who consulted policing stakeholders in order to derive and test the factors that may constitute police performance; and 3) Jackson et al. (2011) in Europe who derived, tested and implemented European indicators of trust in judicial systems.

In Canada, only one national survey administered by Statistics Canada every five years (General Social Survey, Victimization cycle) asks six questions on police performance. This constitutes the only source of public opinion of police performance metrics that are comparable across time and place in Canada. The majority of Canadian municipal police services commission their own annual or bi-annual public opinion community surveys that include numerous general and specific police performance questions. However, the questions tend to be unstandardized and inconsistent, with varying response categories, thereby making them incomparable across time and place, and of little analytic value.

If police services want to improve the measurement of public satisfaction with their services, more research into either testing the existing questions or creating new standardized questions is required. A project similar to the creation of European indicators of trust in justice (Jackson et al, 2011) that involves discussion among experts, proper testing and re-testing of measures of police performance is one example. In the absence of valid, reliable and standardized indicators police services and policy makers will continue to ask different, sometimes vague questions that are not comparable across place and time. Such inconsistent measurement of police performance runs the risk of misevaluating the performance of police services, the policing policies and practices they apply, leading to inefficient and ineffective policing.

Maslov, A. (2014) Measuring the Performance of the Police: The Perspective of the Public. Ottawa, ON: Public Safety Canada.

Additional Sources

For more information on research at the Community Safety and Countering Crime Branch, Public Safety Canada, please contact:
Research Division, Public Safety Canada
340 Laurier Avenue West
Ottawa, Ontario K1A 0P8

Research Summaries are produced for the Community Safety and Countering Crime Branch, Public Safety Canada. The summary herein reflects interpretations of the report authors' findings and do not necessarily reflect those of the Department of Public Safety Canada.

Date modified: