Academic research, police operations in collaboration with municipalities, and surveys as tools - 25 years of experience

• In Event: Measuring and Interpreting Perceptions of Crime

Abstract

The Local Safety Measurement System was introduced in The Swedish Police in the late 1990s. By 2024, more than 1.4 million respondents have responded to police surveys in most of the Swedish Municipalities. Data has been collected with the same questions, sequencing, and data collection methods, with few changes. In 2024, the Swedish Police decided to extend the survey to all 290 municipalities in Sweden. In Sweden, the Local Governments in 2024 have a statutory responsibility for crime prevention and are the main cooperation partner to the local police. Dialogue and collaboration are greatly facilitated by the results from the measurement, as previous experiences show.

Of particular interest is that in 2022, it was decided to move to an innovative model of what was good, less good, and what was a directly bad percentage for an indicator in a residential area. In addition, it is possible to use a summary of the assessments on each indicator to an overall assessment of the severity of the current problem picture in a national context.

Each indicator in the model consists of seven problem levels, 0 – 6. Problem level 2 is the model's starting point and is extra broad as a "normal value for Sweden." An attempt to become almost certain that the value does not deviate negatively or positively from other municipalities.

The length of the intervals is determined for problem levels 0 – 6 by using the average values of the confidence intervals of the indicators using the normal 360 observations.

This is a unique story about analysis using algorithms, The Clopper Pearson interval, confidence interval, forecasting science, and police experience. For the safety and best interests of the residents.

Author

• Kjell Elefalk, Senior Advisor