Understanding local reported-crime statistics

Reported offences are context, not a safety score

Bilby shows municipality-level reported offences from Brå, a rate per 100,000 residents, a compatible police-region comparison and available category shares. It does not calculate a safety score or infer the chance that a particular resident or plot will experience crime. Reported offences and all crime are not the same quantity.

Period and geography come with the dataset

The active dataset configuration selects the reference year, metric, source and regional comparison rates. The card shows that year; it is not automatically the current year. Bilby reads the municipality’s published count and per-population rate from the configured Brå table rather than deriving population from the plot.

Municipality pages already know their municipality/county. Plot context can use supplied identities or reverse-geocoded municipality/county candidates. Matching normalises names, including the word “kommun”. A missing or invalid active dataset, unmatched municipality or provider error produces unavailable statistics, not zero offences.

Rate comparison and category shares

The published rate expresses offences ÷ population × 100,000; Bilby uses the supplied rate rather than independently reconstructing it. Comparison = round((municipality rate − police-region rate) ÷ police-region rate × 100). County-to-police-region mapping selects the comparison region. Both calculation and presentation require matching year and metric.

Available category counts come from the municipality sheet in the configured Brå workbook: offences against the person, theft/robbery, fraud, damage, traffic and drugs. Each displayed share is round(category count ÷ total reported offences × 100). These selected categories need not sum to 100%. Bars are scaled relative to the largest displayed category for readability; read the printed percentage for the actual share.

Fictional municipality comparison

Assume a fictional municipality has 2,000 reported offences and a published rate of 10,000 per 100,000 residents. Its compatible police-region rate is 12,500. The comparison is (10,000 − 12,500) ÷ 12,500 × 100 = −20%, shown as 20% lower. This is a comparison of rates, not “20% safer”. If a selected category contains 300 reports, its share is 300 ÷ 2,000 × 100 = 15%.

If the region value is for a different year, no comparison is shown. If the category workbook cannot be read, the total and rate can still appear without category bars.

Limits of interpretation

Reported counts depend on reporting, recording, classification and enforcement as well as offences occurring. A report is not necessarily a unique victim, convicted person or incident. Commuters, visitors and commercial activity can affect counts relative to resident population. Municipality totals hide neighbourhood differences and cannot establish current conditions at one address.

Small counts and rounded shares can move sharply between years. The UI does not adjust for those differences or turn a lower regional comparison into a recommendation about residents or neighbourhoods. Missing categories are not treated as zero. Changing the selected house, travel preferences or budget does not change the Brå statistics.

Source and review dates

The active-dataset reader, municipality matching, workbook calculations and local-context card were reviewed 18 October 2026. The actual publication year and source link come from the active dataset displayed in the product; no uninspected live dataset year is asserted here. Brå explains what reported-offence statistics measure and their limitations. Brå