Crime statistics

Data extracted in February 2020

Planned article update: July 2020


Robberies down by 30 % in the EU between 2012 and 2017.
There were 4 300 intentional homicides and 589 000 assaults in the EU in 2017.
Crime statistics robbery-Feb-2020.jpg

The statistics presented in this article are based on official figures for police-recorded offences (criminal acts) in Europe during 2008-2017. The results cover the European Union (EU), the United Kingdom, as well as partially the EFTA countries, the candidate countries and the potential candidate countries. Data for the United Kingdom are presented separately for England & Wales, Northern Ireland and Scotland due to different jurisdictions.

Full article

Robberies down by 30 % between 2012 and 2017

Between 2012 and 2017, police-recorded robberies in the EU fell by 30.3 %, to about 317 000. By contrast, there was a 8.8 % increase between 2008 and 2012, the high point during the period 2008-2017, as shown in Figure 1. Between 2016 and 2017 the EU total fell by 22 000, only Ireland, Cyprus, Romania, Slovenia and Sweden had some increase (together an increase of 266 robberies).

Figure 1: Robbery (police-recorded offences), EU-27, 2008-2017
(number of offences)
Source: Eurostat (crim_off_cat)

Figure 2 shows police-recorded robberies relative to population size (number of offences per 100 000 inhabitants). In 2017, the highest rates were observed in Belgium (167), France (150), Spain (144) and Portugal (115), while the lowest rates were found in Romania (16), Estonia and Czechia (both 15), Cyprus (14), Slovenia (12) as well as Slovakia and Hungary (both 9). Within the United Kingdom the rate was highest in England and Wales (132). Among the EFTA countries, Switzerland had the highest rate at 21 police-recorded robberies per 100 000 inhabitants.

Figure 2: Robbery, 2017
(police-recorded offences per hundred thousand inhabitants)
Source: Eurostat (crim_off_cat)

4 300 intentional homicides in the EU in 2017

There were around 4 300 police-recorded intentional homicides in the EU in 2017 (excluding data for the Netherlands), a reduction of 22 % since 2008. Table 1 shows the reported figures by country.

Table 1: Intentional homicide, 2008-2017
(number of police-recorded offences)
Source: Eurostat (crim_off_cat)

Figure 3 shows intentional homicide relative to the population size (police-recorded offences per hundred thousand inhabitants). In 2017, the highest figures were observed in Latvia (5.6), Lithuania (4.0), Estonia (2.2) and Malta (2.0), and the lowest in Czechia and Italy (both 0.6) as well as Luxembourg (0.3).

Figure 3: Intentional homicide, 2017
(police-recorded offences per hundred thousand inhabitants)
Source: Eurostat (crim_off_cat)

589 000 assaults in the EU in 2017

In the EU, police-recorded assaults numbered around 589 000 in 2017. As shown in Figure 4, the trend was generally decreasing during 2010-2017. However, it should be noted that the EU totals for 2016 and 2017 are partly based on previous figures. France did not report for 2017 and Hungary not for 2016-2017. In France there has been an increasing trend from 232 000 in 2010 to 243 000 in 2016. The reported figures for Hungary show a decrease, from 14 600 in 2010 to 12 500 in 2015.

The number of police-recorded assaults varies widely across the EU, even relative to population size. Different laws, reporting rate and recording practices affect comparisons. For instance, in addition to serious assault, some national figures include threats, minor assault, lethal assault (manslaughter, murder, etc.) or sexual assault (which usually is counted separately).

Figure 4: Assault, EU-27, 2010-2017
(number of police-recorded offences)
Source: Eurostat (crim_off_cat)

573 000 cars stolen in the EU 2017

In the EU there were about 573 000 police-recorded car thefts in 2017, a 35 % reduction compared with 2008. As shown in Figure 5, there has been a downward trend in the EU between 2008 and 2017. However, Ireland, Greece, Spain, Latvia, Malta, and Romania had an increase in police-recorded car-thefts between 2016 and 2017. For Belgium, France, Luxembourg, and Hungary, the 2017 figures are not yet reported.

Figure 5: Theft of a motorized land vehicle, EU-27, 2008-2017
(number of police-recorded offences)
Source: Eurostat (crim_off_cat)

Relative to the population size, the figures were highest in Luxembourg (328), Greece (269), Italy (257), Sweden (256), France (247) and Czechia (238), for police-recorded car thefts per 100 000 inhabitants (average 2015-2017). The lowest figures in the EU were observed in Slovakia and Estonia (both 31), Croatia (20), Romania (15) and Denmark (4). Among EFTA countries, Iceland had the highest figure, 138 car thefts per 100 000 inhabitants.

Generally, the figures include theft of motorcycles, passenger cars, buses, coaches, lorries, trucks, bulldozers, etc., but reporting and recording practices vary and affect comparison across countries and years.

Figure 6: Theft of motorized land vehicles, average 2015-2017
(police recorded offences per 100 000 inhabitants)
Source: Eurostat (crim_off_cat)

Source data for tables and graphs

Excel.jpg Source data for tables and graphs

Data sources

Data sources for statistics on crime and criminal justice systems include police and other law enforcement agencies, public prosecutors, law courts, prisons, relevant ministries, and statistical offices. The national authorities are responsible for official figures that are sent to Eurostat and to the United Nations (UN Survey on Crime Trends and Operations of Criminal Justice Systems).

This article presents results based on official figures for police-recorded offences (criminal acts) between 2008 and 2017. For for 1950-1992, summary crime statistics are available, and specified by crime for 1993-2007.

Eurostat updates the web database when countries send new figures, which may differ from figures presented in older versions of web articles such as this one.

One major problem for crime statistics at European level is that some national figures are missing. Several of the EU totals in this article were adjusted due to sporadically missing figures. For instance if a 2017 figure was missing, the figure for 2016 was used from same country for the same crime. In special cases, a missing figure is replaced by an average of the year before and after. Another method to deal with missing data is to compare three-year averages. For some crimes, simply too much data are missing. The web database contains figures as reported (no adjustment).

Concerning intentional homicide, rape, and sexual assault additional data is available on victims, suspects, prosecutions, convictions, and prisoners.

A UN website presents worldwide homicide statistics.


Crime statistics are used by EU institutions, national authorities, media, politicians, organisations, and the general public. International specifications for crime statistics cover only particular crimes. Each state establishes its criminal laws, define crimes, legal proceedings and justice reactions, as well as specifications for official crime statistics. As a consequence, crime statistics is often less comparable between states than statistics with international specifications.

For all their different criminal laws, it could also be argued that there are many similarities between European countries. This, combined with public and political interest, was the background for developing an EU-wide crime statistics. Over the last decade, EU institutions, national authorities and the UN have cooperated to improve European crime statistics. A major quality improvement is to use a common classification of crimes.

Official crime statistics mainly reflect how the authorities register and handle cases. The figures are provided by national authorities such as the police, prosecution, courts, and prisons. Of those, police figures give the broadest picture, as they include recorded offences, whether or not they led to prosecution. Still, the police records do not measure the total occurrence of crime. Simply put, the total occurrence would be the reported plus the unreported, minus the incorrectly reported. It is fair to assume that the reporting rate is high when a police record is required to support an insurance claim (e.g. car theft and burglary).

Direct access to
Other articles
Dedicated section
External links