Data extracted in July 2020.
Planned article update: August 2021.
One prisoner for every 896 people in the EU-27 in 2018
There was one prisoner for every 896 people in the EU-27 in 2018, or 111 prisoners per 100 000 people. This was the lowest number of prisoners per 100 000 people (prisoner rate) since the turn of the century.
The highest prisoner rate in the EU-27 (average 2016-2018) was in Lithuania (232.8), Czechia (208.6), Estonia (206.8), Latvia (196.9), Poland (192.7), Slovakia (185.9), and Hungary (174.5). The lowest rates were in Slovenia (65.5), the Netherlands (63.9), Denmark (60.7), Sweden (58.3), and Finland (55.7). Figure 1 presents data for all countries that provide prison statistics to Eurostat.
Trends in number of prisoners
There were 495 000 prisoners in the EU-27 in 2018, which is 10 % less than in 2012, when there were around 553 000 prisoners, the highest number since 1993. The number of prisoners in the EU-27 rose by 24 % from 1993 to 2012. In that same period, the population of the EU-27 grew by just 4.4 %. This resulted in the prisoner rate increasing from 106 in 1993 to 126 in 2012. From 2012 to 2016, there was a downward trend in the number of prisoners, resulting in an overall reduction of 10.4 %. The data for the last two years (2017 and 2018) show an apparent stability. Due to missing data for these years, the latest results may be less accurate.
A net increase in the number of prisoners is mainly due to more new prison sentences than the number of releases. One of the major causes of the increase in prisoners from 1993 to 2012 was a preceding rise in serious crimes across Europe. After an increase in serious crimes followed by more convictions with longer sentences, the increase in the number of prisoners may remain on a higher level for a while, even after the crime rate drops. This illustrates the indirect and sometimes complicated relationship between crime trends and prisoner trends.
1 out of 20 prisoners are women
Between 2008 and 2018, 1 in 20 adult prisoners in the EU-27 were women. As illustrated in Figure 3, the proportion has increased slightly in recent years. The average percentage of adult women prisoners was 5.0 % from 2008 to 2013 and 5.2 % from 2014 to 2018.
The share of women in prison varies between EU-27 Member States (see Figure 4). Between 2016 and 2018, the highest shares on average were observed in Latvia (8.0 %), Spain, Finland (both 7.5 %), Czechia, Hungary (both 7.4 %), and Slovakia 7.2 %). The lowest shares were in Italy (4.2 %), Denmark (4.0 %), Poland (3.9 %), France (3.5 %), Bulgaria (3.2 %), and Ireland (3.1 %). For Belgium and Germany data are missing for all three years.
Two prisoners per prison employee in the EU-27
From 2008 to 2018, the average number of prisoners per prison employee has been around 2 in the EU-27. The actual number of prisoners per employee in charge can of course vary. Typically, the number of prisoners varies more over time than the number of prison employees, as it takes some time to adjust staff size to the prisoner population. As illustrated in Figure 5, the number of prisoners and prison employees fell between 2012 and 2018, while the ratio of prisoners per prison employee remained relatively stable.
Overcrowding and empty cells
The occupancy rate for a prison is the number of prisoners relative to the official capacity (design capacity) of that prison, multiplied by 100. Overcrowding can be formally defined as any occupancy rate exceeding 100. However, practical adaptations in the prison can make the safe operational capacity higher than the official capacity. On the one hand, a minor or temporary excess does not necessarily indicate an overcrowding problem. On the other hand, measuring overcrowding in a country with several different prisons may hide local overcrowding.
The highest overcrowding rates  were observed in Austria (136), Hungary (121), France (117) and Italy (115), and the lowest in Estonia, Croatia (both 89), Germany (88), Malta (85), Poland, Bulgaria and Lithuania (each 83). For Belgium, Ireland, and the Netherlands data are missing for all three years. As shown in Figure 6, thirteen countries had some extra capacity, or "empty cells". The figures are the 2016-2018 average.
Typically, prisoner numbers vary more than capacity numbers. Prison capacity cannot be changed rapidly, since it takes time to plan and construct secure buildings. Official prison capacity changes when, for instance, a new prison building is finished, adaptations are made to an existing prison, or a prison is decommissioned. Prisoner number depends on other factors, such as conviction rates and lengths of sentences.
Figures 7 and 8 illustrate prison occupancy trends, showing four countries that have reported complete data series 2008-2018.
In Sweden, a decrease in the number of prisoners began in 2009, followed by a capacity reduction that started in 2011. From 2013 to 2018, the number of prisoners in Sweden was relatively stable, and the capacity was gradually reduced, while keeping the rate below 100.
In Portugal, there was a relatively large increase in the number of prisoners between 2008 and 2013, while the capacity remained much the same. Thus, the prison occupancy went from having some extra capacity to being relatively crowded at national level. From 2013, the number of prisoners started falling. This combined with some increase in capacity, improved the situation and in 2018 the occupancy rate was 101.
In Italy, there was a steady increase in prison capacity between 2008 and 2018, while the number of prisoners varied much more. Italy had its highest prison occupancy in 2010 and lowest in 2015.
In Poland, there was a steady increase in capacity between 2008 and 2012, while the number of prisoners generally decreased. From 2013, prison occupancy fell considerably due to falling number of prisoners, while the capacity was maintained.
Government expenditure on prisons
Government expenditure for public order and safety in the EU was 1.7 % of the gross domestic product in 2018 of which 0.9 % was for police, 0.3 % for law courts and 0.2 % for prisons . The remaining government expenditure was for fire-protection and other services.
Source data for tables and graphs
Data sources for prison statistics are national authorities such as prison administration or the central statistical office. Eurostat collects prison data together with the yearly UN global crime statistics. All results presented in this article are based on official figures.
Missing data is a major problem for some types of crime statistics at European level. In this article, some EU totals are presented that are not found in the Eurostat database due to quality considerations. For example, if a 2018 figure was missing, the figure for 2017 (from the same country) was included in the total instead. On occasion, similar adjustments are made for more years. This is not to be considered official EU-27 totals, but may serve to indicate an overall trend. The national figures are presented in the web database as reported (no adjustments). An example may illustrate the reporting. For the EU-27 time series 1993-2018, in all 18 figures were missing for various countries and years, of a total of 26*27=702 figures.
Eurostat updates the web database when countries send new figures, but older statistical articles may refer to previously reported figures.
In general, each country decides criminal laws, criminal justice policies, as well as specifications for relevant statistics. Typically, official statistics on crime and criminal justice is made for administrative purposes and used by national authorities.
Prison statistics is part of official statistics on crime and criminal justice. It relates indirectly to crime occurrence and the performance of the criminal justice system.
The number of prisoners (counted at the end of a year) depends on:
- the number in previous year
- how many were convicted and actually imprisoned during the year
- how many left prison during the year: released, pardoned, died, escaped, etc.
The trend in prisoner number depends on:
- how many crimes are brought to court
- how many are convicted, and actually sentenced to prison
- length of sentences received, and the actual duration
- alternative sanctions, early release, amnesties, etc.
- law changes, for which types of crime imprisonment is prescribed, and for how long
- Crime and criminal justice (ESMS metadata file — crim_esms)
- Data for Belgium missing for all three years.
- Source: Eurostat table "crim_gen". Crimes recorded by the police (sum of acts causing harm or intending to cause harm to the person, injurious acts of a sexual nature and acts against property involving violence or threat against a person).
- Source: ‘Handbook on strategies to reduce overcrowding in prisons’, United Nations Office on Drugs and Crime, 2013
- 100 means that the number of prisoners is equal to the official prison capacity