Social risk factors
4.5 Interpreting crime rates
4.5.3 Outputs and crime rates
In this section we assess the association between several key policy output indicators and crime rates. We distinguish between three successive steps in the judicial process: the probability of being caught, the probability of being punished and the probability of being imprisoned. High probabili-ties are characteristic of a repressive system, assumed to render criminal activities less attractive, enhancing both specific and general deterrence.
The probability of being caught
The probability of being caught reduces crime rates at a country level (Figure 4.10); however, this relationship is insignificant (p<0.15) and weak.
Similar probabilities of being caught coincide with different levels of recorded crime. So the explanatory power of this probability is weak.
The probability of being punished
Arrested perpetrators are not always convicted of a crime and sentenced.
There are many reasons why a case may be dismissed. There may be a lack of evidence, a technical error may have occurred while gathering evidence,
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
0 20 40 60 80 100 120 140
offender / crime ratio
serious crime
AT BE
FR
DE NL
CH
FI SE
GR
MT ES PT
BG CY
CZ HU
PL EE
LT
SI HR
IT
a Number of suspects divided by number of recorded crimes. Source: Eurostat.
Figure 4.10 Relationship between offender/crime ratio a (proxy for the probability of being caught) and level of serious crime, 2012
the prosecutor may deem it not to be in the public interest to prosecute, or the prosecutor may prefer a conditional dismissal, as a tactic to prevent recidivism. The inevitability of actually having to undergo a punishment is expected to reduce the appeal of offending. Unfortunately, no com-plete and consistent inventory is available on the distribution of types of sanctions. Table 4.5 gives an impression of national enforcements of different sanctions. In 2010, Finland, Denmark and the United Kingdom imposed most sanctions per capita, which by and large consisted of fines.
Community service orders were especially popular in Poland, Estonia, the Netherlands and the United Kingdom. Electronic monitoring was still a rare practice in 2010, except in the United Kingdom. Bulgaria, Lithuania and Austria imposed the most prison sentences. No clear regional patterns emerge.
The ratio of the number of convictions to the number of suspects is used as a proxy for the probability of being punished. No relationship was found with the level of serious crime. At country level, therefore, a higher probability of being punished if caught seems to have no effect on rates of recorded crime.
The probability of being imprisoned
In most Western societies, imprisonment is the most severe sanction that a judge can impose. Imprisonment rates are very high in the United States (730 per 100,000 in 2010). No European country comes anywhere near this level (Tabel 4.6). In 2012, Lithuania achieved the highest level (about 300 per 100,000), closely followed by Latvia and Estonia. Northern European countries tend to show fairly low levels of imprisonment. No other region-al patterns of imprisonment emerged, region-although some Southern European countries (Cyprus, Greece) show low levels of imprisonment and most Central and Eastern European countries (except Croatia) show high levels of imprisonment.
In most countries, the level of imprisonment changed little between 1995 and 2012. Some countries show substantial increases (Cyprus, Malta, Slovenia, Croatia and Greece), but these countries started out with low lev-els of imprisonment in 1995. In some other countries the level of impris-onment decreased substantially (Romania, Lithuania, Latvia and Estonia), but these countries started out with high levels in 1995. This leads us to conclude that levels of imprisonment have converged in Europe, especially between 2000 and 2005. In terms of standard deviations, the differences in levels of imprisonment between countries decreased by 27%.
Moving on to the probability of being imprisoned, we find that a higher probability is significantly associated with a lower crime rate (Figure 4.11).
However, the association is weak. Countries with low imprisonment rates, in particular, have varying levels of serious crime. Where crime levels are high, such as in the Northern European states, the safety paradox or other
Community Electronic Unkown /
Region Country Total Fines services monitoring Imprisonment other
Western Austria Table 4.5 Types of sanctions imposed, 2010 (rates per 100,000 of the population)
Source: European Sourcebook of Crime and Criminal Justice Statistics 2014.
Region Country 1995 2000 2005 2010 2012 2012 2012 vs 1995
0 100 200 300 400 Western
Europe United Kingdom 101 +22 +17 +12 +1 153
Luxembourg 111 –19 +69 –29 –11 121
France 92 –9 +14 +9 +10 116
Belgium 74 +11 +5 +12 +2 104
Austria 78 +8 +23 –7 +2 104
Switzerland 81 –2 +4 –2 +4 85
Ireland 56 +20 0 +4 +3 83
Netherlands 70 +10 +30 –24 –5 81
Germany 75 +10 +11 –11 –5 80
Lithuania 364 –116 –15 +33 +34 300
Latvia 379 –9 –66 –3 –27 274
Estonia 305 +36 –13 –75 –8 245
Poland 171 0 +46 –5 +8 220
Czech Republic 189 +30 –33 +24 +9 219
Slovak Republic 149 –17 +40 +12 +18 202
Hungary 121 +31 +4 +8 +9 173
Romania 200 +18 –49 –38 +18 149
Bulgaria 108 +10 +29 –21 +8 134
Croatia 58 0 +20 +39 –9 108
Table 4.6 Number of prisoners in 1995 and changes in 2000, 2005, 2010 and 2012 (per 100,000 of the population)
Source: Eurostat (except non-European countries).
For reading instructions see page 49
institutional and cultural differences may be part of the explanation. We therefore need to be cautious in interpreting the relationship between imprisonment and serious crime as causal. Nonetheless, our finding sug-gests that imprisonment and crime rates are related, whether artificially due to the safety paradox, due to incapacitation, due to in-prison resociali-sation programs, or due to deterrence.
Preliminary conclusion
When the risk factors and inputs and outputs of the safety system are com-bined in one regression model, only the probability of being caught and the probability of imprisonment turn out to be significant. The share of non-Western immigrants is highly correlated with – and therefore over-shadowed by – the level of imprisonment. Together, these factors only explain one third of all differences. Figure 4.12 presents the combined association of the probability of being caught and of imprisonment on the one hand, and serious crime on the other. The share of non-Western immi-grants is not included, because there would be too many missing coun-tries. Many unexplained differences remain. In particular, the crime rates
50 100 150 200 250 300
Figure 4.11 Relationship between imprisonment ratio and level of serious crime, 2012
Source: Eurostat.
in Belgium are hardly explained by the combined arrest and imprisonment ratios, but the serious crime rates in the Netherlands and Greece are also clearly higher than expected based on their arrest and imprisonment ratios.
On the other hand, Switzerland, Germany and Slovak Republic show lower levels of serious crime than might be expected.
All in all, our findings suggest that both the probability of arrest and the probability of imprisonment may contribute to effective law enforcement.
Durlauf and Nagin (2011) argue that it is more effective to shift available resources from severity-based policies (risk and severity of punishment) to risk-based policies (risk of arrest). Scholars disagree about the deterrent effect of the severity of punishment. Some claim that the certainty of punish-ment is much more effective (Lee and McCrary 2009; Wright 2010), whereas others do find a significant effect of the duration of imprisonment (Abrams 2011; Drago et al. 2009). Impact evaluations are generally pessimistic about the contribution of incarceration to desistance. Sober and strict prison regimes have been found to foster deviance rather than combating it. This is different when the time spent behind bars is used effectively for individual rehabilitation programmes, which actually further a convict’s future pros-pects (Van Noije and Wittebrood 2008; Sapouna et al. 2011). However, this is a costly strategy that is by no means equally adopted across our selection of countries. Nonetheless, imprisonment also contributes to lower crime rates by temporarily incapacitating convicts, and potentially by deterring others in the outside world.
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2
0 20 40 60 80 100 120 140
combi offender/imprisonment ratio
serious crime
AT BE
FR
DE NL
CH
FI SE
GR
IT
MT
PT ES
BG CZ CY
HU
PL EE
LT
SI HR
Figure 4.12 Relationship between combination of offender/imprisonment ratio and level of serious crime, 2012