Police traffic enforcement works on the basis of the objective and subjective probability of detection, and the subsequent sanctions or measures. Police surveillance along the road determines the objective probability of detection or the enforcement pressure [16]. Road users assess the chance of being caught for a violation on the basis of this enforcement pressure and also on what they read in newspapers or hear from friends and acquaintances: this is the subjective probability of detection. When this probability is deemed to be sufficiently high, violations will be avoided.
In order to ensure that the subjective probability of detection is sufficiently high, it is important to have regular police checks that are unpredictable, highly visible, hard to evade, and accompanied by significant publicity. If the subjective probability of detection is sufficiently high, the preventive effect of police surveillance is generally larger as the certainty of punishment is greater, as the punishment follows the violation more rapidly, and as there is more public support for enforcement [16]. Each of these elements are links in the traffic enforcement chain.
Penalty severity only affects the preventive effect of enforcement to a limited extent. Research has shown that it is less likely to affect road user behaviour than the probability of detection is. Moreover, higher penalties will not affect drink driving, probably because the current penalties are already rather severe. Minor offences, such as speeding, not wearing a seat belt, and red light negation, do decrease somewhat as penalties are higher, but only on those roads that are frequently policed [17] (also see the question Do higher penalties work?).
Apart from a preventive instrument, police enforcement is also a selection instrument. By revoking a driving licence and by disqualifying drivers, dangerous drivers may be excluded from traffic (selection), which will enhance road safety. Repeat offenders are only a small proportion of the entire population (less than 0.5 %), yet, as drivers, they are involved in 6 % of the crashes [18] [19].
