The collected data of the Freiburg Cohort Study, which includes all police records and court decisions of six birth cohorts from Baden-Württemberg (a federal state in the south of Germany), has been analysed in relation to both the incidence and prevalence of violent sexual offences, as well as the rate of recidivism amongst sexual offenders. Furthermore, it was questioned as to how well the occurrence of violent sexual offences could be predicted from the officially registered information concerning past offences (including those offences of a violent sexual nature). Due to the longitudinal data collected in the Freiburg Cohort Study, a number of unique contributions can be made to the focal point “dangerous offenders”. This is all the more so because the data of the Freiburg Cohort Study was not pre-selected – for example, by a specific conviction – as is often the case in other studies concerning sexual violence. The data set contains all delinquent acts of the offender as far as they are known by the police. In the meantime, the data set of the study became so large, that even such rare events as rape and indecent assault could be analysed on a statistically sufficient base of cases.

Some cent­ral res­ults for rape and in­de­cent as­sault (here­after men­tioned only as rape) are presen­ted here:

It turned out that the age de­pend­ent fre­quency of rape is nearly con­stant, after a strong in­crease dur­ing early pu­berty. This could be wit­nessed for the age range start­ing at about 17 years up to an age of 30 years (as demon­strated by the rates shown in the fig­ure). This is a no­tice­able res­ult, be­cause in gen­er­al the age of de­pend­ency of de­lin­quent be­ha­viour shows a roll-back ef­fect after a max­im­um peak dur­ing late ad­oles­cence. It should be con­sidered, however, that this res­ult is based on lon­git­ud­in­al data, mean­ing that the ef­fects of in­di­vidu­al age­ing may also be in­flu­enced by pos­sible so­ci­et­al changes. Ac­tu­ally, it seems as if the nearly con­stant shape with age in the case of rape is a res­ult of an in­crease in the re­port­ing of cases of rape that oc­cur in the fam­ily or in oth­er close re­la­tion­ships. In par­tic­u­lar, dur­ing the 1990s there was a marked in­crease in the num­ber of re­por­ted rape cases where the of­fend­er and the vic­tim had a close re­la­tion­ship, while the num­ber of cases in which the vic­tim did not know the of­fend­er re­mained con­stant.

Fig­ure: an­nu­al po­lice rates and cu­mu­lated rates of Gre­man males, rape and in­de­cent as­sault (gen­er­al de­lin­quency) Baden-Württem­berg

A great ad­vant­age of the Freiburg Co­hort Study is that the per­sons in prin­ciple are fol­lowed over their life course. So it is pos­sible, even for cases in­volving rape, to dis­tin­guish between a first of­fence and any sub­sequent fol­low­ing of­fences. It turns out that nearly every case of rape is a first time re­gis­tra­tion (over 80%, see the rate of first time oc­cur­rences of rape in the fig­ure in com­par­is­on to all re­gis­tra­tions of rape). This im­plies both that nearly all people re­gistered for rape have no spe­cif­ic his­tory in this area of de­lin­quency, and, as fol­lows, re­cidiv­ism has a low rate of oc­cur­ring in the case of sexu­al vi­ol­ence. From the data of the Freiburg Co­hort Study the es­tim­ate for re­cidiv­ism is be­low 20% over the fol­low­ing 10 years.

Over the course of life up to age 30, about 0.4% of Ger­man males are re­gistered at least (and mostly only) once for rape (see the cu­mu­lated rates in the fig­ure). The group size of re­gistered rape of­fend­ers in­creased in a nearly lin­ear fash­ion with age. This cor­res­ponds to the con­stant (flat) age de­pend­ency of the (first time) re­gis­tra­tion rates men­tioned above. Moreover, this is in con­trast to the typ­ic­al age de­pend­ency of de­lin­quency, which is also shown in the fig­ures (gen­er­al de­lin­quency). As also men­tioned above, this is prob­ably caused by an in­crease of re­ports be­ing made.

If the afore­men­tioned facts would lead to the im­pres­sion that rape is mostly a sin­gu­lar event, this is only true for this spe­cif­ic kind of de­lin­quency, and not for oth­er de­lin­quent acts of these sexu­al of­fend­ers. In gen­er­al, throughout the life course of sexu­al of­fend­ers there are oth­er de­lin­quent acts be­fore or after the case of rape. From all the Ger­man males which were re­gistered (mostly once) for rape oc­cur­ring be­fore and up to age 27, about 16% could be said to be life course per­sist­ent of­fend­ers. On the oth­er hand, a group of about 14% had no ad­di­tion­al re­gis­tra­tions at all. The ma­jor­ity of these sexu­al of­fend­ers (70%) fall between these ex­tremes.

Fur­ther­more, it could be demon­strated that the data re­gistered by the po­lice or the courts would not be suf­fi­cient for pro­gnost­ic pur­poses. This re­mains true even if the dif­fer­ent kind of de­lin­quent acts are taken in­to ac­count sep­ar­ately. The best es­tim­ate of a mul­tivari­ate ana­lys­is had been an ex­plan­a­tion of about 8% of the vari­ance (R² = 0.08).