Any summary of the findings from existing studies of sex offender recidivism must begin by acknowledging that there is a great deal of variation in the offender populations studied, the size of the sample, the definition of recidivism, the length of the follow-up period, and in the use of control or comparison groups (Furby et al., 1989; Prentky et al., 1997). Most studies of sex offender recidivism are studies of only those persons released from prison or prison-based treatment programs (Barbaree, Seto, Langston and Peacock, 200 1 ; Beech, Friendship, Erikson and Hanson, 2002; Dempster and Hart, 2002; Dobson and Konicek, 1998; Escarela, Francis and Soothill, 2000; Nunes, Firestone, Bradford, Greenbert and Broom, 2002; Prentky et al., 1997), and most typically these studies sample a mixed group of sex offenders (Barbaree et al., 2001; Dempster and Hart, 2002; DiFazio, Abracen and Looman, 2001; Dobson and Konicek, 1998; Hanson and Harris, 2000; Nunes et al., 2002). Recidivism studies of persons convicted of the same type of sexual offense, such as rapists (Prentky et al., 1997; Rice, Harris and Quinsey, 1990), child molesters (Hanson, Steffy and Gauthier, 1993; Hanson, Scott and Steffl, 1995) or, more specifically, extrafamilial child molesters (Firestone, Bradford, McCoy, Greenberg, Curry and Larose, 2000; Prentky et al., 1997; Rice, Quinsey and Harris, 1991) are less common….
Given these differences in the nature of the sample studied, the length of observation and the definition of recidivism, it is not surprising that reported rates of recidivism among sex offenders vary widely. Rates of reconviction for a new sex offense among mixed groups of sex offenders who have been observed for five years following release have been reported variously as (1) 4.3 percent of 5,098 offenders released from Ohio’s state prisons (Dobson and Konicek, 1998), (2) 9.3 percent of 321 offenders released from the Canadian federal correctional system (Dempster and Hart, 2002) and 28 percent of 178 offenders released from a Canadian maximum security psychiatric 3facility (Quinsey et al., 1995). Hanson and Bussiere (1998) reviewed 61 sex offender data bases containing information on a total of 28,972 offenders; they conclude that 36.3 percent of these offenders committed a new sexual offense and 13.4 percent were reconvicted for a new sexual offense during the first five years following release fiom prison.
Recidivism rates vary by offense type. Hanson and Bussiere’s (1998) analysis notes that the rates of reoffending and reconviction for a new sex offense among the 9,603 child molesters are 36.3 and 12.7 percent, respectively, and that the rates of reoffending and reconviction among the 1,839 rapists are 46.2 percent and 18.9 percent, respectively. Firestone and his colleagues report that whereas 15.1 percent of their sample of extrafamilial molesters committed a new sexual offense during a follow-up period that averaged 7.8 years (Firestone et al., 2000), 6.4 percent of a sample of intrafamilial molesters committed a new sexual offense during a follow-up period that averaged 6.5 years (Firestone et al., 1999). Prentky et al. (1997) studied a group of 115 extrafamilial child molesters released fiom prison and report that 14 percent committed a new sexual offense during the first five years following release and that 52 percent committed a new sexual offense over 25 years of observation. Similarly, a 19 year follow-up study of child molesters released from prison notes that half of all those reconvicted for a new sex offense were reconvicted 10 years after release (Hanson et al.,
1993).