Loading...
HomeMy WebLinkAboutCC Ex D Sex Offender Duwe Reporthttp://cjb.sagepub.com Criminal Justice and Behavior DOI: 10.1177/0093854807313690 2008; 35; 484 Criminal Justice and Behavior Grant Duwe, William Donnay and Richard Tewksbury Recidivism Does Residential Proximity Matter? A Geographic Analysis of Sex Offense http://cjb.sagepub.com/cgi/content/abstract/35/4/484 The online version of this article can be found at: Published by: http://www.sagepublications.com On behalf of: International Association for Correctional and Forensic Psychology can be found at:Criminal Justice and Behavior Additional services and information for http://cjb.sagepub.com/cgi/alerts Email Alerts: http://cjb.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.navReprints: http://www.sagepub.com/journalsPermissions.navPermissions: http://cjb.sagepub.com/cgi/content/refs/35/4/484 Citations at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from 484 DOES RESIDENTIAL PROXIMITY MATTER? A Geographic Analysis of Sex Offense Recidivism GRANT DUWE WILLIAM DONNAY Minnesota Department of Corrections RICHARD TEWKSBURY University of Louisville In an effort to reduce sex offense recidivism, local and state governments have recently passed legislation prohibiting sex offenders from living within a certain distance (500 to 2,500 feet) of child congregation locations such as schools, parks, and daycare centers. Examining the potential deterrent effects of a residency restrictions law in Minnesota, this study analyzed the offense patterns of every sex offender released from Minnesota correctional facilities between 1990 and 2002 who was reincarcerated for a new sex offense prior to 2006. Given that not one of the 224 sex offenses would have likely been prevented by residency restrictions, the findings from this study provide little support for the notion that such restrictions would signifi- cantly reduce sexual recidivism. Keywords:sex offenders; recidivism; deterrence; offense patterns; residency requirements In light of the perception that sex offenders pose a major threat to their communities because they are highly incorrigible, local and state governments have recently enacted policies that restrict where sex offenders are allowed to live. For example, 22 states have passed legislation that may prohibit sex offenders from living near schools, daycare centers, parks, and other areas where potential vulnerable victims may be present (Nieto & Jung, 2006). In some states, local governments have passed ordinances restricting the placement of sex offenders. In Minnesota, for example, local governments in Taylors Falls and Wyoming have both passed such ordinances. Designed to enhance the safety of children, residency restrictions are targeted mainly toward child molesters, who often gain access to their victims through marriage, occupation, or the neighborhood in which they live (Walker, Golden, & VanHouten, 2001). Because resi- dency restrictions are intended to prevent child molesters from making direct contact with children, they are primarily applicable with the third type of access—neighborhood (Walker et al., 2001). But are such policy measures consistent with the reality of sexual recidivism? That is, are sex offenders highly likely to recidivate? And, when they do reoffend, are they likely to do so by selecting victims who reside in close proximity to where they live? CRIMINAL JUSTICE AND BEHAVIOR, Vol. 35 No. 4, April 2008 484-504 DOI: 10.1177/0093854807313690 © 2008 International Association for Correctional and Forensic Psychology AUTHORS’NOTE:The views expressed in this study are not necessarily those of the Minnesota Department of Corrections. The authors wish to thank Karl Hanson and Michael Miner for their helpful comments on an earlier version of this article and Nicole Hansen from the Minnesota Department of Corrections for her work in collecting and entering the data. Correspondence concerning this article should be addressed to Grant Duwe, Minnesota Department of Corrections, 1450 Energy Park Drive, Suite 200, St. Paul, MN 55108; e-mail: GDuwe@co.doc.state.mn.us. at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from LITERATURE REVIEW SEX OFFENDER RECIDIVISM Prior research has clearly indicated that sex offenders are, compared to other offenders, among the least likely to reoffend (Harris & Hanson, 2004; Langan & Levin, 2002; Sample & Bray, 2006). Moreover, when sex offenders recidivate, they are much more likely to do so with a nonsexual offense. Examining recidivism among 9,691 sex offenders released from prison in 1994, Langan, Schmitt, and DuRose (2003) found that only 12% of the rear- rests in the 3-year postrelease period involved a sex offense. However, despite the fact that sex offenders are among the least likely to recidivate in general, they are still, compared to other offenders, more likely to reoffend sexually (Langan & Levin, 2002). When sex offenders recidivate with a sex offense, at least 75% victimize individuals (both adults and children) whom they already know (Greenfield, 1997; Snyder, 2000). A number of factors influence recidivism, and the same factors are not equally influential on all varieties of sex offenders (Levenson & D’Amora, 2007). Existing research has demon- strated, for example, that the risk of sexual recidivism is significantly greater for offenders who have an antisocial orientation (i.e., history of rule violation), deviant sexual interests, a history of victimizing strangers, conflicts in intimate relationships, an emotional identifi- cation with children, and prior noncontact sex offenses (Hanson & Morton-Bourgon, 2004). Conversely, the risk of sexual recidivism is less for incest offenders, first-time sex offenders, those older than the age of 50, and those who target female children rather than male children (Harris & Hanson, 2004). Consequently, such research has led to the conclu- sion that “not all sex offenders should be treated the same” (Harris & Hanson, 2004, p. 1). However, when considering sex offender registration and accompanying residency restric- tions, all are treated the same. RESIDENCY AND OFFENSE CHARACTERISTICS OF SEX OFFENDERS Recent research has shown that sex offenders—at least those who have been previously identified, convicted, and placed on sex offender registries—are likely to live in certain types of locations. Specifically, registered sex offenders are especially likely to live in neighborhoods that have high levels of social disorganization, greater proportions of youth, and lower proportions of high school and college graduates, more minorities, fewer owner- occupied homes, and lower housing values and household incomes (Mustaine, Tewksbury, & Stengel, 2006a, 2006b, in press); these factors are especially salient for African American sex offenders. Most also live in single family homes but do not live in neighborhoods with a posted neighborhood watch (Tewksbury & Mustaine, 2006). Other research, (Stengel, Tewksbury, & Mustaine, in press) has shown that sex offenses are not associated with most measures of proximity to pools of available suitable targets of victims. Although sex offenses are more likely in census tracts with larger proportions of children younger than age 10 and more daycare centers, the presence of schools, youth with disabilities, and women living alone is not associated with a greater number of sex offenses (Stengel et al., in press). Although Walker et al. (2001) reported that registered sex offenders in one Arkansas county may be likely to live near schools, daycare centers, and parks, other research has suggested that sex offenders are highly unlikely to reoffend close to their homes (Colorado Department of Public Safety, 2004; Minnesota Department of Corrections, Duwe et al. / RESIDENTIAL PROXIMITY & SEX OFFENSE RECIDIVISM 485 at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from 2003). In addition, a higher concentration of registered sex offenders residing in a neigh- borhood has been shown to have no statistical relationship with the number of sex offenses in a community (Stengel et al., in press), and sex offenders who recidivate are no more likely than nonrecidivating sex offenders to live near schools or day care centers (Colorado Department of Public Safety, 2004). A slightly different conclusion, however, is reached by the one study to assess where sex offenders known to have offended against child victims reside (Tewksbury & Mustaine, in press–a). Child-victimizing sex offenders tend to live in less socially disorganized neigh- borhoods than do other sex offenders. Research has shown that such offenders, when com- pared with communities in general, tend to live in neighborhoods with greater proportions of youth in the population, more minority residents, more residents with high school and college educations, more female-headed households, and higher household incomes and housing values. Scholars generally believe that the residential locations of registered sex offenders are caused by a process of social and economic relegation, facilitated by low levels of social cohesion and social capital in communities where known sex offenders do reside. Such com- munity conditions leave these community members with few (if any) resources to discourage sex offenders from moving into such communities. Most known sex offenders do change residences, often frequently (Mustaine et al., 2006b; Turley & Hutzel, 2001). When they move, they typically go to a more socially disorganized neighborhood than where they lived previously (Mustaine et al., 2006b). Overall, economic factors, not proximity to pools of available, potential victims are what appear to drive where registered sex offenders reside (Tewksbury & Lees, 2006; Tewksbury & Mustaine, in press–b). THE IMPACT OF REGISTRATION AND RESIDENCY RESTRICTIONS ON OFFENDERS Other research has examined the impact of registration on offenders themselves, includ- ing a focus on where registrants can, and do, live. As Mustaine et al. (2006a) point out, sex offenders are frequently relegated to neighborhoods and communities marked by social dis- organization and economic deprivation. Furthermore, residency restrictions often force offenders to move from their residences. Levenson and Cotter (2005) surveyed 135 sex offenders in Florida who were subject to residency restrictions that prohibited them from living within 1,000 feet of a school, daycare center, park, playground, or other place where children regularly congregate. Levenson and Cotter found that 50% of the 135 offenders, of whom 97% were child molesters, reported being forced to move on account of the 1,000-foot rule. In addition, the results indicated that the housing restrictions also led to increased isola- tion, decreased stability, and greater emotional and financial stress. In addition, in jurisdictions with residency restriction laws, registered sex offenders may have extremely limited options regarding where they may legally live (Tewksbury, 2007; Zandbergen & Hart, 2006). Using mapping technology, Zandbergen and Hart (2006) found that less than one quarter of all housing in Orange County, Florida is not within a restricted zone. If school bus stops were added to the list of locations included within a restriction zone, only 4% of all housing in the county would be available to registered sex offenders. When residential restriction laws are in place, there are numerous accompanying negative consequences and an intensification of collateral consequences of sex offender registration in general (Tewksbury, 2007). 486 CRIMINAL JUSTICE AND BEHAVIOR at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from One of the clearest findings of this body of research is that residency restrictions and, more generally, sex offender registration create a hardship for sex offenders. However,despite both registration and residential restrictions being focused on promoting community safety, the impact of such laws on sexual recidivism remains unclear. Given that the residential proximity issue has not been fully addressed empirically, there may be truth to the notion that when sex offenders reoffend sexually, they are likely to do so very close to where they live. If so, then policies restricting where sex offenders live may have merit in terms of enhancing public safety. THE JOURNEY TO CRIME A growing literature addressing where criminal offenders commit their offenses has consistently shown that the number of criminal offenses that an offender commits decreases as distance from an offender’s residence increases (Brantingham & Brantingham, 1984; Rengert, 2004; Rengert, Piquero, & Jones, 1999). However, for violent offenders (including sex offenders), this pattern typically does not hold. Confrontational offenders—who actually encounter their victims personally—seek offending locations where they are unlikely to be recognized (and therefore apprehended). Data from 565 rapes committed by serial rapists showed that offenses occurred an average of more than 3 miles from offenders’ homes (Warren et al., 1998). Also focusing on serial rapists, Canter and Gregory (1994) found that 86% of the offenders they studied did not offend against victims who lived nearby, but instead they “marauded” outward into an area of an average of 180 square miles. In New Zealand, serial sexual assault offenders committed their offenses an average of 3 kilometers (1.86 miles) away from their residences (Lundrigan & Czarnomski, 2006). For other types of violent crime, Groff and McEwen (2006) reported that homicide offenders committed the offense, on average, 0.69 miles from their homes. In addition, Tita and Griffiths (2005) showed that across 9 years of homicides in Pittsburgh, homicide offend- ers rarely killed in their own neighborhoods. Also, they found that characteristics of events and relationships were more important than characteristics of victims and opportunities in determining where offenses took place. THE PRESENT STUDY: RESIDENCY RESTRICTIONS IN MINNESOTA In a 2006 report, Nieto and Jung identified 22 states (Alabama, Arkansas, California, Florida, Georgia, Illinois, Indiana, Iowa, Kentucky, Louisiana, Michigan, Missouri,Minnesota, New Mexico, Ohio, Oklahoma, Oregon, South Dakota, Tennessee, Texas, Washington, and West Virginia) that have residency restriction legislation for sex offenders. Of these, how- ever, 4 (New Mexico, Oregon, Texas, and Minnesota) do not have statutory language specifi- cally prohibiting sex offenders from living within a certain distance of a child congregation location. In New Mexico, schools must be notified of sex offenders living within a 1-mile radius, but the law does not restrict where they can live. In Oregon and Texas, a government body (e.g., Department of Corrections or the Parole Board) is responsible for determining where and how close a sex offender can live to a child congregation location. In Minnesota, Nieto and Jung (2006, p. 21) reported that “the Parole Commissioner determines if and where a level II and III sex offender may reside within 1,500 feet of school zones.” Nieto and Jung’s (2006) description of Minnesota’s residency restriction statute is flawed, however. First, the actual statute does not mention a Parole Commissioner nor does the State Duwe et al. / RESIDENTIAL PROXIMITY & SEX OFFENSE RECIDIVISM 487 at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from of Minnesota actually have one. Second, and more important, there is no mention in the statute of either Level II offenders or restrictions about living within 1,500 feet of school zones. Instead, the statute, which was enacted in 1999, indicates that for offenders given a Level III assignment, the End of Confinement Review Committee (ECRC) “shall determine whether residency restrictions shall be included in the conditions of the offender’s release based on the offender’s pattern of offending behavior” (Minnesota Statutes 2006, section 244.052, subdivision 3, paragraph [k]). Of the 224 sex offender recidivists examined in this study, there were 22 who were released on or after July 1, 1999—the inception date for this statute. Of the 22, only 2 were eligible for this provision in that they were given a Level III assignment. The ECRC requested residency restrictions for 1 of the 2 offenders; however, on further review of this particular offender’s release plan, residency restrictions were not included in his conditions for release. As a result, not one of the offenders analyzed in this study was subject to residency restrictions on release from prison.1 The sex offenders examined here thus provide an ideal opportunity to assess the likelihood of whether residency restrictions might have been effective in preventing the occurrence of their sex reoffenses. This is accomplished by identifying whether these sexual recidivists did, in fact, target children who lived in close proximity to their own residences. In closely analyzing the spatial patterns of sexual reoffending among the 224 recidivists released from Minnesota correctional facilities between 1990 and 2002, this study focuses on several key questions. First, where did offenders initially establish contact with their victims, and where did they commit the offense? Second, what were the physical distances between an offender’s residence and both the offense and first contact locations? Finally, were other factors such as victim–offender relationship, supervision status, use of alcohol or drugs, and use of force associated with both residential proximity and the sexual reoffense? METHOD To address these questions, data were gathered on sex offenders who recidivated with a new sex offense. Between 1990 and 2002, there were a total of 3,166 sex offenders released from eight state correctional facilities in Minnesota.2 Of these offenders, 374 were rear- rested for a new sex offense, 304 were reconvicted for a new sex crime, and 224 were rein- carcerated for a new sex offense prior to January 1, 2006 (Minnesota Department of Corrections, 2007). As a result, there were 70 offenders who were rearrested, but were not reconvicted or reincarcerated, for a new sex offense, and 80 offenders who were reconvicted of a new sex crime but were not reincarcerated. Because of the greater availability of data on the offenders who returned to prison for a new sex crime, we did not examine the 70 rearrests and 80 reconvictions that did not result in reincarceration. Instead, we focused our analyses on the 224 sex offenders who were reincarcerated for a new sex offense. As a result, our sample represents what are likely the most serious cases of sexual recidivism that occurred in Minnesota between 1990 and 2005. MEASURES Several different sources of data were used to examine proximity: the criminal complaint for the sexual reoffense, the pre-sentence investigation (PSI) report, the Statewide 488 CRIMINAL JUSTICE AND BEHAVIOR at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from Duwe et al. / RESIDENTIAL PROXIMITY & SEX OFFENSE RECIDIVISM 489 Supervision System (SSS), the Minnesota Bureau of Criminal Apprehension (BCA) offender registry, and the Correctional Operations Management System (COMS), the data- base maintained by the Minnesota Department of Corrections. These sources were reviewed for each of the 224 recidivists, and data were recorded for each of the following items: 3 offender’s address at the time of the reoffense, the address of the location where the new offense occurred, the location or address where the offender first established contact with the victim, the type of location where the offense took place (e.g., offender’s residence, victim’s residence, public building, etc.), the type of location where the offender first estab- lished contact with the victim (e.g., offender’s residence, victim’s residence, bar/nightclub, etc.), the relationship between the offender and victim (e.g., stranger, girlfriend’s daughter, babysitter, etc.), the amount of force used (e.g., no force, force with injury, etc.), and the presence of alcohol and/or drug use by the offender and/or the victim around the time of the offense. Of the 22 states that have enacted residency restriction legislation, most have applied restrictions for all sex offenders regardless of their offending history or perceived risk of reoffense (Nieto & Jung, 2006). It is possible, however, that residency restrictions may be a more effective deterrent for some offenders than for others. In an effort to identify the types of sex offenders for whom residency restrictions might be more effective, we collected and analyzed additional offender and victim data from the COMS database. The appendix lists the variables derived from COMS and describes how they were created. ANALYSIS Once all data were collected, the physical distances between the offender’s residence and both the offense and first contact locations were calculated, using Google Earth. For example, using the “Directions” feature, the offender’s address was entered in the first address loca- tion (i.e., “From”), whereas the offense or first contact location was entered in the second address location (i.e., “To”). The “Ruler” feature in Google Earth was then used to deter- mine the Euclidian, or straight-line, distance (in both feet and miles) between the first and second address locations. Four criteria were used to determine whether residency restrictions might have prevented a sex crime from occurring. As noted above, housing restrictions are geared primarily toward deterring sex offenders—namely, child molesters—from initiating contact with potential victims by prohibiting them from living within a certain distance of a school, park, daycare center, or other area where children might be present. The first criterion, then, con- cerns the means by which the offenders established contact with their victims. Therefore, our analyses focused on direct-contact offenders, who typically initiated contact with their victims by approaching them on the street, meeting them in a bar, or breaking into the victim’s home. In addition, our analyses also assessed offenders who gained access through indirect means (e.g., girlfriend’s daughter, babysitter, friend’s son or daughter, etc.) to provide a more complete picture of the patterns of sexual reoffending. The second criterion concerns the distance between an offender’s residence and where he (all 224 offenders were male) first established contact with the victim. There is no clear consensus on the distance requirement across jurisdictions that have implemented housing restrictions; statutes range from 500 to 2,500 feet. The distances in most states, however, are often between 1,000 and 2,500 feet (Levenson & Cotter, 2005). This study, therefore, determined residential proximity on the basis of a 1,000-foot zone (0.2 miles) as well as a at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from 490 CRIMINAL JUSTICE AND BEHAVIOR 2,500-foot zone (0.5 miles). To ensure that neither distance was overly restrictive, residen- tial proximity was also determined on the basis of a 1-mile zone (5,280 feet). As a result, three distances were used in this study to determine residential proximity: 1,000 feet, 2,500 feet, and 5,280 feet (1 mile). The third criterion concerns the type of location where the offender established contact with the victim. In order for a case to be considered one that might have been prevented by a residency restrictions law, the offender had to have established victim contact in or near one of the prohibited areas: a school, park, playground, daycare center, or other location where children are known to congregate. The fourth criterion concerns the age of the victim. Because housing restrictions focus on the protection of children, the victim(s) had to have been younger than age 18 at the time of the offense for it to be considered a case where a residency restriction law might have made a difference. All four of the criteria outlined above had to be met for an offense to be classified as one that might have been prevented by housing restrictions. If, for example, an offender estab- lished direct contact with a juvenile victim 0.3 miles (1,584 feet) away from his residence at a park and committed the offense in the same location, residential proximity would be rele- vant for both the 0.5 mile (2,500 feet) and 1.0 mile distances but not for the 0.2 mile (1,000 feet) distance. If, however, the victim in this situation were an adult, then the case would not be classified as one that might have been deterred by residency restrictions. Similarly, if an offender broke into a neighbor’s home 0.1 miles (500 feet) away from his own residence and victimized a juvenile female victim, the case would not meet the criteria for classification because the first contact location would be the victim’s home rather than a child congregation location. RESULTS The results show that 85% (n =190) of the sexual offenders’ reoffenses occurred in a residential location (see Table 1). The other 15% (n =34) took place in a public location, of which most were an exterior location such as a street, alley, or park. Slightly more than one half (54%) of the recidivists committed the reoffense in their own residence. Of these 121 reoffenses, the offender shared the residence with the victim in 37% of the cases. In 42% (n =94) of the cases, the offense took place within the victim’s home. The victim shared the residence with the offender in 48% (n =45) of these cases. For 27 of the 224 cases, it was not possible to estimate the distance between the offender’s residence and the location where the offense took place because of the absence of specific address information for at least one of the locations. This was especially true for the older cases (i.e., those that took place in the early to mid-1990s) in that specific address information for either the offender’s residence or the offense location was less likely to be available in any of the data sources used. OFFENDER RESIDENCE–OFFENSE LOCATION DISTANCE The results in Table 1, which also display the straight-line distance between the offender’s residence and the offense location, suggest that offenders were more likely to commit offenses in or near their place of residence. Of the 197 cases where specific address at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from Duwe et al. / RESIDENTIAL PROXIMITY & SEX OFFENSE RECIDIVISM 491 information was available,61% (n =121) took place inside the offender’s residence. For the remaining 39% (n =76) who committed the offense outside of their residence, there was an inverse relationship between the offender residence–offense location distance and the number of offenses. That is, as the distance between the offender’s residence and the offense location increased, the number of offenses decreased. For example, 19% (n =38) of the offenses took place within 5 miles of the offender’s residence, 7% (n =14) within 6 to 10 miles,6% (n =12) within 11 to 20 miles, 4% (n =8) within 21 to 50 miles, and 2% (n =4) beyond 50 miles. However, relatively few of the offenses (9%) took place within 1 mile of the offender’s residence. Compared to the offender residence–offense location distance, estimating the distance between the offender’s residence and the first contact location was more difficult for sev- eral reasons. First, the address information regarding the specific location where offenders first established contact with their victims was frequently unavailable in the criminal com- plaint. Second, for some cases, geographic distance was irrelevant in that several offenders first established contact over the telephone (one even while incarcerated for the prior sex offense) or the Internet (i.e., dating personals). Finally, and perhaps most important, the majority of the offenders knew their victims at the time of the offense, often for some time before the crime took place. For example, determining the specific location where an TABLE 1: Location, Distance, and Victim–Offender Relationship of Sex Reoffenses Number % Location of offense Offender’s residence 76 33.9 Victim’s residence 49 21.9 Shared residence 45 20.1 Residence of acquaintance/family member 15 6.7 Other residence (e.g., hotel room) 6 2.7 Exterior public location 28 12.5 Interior public location 6 2.2 Total 224 100.0 Distance between offender’s residence and offense location Same offender residence/offense location 76 38.6 Same victim–offender residence/offense location 45 22.8 Less than 1 mile 18 9.1 1–5 miles 20 10.2 6–10 miles 14 7.1 11–20 miles 12 6.1 21–50 miles 8 4.1 More than 50 miles 4 2.0 Total 197 100.0 Victim–offender Relationship Stranger 48 21.4 Acquaintance/other known 51 22.8 Babysitter 13 5.8 Neighbor 8 3.6 “Romantic/dating” 13 5.8 Friend of family 20 8.9 Significant other’s son or daughter 39 17.4 Family/biological 32 14.3 Total 224 100.0 at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from offender first met his stepdaughter (his victim) is largely irrelevant to the issue of residential proximity. Consequently, the findings regarding the offender residence–first contact distance will focus only on the direct-contact offenders and will be presented later in this study. VICTIM–OFFENDER RELATIONSHIP The findings regarding the victim–offender relationship for the 224 reoffenses indicate that 21% of the offenders victimized someone they did not know (see Table 1). This percentage is higher than that normally seen in sex offender populations because this is a sample of recidivists, who are more likely to victimize strangers. Consistent with research on sex offenders in general, the vast majority (79%) of offenders, however, victimized someone they knew. Acquaintance/other known was the most common victim–offender relationship (23%), followed closely by offenders who victimized the daughter or son of the woman with whom they had developed a romantic relationship (17%). This category includes men who molested their stepdaughters or stepsons. In 14% (n =32) of the cases, offenders victimized family members such as their own daughter, niece, or granddaughter. In Table 2, we examine whether the offender residence–offense location distance varied by the type of victim–offender relationship. When the offenders victimized a stranger, 28% (n =12) committed the offense in their own residence. When they committed the offense outside their residence, however, most did so more than 1 mile away from their home; fully 49% (n =21) of the stranger-on-stranger reoffenses took place more than 1 mile from the offender’s residence. In contrast, 23% (n =10) occurred within 1 mile of the offender’s residence. Of the 17 offenses that took place within 1 mile of the offender’s residence, 10 involved strangers, 5 involved neighbors, 1 involved a babysitter, and 1 involved a “consen- sual” romantic relationship. There were 124 cases (63%) that occurred within the offender’s residence, of which 90% (n =112) involved offenders who knew their victims. The percentage of cases occurring inside the offender’s residence was greater for those who victimized their significant other’s son or daughter (89%), a biological family member (81%), a child they were babysitting (77%), or an acquaintance (74%). In Table 3, we examine more closely the relationship between the offender’s residence, the victim’s residence, and the location where the offense occurred. There were five possible offender residence, victim residence, and offense location combinations, as follows: 1. The offense occurred at the residence where both the offender and victim lived. 2. The offense took place at the offender’s residence, which was different from that of the victim. 3. The offense took place at the victim’s residence, which was different from that of the offender. 4. The offender and victim shared a residence, but the offense occurred at a different location. 5. The offender residence, victim residence, and offense locations were all different from one another. The results from a chi-square significance test showed that the 45 offenses occurring in the residence shared by both the victim and the offender were significantly less likely to involve the use of physical force and alcohol or drugs. These offenses were more likely to involve offenders who victimized children and family members, however. The 76 offenses that took place at the residence where only the offender lived were significantly more likely to involve the use of alcohol or drugs and acquaintance victims. The 49 offenses that 492 CRIMINAL JUSTICE AND BEHAVIOR at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from 493 TA B L E 2 : O f f e n s e L o c a t i o n – O f f e n d e r R e s i d e n c e D i s t a n c e b y V i c t i m – O f f e n d e r R e l a t i o n s h i p Ac q u a i n t a n c e / F a m i l y S i g n i f i c a n t O t h e r Di s t a n c e S t r a n g e r O t h e r K n o w n B a b y s i t t e r N e i g h b o r R o m a n t i c F r i e n d O t h e r ’ s C h i l d F a m i l y T o t a l Of f e n d e r ’ s r e s i d e n c e 2 7 . 9 7 3 . 8 7 6 . 9 3 7 . 5 4 1 . 7 5 8 . 8 8 8 . 8 8 0 . 8 6 2 . 9 < 1 m i l e 2 3 . 3 0 . 0 7 . 7 6 2 . 5 8 . 3 0 . 0 0 . 0 0 . 0 8 . 6 1– 5 m i l e s 1 8 . 6 7 . 1 0 . 0 0 . 0 1 6 . 7 2 3 . 5 2 . 8 3 . 8 9 . 6 6– 1 0 m i l e s 1 1 . 6 2 . 4 1 5 . 4 0 . 0 1 6 . 7 0 . 0 5 . 6 7 . 7 7 . 1 11 – 2 0 m i l e s 1 4 . 0 7 . 1 0 . 0 0 . 0 8 . 3 5 . 9 2 . 8 0 . 0 6 . 1 21 – 5 0 m i l e s 2 . 3 9 . 5 0 . 0 0 . 0 8 . 3 5 . 9 0 . 0 3 . 8 4 . 1 > 50 m i l e s 2 . 3 0 . 0 0 . 0 0 . 0 0 . 0 5 . 9 0 . 0 3 . 8 1 . 5 N 43 4 2 1 3 8 1 2 1 7 3 6 2 6 1 9 7 at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from occurred at the residence where only the victim lived were also significantly more likely to involve offenders who victimized acquaintances. These offenders, however, were signifi- cantly more likely to victimize adults and to use physical force during the assault. Finally, the 51 offenses in which the offender residence, victim residence, and offense location were different from one another were significantly more likely to involve offenders who not only used physical force but who also victimized adults and strangers. TYPE OF VICTIM CONTACT As noted above, determining the location where offenders first established contact with their victims was often difficult, particularly for offenders who had known their victims for some time. More important, however, the data show that residential proximity had only modest relevance in a majority of the 224 reoffenses. More specifically, 79 offenders (35%) directly established contact with the victim. For these direct-contact offenders, they met their victims by approaching them on the street, meeting them in a bar, or breaking into the victim’s home. For the remaining 65% (n =145), however, the offenders were biologically related to their victims (14%), or they gained access to their victims through a form of col- lateral contact such as a girlfriend, wife, coworker, friend, or acquaintance (51%). Thus, for the biological-contact and collateral-contact offenders, residential proximity was not nearly as important as social or relationship proximity. In Table 4, we examine the characteristics of the 224 offenders by the three types of victim contact. As indicated by the results from a chi-square significance test, the 79 direct-contact offenders had a significantly greater number of institutional disciplinary convictions in the past 12 months prior to release than did either the collateral- or biological- contact offenders. These offenders were also significantly more likely to be released to intensive supervision. In their sex reoffense, where they established direct contact with their victims, all of the direct-contact offenders victimized acquaintances and strangers. Indeed, it was almost evenly split between the two, although nearly one third assaulted a stranger adult female victim. Direct-contact offenders were significantly less likely to victimize those younger than the age of 13 (i.e., “child”) in either their previous or current offense. Instead, they were much more likely to victimize adults. In fact, adults were the victims in 54% (n =42) of their reoffenses, which is more than 4 times greater than for the other recidivists. Furthermore, these offenders were more likely to have a history of victimizing adult strangers. For example, in their previous sex offense, 35% (n =28) had victimized adults, whereas 34% (n =27) had victimized strangers. Direct-contact offenders were significantly more likely to commit the reoffense at a location that was different from both their residence and that of the victims. Perhaps as a consequence of their tendency to victimize adult strangers at a location where neither they nor their victims lived, direct-contact offenders were signifi- cantly more likely to use physical force. Compared to the other recidivists, the 113 collateral-contact offenders were significantly more likely to victimize acquaintances in both their prior and current sex crimes. These offenders, moreover, were much more likely than direct-contact offenders to victimize children in both their previous and current sex offenses. More specifically, 53% (n =60) of the acquaintance victims in their reoffenses were females younger than the age of 18. Given that 50% (n =56) offended against juvenile female acquaintances in their prior offense, these 494 CRIMINAL JUSTICE AND BEHAVIOR at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from Duwe et al. / RESIDENTIAL PROXIMITY & SEX OFFENSE RECIDIVISM 495 TABLE 3: Recidivist Characteristics by Residence and Offense Location (OL) Same V–O Different V–O Same V–O Same Same Residence, Residences Residence/OL’s Offender/OL Victim/OL Different Location and Location Avg. distance (miles)* 0.0 0.0 9.4 22.1 13.3 Median distance (miles) 0.0 0.0 4.5 8.1 4.0 Offender demographics Race White (%) 80.0 68.4 61.2 100.0 62.7 African American (%) 17.8 26.3 34.7 0.0 27.5 American Indian (%) 2.2 3.9 4.1 0.0 9.8 Asian (%) 0.0 1.3 0.0 0.0 0.0 Avg. age at release (years) 32.0 34.7 33.2 33.7 31.7 Metro-area commit (%) 43.2 52.6 65.3 66.7 56.0 Criminal history Multiple prior sex crimes (%) 40.0 28.9 30.6 66.7 33.3 Multiple prior felonies (%) 60.0 53.9 71.4 66.7 54.9 Institutional Recent discipline convictions 1.9 3.5 4.8 0.7 4.0 Length of stay (months) 22.7 26.9 33.0 26.5 27.9 Completed treatment (%) 9.1 9.2 14.3 0.0 6.0 Treatment dropout (%) 11.4 6.6 8.2 0.0 8.0 Did not enter treatment (%) 77.8 84.2 77.6 100.0 84.3 Postrelease Length of supervision (months) 19.8 23.2 24.1 10.5 20.3 ISR (%) 6.7 11.8 20.4 0.0 11.8 SR (%) 91.1 80.3 73.5 100.0 78.4 Discharge 2.2 7.9 6.1 0.0 9.8 SRRS (number) 0.5 0.5 0.6 0.0 0.5 Reoffense Alcohol/drugs (%)** 9.5 34.3 25.0 50.0 19.6 Physical force (%)* 22.2 34.2 57.1 0.0 52.9 Supervised at time of offense (%) 20.0 25.0 36.7 0.0 37.3 Time unsupervised (months) 44.3 39.0 40.0 52.2 29.1 Prior victim characteristics Female (%) 88.1 80.0 89.6 100.0 84.4 Child (%) 40.0 40.8 40.8 0.0 39.2 Adolescent (%) 37.8 42.1 32.7 66.7 21.6 Adult (%) 15.6 15.8 24.5 33.3 27.5 Family (%) 22.2 27.6 16.3 33.3 15.7 Acquaintance (%) 73.3 60.5 59.2 33.3 60.8 Stranger (%)** 4.4 11.8 24.5 33.3 23.5 Reoffense victim characteristics Female (%) 88.9 78.9 93.9 100.0 88.2 Child (%)* 62.2 44.7 32.7 0.0 31.4 Adolescent (%)** 22.2 35.5 20.4 100.0 37.3 Adult (%)* 15.6 19.7 46.9 0.0 31.4 Family (%)* 57.8 17.1 12.2 100.0 9.8 Acquaintance (%)* 42.2 67.1 61.2 0.0 39.2 Stranger (%)* 0.0 15.8 26.5 0.0 51.0 N (%)45 (20.1) 76 (33.9) 49 (21.9) 3 (1.3) 51 (22.8) Note.V–O =victim and offender; SR =supervised release; ISR =intensive supervised release; SRRs =supervised release revocations. *Pearson χ2 significant at the .01 level. **Pearson χ2 significant at the .05 level. at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from 496 CRIMINAL JUSTICE AND BEHAVIOR TABLE 4: Recidivist Characteristics by Type of Victim Contact Characteristics Direct Contact Collateral Contact Biological Contact Total Offender demographics Race White (%) 60.8 69.0 84.4 68.3 African-American (%) 32.9 24.8 15.6 26.3 American Indian (%) 5.1 6.2 0.0 4.9 Asian (%) 1.3 0.0 0.0 0.4 Average age at release (years) 32.8 32.3 36.7 33.1 Metro-area commit (%) 57.7 54.5 46.9 54.5 Criminal history Multiple prior sex crimes (%) 30.4 38.1 21.9 33.0 Multiple prior felonies (%) 53.2 64.6 56.3 59.4 Institutional Recent discipline convictions** 5.6 2.5 2.4 3.5 Length of stay (months) 29.8 26.2 27.5 27.6 Completed treatment (%) 9.0 8.9 12.5 9.5 Treatment dropout (%) 6.4 9.8 6.3 8.1 Did not enter treatment (%) 83.5 80.5 81.3 81.7 Postrelease Length of supervision (months) 25.2 20.9 17.2 21.9 ISR (%)* 21.5 8.0 6.3 12.5 SR (%)** 70.9 85.8 87.5 80.8 Discharge (%) 7.6 6.2 6.3 6.7 SRR’s (number) 0.56 0.46 0.66 0.53 Reoffense Alcohol/drugs (%) 36.7 30.1 18.7 30.8 Physical force (%)* 59.5 31.9 25.0 40.6 Supervised at time of offense (%) 34.2 24.8 31.3 29.0 Time unsupervised (months) 34.8 42.5 32.9 38.7 Distance to sex reoffense Same residence/location* 3.8 29.2 28.1 20.1 Same offender/offense location 30.4 37.2 31.3 33.9 Same victim/offense location 25.3 20.4 18.8 21.9 Same V–O residence/different location** 0.9 0.0 6.3 1.3 Different residences/location* 40.5 12.4 15.6 22.8 Prior victim characteristics Female (%) 82.7 85.0 90.3 85.0 Child (%)* 24.1 49.6 59.4 42.0 Adolescent (%) 33.6 40.5 31.3 35.7 Adult (%)* 35.4 16.8 9.4 22.3 Family (%)* 8.9 22.1 50.0 21.4 Acquaintance (%)* 57.0 72.6 40.6 62.5 Stranger (%)* 34.2 5.3 9.4 16.1 Reoffense victim characteristics Female (%) 83.5 88.5 87.5 86.6 Child (%)* 19.0 52.2 62.5 42.0 Adolescent (%) 27.8 33.6 28.1 30.8 Adult (%)* 53.5 14.2 9.4 27.2 Family (%)* 0.0 18.6 100.0 23.7 Acquaintance (%)* 45.6 74.3 0.0 53.6 Stranger (%)* 54.4 7.1 0.0 22.8 N (%)79 (35.3) 113 (50.4) 32 (14.3) 224 Note.V–O =victim and offender; SR =supervised release; ISR =intensive supervised release; SRRs =supervised release revocations. *Pearson χ2 significant at the .01 level. **Pearson χ2 significant at the .05 level. at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from Duwe et al. / RESIDENTIAL PROXIMITY & SEX OFFENSE RECIDIVISM 497 offenders had a relatively high rate of specialization. In contrast to those who established direct contact with the victims, collateral contact offenders were significantly more likely to commit the reoffense in the residence that they shared with the victim. The 32 recidivists who offended against a biological family member were significantly more likely to victimize a child in both their previous and current sex crimes. These offenders were also significantly more likely to have a history of offending against family members. Indeed, 50% (n =16) victimized a family member in their previous offense. Like collateral- contact offenders, these recidivists were, compared to direct-contact offenders, significantly less likely to use physical force but significantly more likely to commit the offense at the residence that they shared with the victim. Of the three reoffenses that took place in which the offense location was different from the residence shared by the victim and offender, two involved biological-contact offenders. OFFENDER RESIDENCE–FIRST CONTACT DISTANCE FOR DIRECT-CONTACT OFFENDERS In assessing the potential deterrent effects of residency restrictions on sexual reoffend- ing, it is, as noted earlier, necessary to focus on the 79 direct-contact cases. As shown in Table 5, it was not possible to estimate the offender residence–first contact distance for 13 of the cases because of unavailable address information for either the offender’s residence or the first contact location. However, even if it were possible to estimate the first contact distance, none of the cases would have likely been affected by residency restrictions according to the criteria outlined above. For example, in 10 of the 13 cases, the victim was an adult. In the 3 cases involving juvenile victims, 1 offender met the victim through his occupation. In the other two cases, the offenders established “consensual” romantic relation- ships with the victims, both of whom were 14 years old. One of the offenders, who was 24 at the time, met the victim at a party attended by mutual friends, and the other offender, who was 19 years old, “picked up” the victim as she was taking a walk from her home. Given that 4 offenders established contact via telephone and 1 offender initiated contact via the Internet, there were 61 direct-contact cases in which address information was avail- able. Of the 61 cases, more than half (n =31) contacted their victims beyond a mile from where they were residing at the time of the offense. In the other 30 cases, the offenders met their victims less than a mile away from their home. However, 1 of these offenders victimized an inmate while he was incarcerated at a county jail, whereas another offender molested his roommate at a halfway house following his release from prison. Because residency restric- tions would not apply in either situation, both cases were excluded, lowering the total to 28. Of the 28 cases, 21 would qualify for less than a 2,500-foot (less than 0.5 miles) zone, whereas this number would drop to 16 for a 1,000-foot (less than 0.2 miles) zone. RESIDENTIAL PROXIMITY FOR DIRECT-CONTACT OFFENDERS In Table 6, we compare the characteristics of the 28 offenders who established direct vic- tim contact within 1 mile of their residence with the remaining 196 offenders. When the 28 offenders established direct victim contact within 1 mile of their residence, they were likely to target an adult female stranger. Indeed, 43% (n =12) of the victims in their reoffense were adults, 79% (n =22) were females, and 68% (n =19) were strangers. More specifi- cally, the results from a t test reveal that the percentages of adult and stranger victims were at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from 498 CRIMINAL JUSTICE AND BEHAVIOR significantly higher than those for the other recidivists. The 28 residential proximity offend- ers were also more likely to have victimized a stranger in their prior offense. Because of their greater victimization of adult female strangers, these offenders were significantly more likely to have used physical force during the sex crime. Finally, compared to the other recidivists, they were significantly less likely to commit the reoffense at a location where they shared the residence with the victim. But how many of the 28 cases might have been prevented by a law barring sex offenders from living near prohibited areas such as schools, daycare centers, or parks? Twelve cases would be eliminated from consideration because they involved adult victims. Of the remaining 16 cases with juvenile victims, 12 involved offenders who established direct contact within 1,000 feet, and 2 additional cases involved an offender who initiated contact within 2,500 feet. Not one of the 16 cases, however, was facilitated by close proximity to a school, daycare, or park. Instead, the offenders in these 16 cases victimized neighbors, or they made contact with victims near their own property. For example, in 8 of the cases, the victim was a neighbor to the offender in that they lived in the same residential block, trailer park, or apartment building. In 4 of the cases, the offenders made contact with the victims just outside their own property. In 1 incident, the offender met the victim, a 17-year-old male runaway, at a nearby fast food restaurant. One of the offenders molested a child who lived in the same apartment building of an acquaintance he was visiting. In another incident, the offender lived near a shopping mall, which is where he initiated contact with a juvenile victim. And in the final case, the offender gained entrance by breaking into the victim’s home. In general, though, the offenders typically gained access to the victims by enticing them with a ruse: for example, an offer to use the offender’s phone or paying the victim money to clean the offender’s residence. Of the 224 cases, there were a total of 3 in which the offender established contact with the victim at a possible prohibited area where children are known to be present. The loca- tion was a park in 2 of the incidents and a school in the other incident. In 2 of the cases, however, the offender lived more than 10 miles away from the first contact location, whereas the victim in the other case was an adult. Therefore, none of the 224 incidents of sex offender recidivism fit the criteria of a known offender making contact with a child victim at a loca- tion within any of the distances typically covered by residential restriction laws. TABLE 5: Offender Residence–First Contact Distances for Direct-Contact Offenders Distance Number % <1,000 ft. (0.00–0.19 miles) 18 22.8 1,000–2,500 ft. (0.20–0.47 miles) 5 6.3 2,501–5,280 ft. (0.48–0.99 miles) 7 8.8 1–2 miles 6 7.6 3–5 miles 10 12.7 6–10 miles 4 5.1 11–20 miles 4 5.1 >20 miles 7 8.8 Telephone 4 5.1 Internet 1 1.2 Unknown 13 16.5 Total 79 100.0 at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from Duwe et al. / RESIDENTIAL PROXIMITY & SEX OFFENSE RECIDIVISM 499 TABLE 6: A Comparison of Residential Proximity Among Recidivist Sex Offenders Characteristics Residential Proximity Nonresidential Proximity Total Offender demographics Race White (%) 57.1 69.9 68.3 African American (%) 39.3 24.5 26.3 American Indian (%) 3.6 5.1 4.9 Asian (%) 0.0 0.5 0.4 Average age at release (years) 35.2 32.9 33.1 Metro-area commit (%) 50.0 55.2 54.5 Criminal history Prior sex crime (%) 28.6 33.7 33.0 Prior felony (%) 57.1 60.0 59.4 Institutional Recent discipline convictions 5.50 3.26 3.55 Length of stay (months) 32.2 27.0 27.6 Completed treatment (%) 0.0 10.8 9.5 Treatment dropout (%) 7.1 8.3 8.1 Did not enter treatment (%) 92.9 80.1 81.7 Postrelease ISR (%) 17.9 11.7 12.5 SR (%) 78.6 81.1 80.8 Discharge (%) 3.6 7.1 6.7 Length of supervision (months) 28.1 21.0 21.9 SRR’s (number) 0.46 0.54 0.53 Reoffense Alcohol/drugs (%) 32.1 30.6 30.8 Physical force (%)** 60.7 37.8 40.6 Supervised at time of offense (%) 42.9 27.0 29.0 Time unsupervised (months) 40.9 38.4 38.7 Distance to sex reoffense Same residence/location* 0.0 100.0 20.1 Same offender/offense location 50.0 31.6 33.9 Same victim/offense location 25.0 21.4 21.9 Same V–O residence/different location 0.0 1.5 1.3 Different residences/location 25.0 22.4 22.8 Prior victim Female (%) 75.0 86.8 85.3 Child (%) 32.1 43.4 42.0 Adolescent (%) 39.3 35.2 35.7 Adult (%) 28.6 21.4 22.3 Family (%) 10.7 23.0 21.4 Acquaintance (%) 46.4 64.8 62.5 Stranger (%)* 42.9 12.2 16.1 Reoffense victim Female (%) 78.6 87.8 86.6 Child (%) 28.6 43.9 42.0 Adolescent (%) 28.6 31.1 30.8 Adult (%)** 42.8 25.0 27.2 Family (%)* 0.0 27.0 23.7 Acquaintance (%)* 32.1 56.6 54.0 Stranger (%)* 67.9 16.4 22.3 N (%) 28 (12.5) 196 (87.5) 224 Note. SR =supervised release; ISR =intensive supervised release; SRRs =supervised release revocations. *t test statistically significant at the .01 level. **t test statistically significant at the .05 level. at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from DISCUSSION Only a minority (35%) of the 224 sex offender recidivists directly established contact with their victims. For those who did, they were much more likely to initiate contact with an adult. But even when offenders contacted juvenile victims directly, it was often more than a mile away from where they lived. Of the few offenders who directly contacted a juvenile victim within close proximity of their residence, none did so near a school, park, playground, or other location included in residential restriction laws. There are several limitations with this study, however, that are worth mentioning. First, as noted earlier, by focusing on the 224 cases resulting in reincarceration, we did not exam- ine the 80 reconvictions and 70 rearrests that took place between 1990 and 2005. Although we found no evidence that housing restrictions would likely have had a deterrent effect on the 224 most serious instances of sexual recidivism, the possibility still remains that they might have an impact on the less serious cases. Second, we did not examine whether the 224 offenders lived near a school, park, or other possible prohibited area at the time of the offense. Thus, there may have been some child molesters examined here who became aroused from living near a location where children congregate but who chose to go to a different location (i.e., more than 1 mile away) to commit the offense. As noted above, however, there were only 3 cases in which the offender established contact with a victim at a possible prohibited area, and 1 involved an adult victim whereas the other 2 were more than 10 miles from the offenders’ residences. Despite these limitations, the results presented here provide very little support for the notion that residency restriction laws would lower the incidence of sexual recidivism, particularly among child molesters. Why, then, does residential proximity appear to matter so little with regard to sexual reoffending? Much of it has to do with the patterns of sexual offending in general. Sex offense rates are generally not related to the number of known sex offenders in a community (Stengel et al., in press). Moreover, sex offenders are much more likely to victimize someone they know. For example, one of the most common victim– offender relationships found in this study was that of a male offender developing a romantic relationship with a woman who has children. The sex offender recidivists examined here used their relationships with these women to gain access to their victims—the women’s children. Similarly, it was relatively common for offenders to gain access to victims through babysitting for an acquaintance or coworker or by living with friends who had children. Thus, in one half of the cases, the offenders established contact with their victims—most of whom were juveniles—through their relationship or acquaintance with another person, almost invariably an adult. Even when offenders established direct contact with victims, they were unlikely to do so close to where they lived. This may be largely because of the fact that offenders are more likely to be recognized within their own neighborhoods. Indeed, many of the child moles- ters whom Levenson and Cotter (2005) interviewed for their study indicated that they were careful not to reoffend close to their own home. As a result, when direct-contact offenders look for a victim, they may be more likely to go to an area relatively close to home (i.e., less than 20 miles) but still far enough away (i.e., greater than 1 mile) to decrease the chances of being recognized. These findings fit with previous research showing that repeat sex offenders typically offend outside their immediate neighborhoods (Canter & Gregory, 1994; Lundrigan & Czarnomski, 2006; Rengert, 2004; Warren et al., 1998). 500 CRIMINAL JUSTICE AND BEHAVIOR at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from Duwe et al. / RESIDENTIAL PROXIMITY & SEX OFFENSE RECIDIVISM 501 Residency restriction laws would likely offer, at best, a marginal impact on the incidence of sexual recidivism. Together with emerging research suggesting that sex offender regis- tration and notification processes have a negligible effect on recidivism (Walker, Madden, Vasquez, VanHouten, & Ervin-McLarty, 2006), this finding casts doubt on the efficacy of such policies. This is not to say that housing restrictions would never prevent a sex offender from reoffending sexually. Based on the results presented here, however, the chances that it would have a deterrent effect are slim. Indeed, during the past 16 years, not one sex offender released from a MCF has been reincarcerated for a sex offense in which he made contact with a juvenile victim at or near a school, park, or daycare center close to his home. In short, it is unlikely that residency restrictions would have a deterrent effect because the types of offenses that such laws are designed to prevent are exceptionally rare and, in the case of Minnesota, virtually nonexistent in the past 16 years. It is still possible, however, that a housing restrictions law could have an impact because laws sometimes have unintended consequences. It is debatable, though, whether the impact would be a positive one. In 2002, Iowa passed a residency restrictions law prohibiting sex offenders from living within 2,000 feet of a child congregation location. Although there are no hard data on the impact of the law, anecdotal evidence suggests that residency restric- tions may limit offender employment prospects, reduce suitable housing opportunities, and threaten the reliability of the sex offender registry by causing more offenders to become homeless, change residences without notifying the authorities, or register false addresses. Moreover, the forced removal of offenders from established residences may have an adverse impact on family members, causing children to be pulled out of school and away from friends and resulting in the loss of jobs and community connections for spouses (Iowa County Attorneys Association, 2006). Recognizing the harmful impact that residency restrictions can have on others besides the offender, the American Correctional Association (2007) has passed a resolution calling for legislatures to examine and consider the negative consequences that these laws create. By making it more difficult for sex offenders to successfully re-enter society, housing restrictions might promote conditions that militate against the goal of reducing the extent to which they recidivate sexually. APPENDIX Coding of Variables Offender race: Dichotomized as White (1) or minority (0). Age at release: The age of the offender in years at the time of release based on the date of birth and release date. Metro-area: A rough proxy of urban and rural Minnesota, this variable measures an offender’s county of commitment, dichotomizing it into either metro-area (1) or greater Minnesota (0). The seven metro-area counties include Anoka, Carver, Dakota, Hennepin, Ramsey, Scott, and Washington. The remaining 80 counties were coded as non-metro area or greater Minnesota counties. Multiple prior felony convictions: Offenders who had more than one prior felony conviction (including the instant offense) were assigned a value of 1, whereas those without multiple prior felony convictions were assigned a value of 0. (continued) at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from 502 CRIMINAL JUSTICE AND BEHAVIOR Multiple prior sex-crime convictions: Offenders with more than one prior sex-crime conviction (including the instant offense) were given a value of 1, whereas those without multiple prior sex crime convictions were given a value of 0. Recent disciplinary history: This variable measures the number of disciplinary convictions that an offender received in the final 12 months prior to his initial release from prison. Sex offender treatment: Three dichotomous dummy variables were created to measure prison-based sex offender treatment during the offender’s term of imprisonment prior to release. Offenders’ involvement in sex offender treatment was not included, however, if they participated in treatment during either a prior commitment to prison or a subsequent return to prison (either for a supervised release violation or for the new sex reoffense). The three variables were offenders who successfully completed treatment or were participating until the time of release (1 =treatment completers/participants; 0 =treatment dropouts or nonparticipants), offenders who were terminated from treatment or voluntarily quit (1 =treatment dropouts; 0 =treatment completers/participants or nonparticipants), and those who never entered treatment (1 =nonparticipants; 0 =treatment completers/ participants, and dropouts). Length of postrelease supervision: The number of months between an offender’s first release date and the end of postrelease supervision, for example, the sentence expiration or conditional release date, the greater of the two. Intensity of postrelease supervision: Three dichotomous dummy variables were created to measure the level of postrelease supervision to which offenders were released. The three variables were intensive supervised release (ISR) (1 =ISR,0 =non-ISR), supervised release (SR) (1 =SR,0 =non-SR), and discharge (1 =discharge or no supervision,0 = released to supervision).4 Supervised release revocations (SRR’s): The number of times during an offender’s sex crime sentence when he returned to prison as a supervised release violator. Prior victim gender: Dichotomized as either a female victim (1) or a male victim (2), this variable measures whether offenders victimized a female or male victim in the sex offense for which they were previously incarcerated. Victim gender was coded the same for the reoffense. Prior victim age: Three dichotomous dummy variables were created to measure the age of the victim for the sex offense for which the recidivists were previously incarcerated. The three variables were child victims (1 =victim younger than the age of 13; 0 =victims 13 and older), adolescent victims (1 =victims between the ages of 13 and 17; 0 =child and adult victims), and adult victims (1 =victims older than age 17; 0 =victims younger than age 18). The same coding structure was used for the victim’s age in the reoffense. Prior victim–offender relationship: Three dichotomous dummy variables were created to measure the offender’s relationship to the victim for the sex crime for which he was previously incarcerated. The three variables were stranger victims (1 =stranger victim; 0 =known or nonstranger victim), acquaintance victims (1 =acquaintance victim; 0 =nonacquaintance victim), and family member victims (1 =family member victim; 0 =non-family-member victim). Victim–offender relationship was coded the same for the reoffense. APPENDIX (continued) at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from NOTES 1. Taylors Falls and Wyoming are currently the only local governments in Minnesota that have passed residency restric- tions for sex offenders. None of the 224 offenders examined in this study were released to either of these two locations. 2. In Minnesota, there are currently 11 correctional facilities that house adult prisoners committed to the Commissioner of Corrections. The 3,166 sex offenders were released from 8 of these facilities. The only 3 facilities from which these offend- ers were not released were MCF–Red Wing, which contains mostly juvenile offenders, and MCF–Willow River and MCF–Togo, which hold the male and female boot camp (i.e., Challenge Incarceration Program) populations, respectively. The 224 recidivists examined here were released from 7 facilities. The eighth facility, MCF–Shakopee (the main correctional facility for female offenders), did not account for any releases due to the absence of females among the 224 recidivists. 3. Information regarding the specific data source(s) used for each item can be obtained from the corresponding author. 4. The standards for intensive supervised release (ISR) are more rigorous in comparison to those for supervised release. In particular, offenders placed on ISR experience a greater frequency of contacts with supervision agents, are required to maintain steady employment, are expected to comply with random alcohol and/or drug testing, and are subjected to unan- nounced face-to-face contacts with their supervision agents at both their residence and place of work. Because of the increased supervision requirements, ISR agents handle a smaller caseload, which has been capped at 15 offenders per agent according to Minnesota law. REFERENCES American Correctional Association. (2007). Resolution on Neighborhood Exclusion of Predatory Sex Offenders. Retrieved from http://www.aca.org/government/policies.asp Brantingham, P. J., & Brantingham, P. L. (1984). Patterns in crime. New York: Macmillan. Canter, D.V., & Gregory, A. (1994). Identifying the residential location of rapists. Journal of the Forensic Science Society, 34, 169-175. Colorado Department of Public Safety. (2004). Report on safety issues raised by living arrangements for and location of sex offenders in the community. Denver, CO: Sex Offender Management Board. Greenfield, L. (1997). Sex offenses and offenders: An analysis of data on rape and sexual assault. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics. Groff, E., & McEwen, T. (2006). Exploring the spatial configuration of places related to homicide events, final report. Alexandria, VA: Institute for Law and Justice. Hanson, R. K., & Morton-Bourgon, K. (2004). Predictors of sexual recidivism: An updated meta-analysis (User Report No. 2004-02). Ottawa: Department of the Solicitor General of Canada. Harris, A. J. R., & Hanson, R. K. (2004). Sex offender recidivism: A simple question. Ottawa: Public Safety and Emergency Preparedness Canada. Iowa County Attorneys Association. (2006). Statement on sex offender residency restrictions in Iowa. Des Moines: Author. Langan, P. A., & Levin, D. J. (2002). Recidivism of prisoners released in 1994. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics. Langan, P. A., Schmitt, E., & DuRose, M. (2003). Recidivism of sex offenders released from prison in 1994. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics. Levenson, J. S., & Cotter, L. P. (2005). The impact of sex offender residence restrictions: 1,000 feet from danger or one step from absurd? International Journal of Offender Therapy and Comparative Criminology,49, 168-178. Levenson, J. S., & D’Amora, D. A. (2007). Social policies designed to prevent sexual violence: The emperor’s new clothes? Criminal Justice Policy Review, 18,168-199. Lundrigan, S., & Czarnomski, S. (2006). Spatial characteristics of serial sexual assault in New Zealand. The Australian and New Zealand Journal of Criminology,39, 218-231. Minnesota Department of Corrections. (2003). Level three sex offenders’residential placement issues. St. Paul: Author. Minnesota Department of Corrections. (2007). Sex offender recidivism in Minnesota. St. Paul: Author. Mustaine, E. E., & Tewksbury, R. (in press). Registered sex offenders, residence, and the influence of race. Journal of Ethnicity in Criminal Justice. Mustaine, E. E., Tewksbury, R., & Stengel, K. M. (2006a). Social disorganization and residential locations of registered sex offenders: Is this a collateral consequence? Deviant Behavior,27, 329-350. Mustaine, E. E., Tewksbury, R., & Stengel, K. M. (2006b). Residential location and mobility of registered sex offenders. American Journal of Criminal Justice,30(2), 177-192. Nieto, M., & Jung, D. (2006). The impact of residency restrictions on sex offenders and correctional management practices: A literature review. Sacramento: California Research Bureau. Rengert, G. F. (2004). Journey to crime. In G. Bruinsma, H. Elffers, & J. de Keijser (Eds.),Punishment, places and perpe- trators: Developments in criminology and criminal justice research. Portland, OR: Willan Publishing. Duwe et al. / RESIDENTIAL PROXIMITY & SEX OFFENSE RECIDIVISM 503 at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from Rengert, G. F., Piquero, A., & Jones, P. (1999). Distance decay reexamined. Criminology,37, 427-446. Sample, L., & Bray, T. M. (2006). Are sex offenders different? An examination of rearrest patterns. Criminal Justice Policy Review,17, 83-102. Snyder, H. N. (2000). Sexual assault of young children as reported to law enforcement: Victim, incident and offender charac- teristics. Washington, DC: Bureau of Justice Statistics. Stengel, K. M., Tewksbury, R., & Mustaine, E. E. (in press). Examining rates of sexual offenses from a routine activities perspective. Victims & Offenders. Tewksbury, R. (2007). Exile at home: The unintended collateral consequences of sex offender residency restrictions.” Harvard Civil Rights—Civil Liberties Law Review,42, 531-540. Tewksbury, R., & Lees, M. B. (2006). Perceptions of sex offender registration: Collateral consequences and community expe- riences. Sociological Spectrum, 26, 309-334. Tewksbury, R., & Mustaine, E. E. (2006). Where to find sex offenders:An examination of residential locations and neighbor- hood conditions. Criminal Justice Studies,19,61-75. Tewksbury, R., & Mustaine, E. E. (in press–a). Collateral consequences and community re-entry for registered sex offenders with child victims: Are the challenges even greater? Journal of Offender Rehabilitation. Tewksbury, R., & Mustaine, E. E. (in press–b). Where registered sex offenders live: Community characteristics and proximity to possible victims. Victims & Offenders. Tita, G., & Griffiths, E. (2005). Traveling to violence: The case for a mobility-based spatial typology of homicide. Journal of Research in Crime and Delinquency, 42, 275-308. Turley, E., & Hutzel, L. (2001). West Virginia Sex Offender Study. Charleston, WV: Division of Criminal Justice Services. Retrieved November 8, 2005, from http://www.wvdcjs.com/publications/ Walker, J. T., Golden, J. W., & VanHouten, A. C. (2001). Geographic link between sex offenders and potential victims: A routine activities approach. Justice Research and Policy,3, 15-34. Walker, J. T., Madden, S., Vasquez, B. E., VanHouten, A. C., & Ervin-McLarty, G. (2006). Influence of sex offender regis- tration and notification laws in the United States. Little Rock: Arkansas Crime Information Center. Warren, J., Reboussin, R., Hazelwood, R. R., Cummings, A., Gibbs, N., & Trumbetta, S. (1998). Crime scene and distance correlates of serial rape. Journal of Quantitative Criminology,14, 35-59. Zandbergen, P. A., & Hart, T. C. (2006). Reducing housing options for convicted sex offenders: Investigating the impact of residency restriction laws using GIS. Justice Research and Policy,8, 1-24. 504 CRIMINAL JUSTICE AND BEHAVIOR at University of British Columbia Library on April 27, 2010 http://cjb.sagepub.comDownloaded from