The Efficacy of County-Level Sex Offender Residence Restrictions in New York more

The Efficacy of County-Level Sex Offender Residence Restrictions in New York D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on Keywords: Sex Offenders, Residence Restrictions, New York, Incapacitation, Deterrence Author Bio: Kelly M. Socia is an assistant professor in the Department of Sociology at the University of New Mexico. He recently received his PhD in criminal justice from the School of Criminal Justice at the University at Albany. His research interests include offender reentry and recidivism, registered sex offenders, public policymaking, geographic information systems, and spatial analyses. Author Note: This project was supported by Award No. 2010-IJ-CX-0004, awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. Further, sex crime arrest data was supplied by the New York State Division of Criminal Justice Services. The opinions, findings, and conclusions or recommendations expressed in this publication are those of the author and do not necessarily reflect those of either the Department of Justice or the New York State Division of Criminal Justice Services. The author would like to thank Andy Davies, M.A., Chris Dum, M.A., Andy Wheeler (M.S., Statistics), and Drs. Robert Apel, Jill Levinson, Alan Lizotte, David McDowall, Steven Messner, Greg Pogarsky, Jeffrey Sandler, Janet Stamatel for help comments on earlier drafts. Unfortunately, all errors and omissions remain the responsibility of the author. Kelly M. Socia Assistant Professor Department of Sociology, University of New Mexico 1915 Roma NE, Suite 1108 – MSC05 3080 Albuquerque, NM 87131, USA ksocia@gmail.com v. 505-277-0250 f. 505-277-8805 Abstract This study examines whether county sex offender residence restrictions were associated with reduced sex crime arrest rates in New York State. In doing so, this study draws on the limited prior research regarding the effectiveness of residence restrictions, and on the extensive literature regarding the incapacitation and deterrence of crime through public policy measures. Results indicate that residence restrictions were not associated with significantly reduce arrests for sex crimes committed either by RSOs or non-RSOs against child victims. However, results suggested that these policies may generally deter some non-RSOs from sexually victimizing adults. Implications for future research and policymaking are discussed. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on 1 Since at least the early 1990s, a number of public policies have applied specifically to registered sex offenders (RSOs) either returning to or already living in the community. These policies have included mandatory public registration, community notification, temporary or lifetime electronic monitoring, chemical castration, civil commitment, and residence restrictions (see Bonnar-Kidd, 2010; Cohen & Jeglic, 2007; CSOM, 2008; Farkas & Stichman, 2002; D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on Levenson & D'Amora, 2007; Matson & Lieb, 1997; Socia & Stamatel, 2010; Zgoba, 2011). These policies attempt to reduce sex crime rates through either incapacitative measures (e.g., electronic monitoring, chemical castration, civil commitment, residence restrictions), or specific deterrence measures (e.g., public registration, community notification). Residence Restrictions Sex offender residence restrictions are one of the most recent and controversial of these policies. Residence restrictions prohibit certain RSOs from living within a given distance of specific places where children might gather, such as schools, daycares, parks, and playgrounds. These locations may be collectively referred to as child congregation locations. In doing so, the expectation is that RSOs will have a harder time finding and approaching young children that they can sexually assault, which will reduce sexual recidivism rates and, as a result, reduce the overall number of recidivistic sex crimes.1 First passed in 1995 at the state level and in 2005 at the county and local level, residence restrictions have become extremely popular throughout the United States, but without evidence that they are effective.2 In fact, only a few studies have analyzed whether residence restrictions have (or would have) reduced sex crime rates (e.g., Blood, Watson, & Stageberg, 2008; Duwe et al., 2008; Minnesota Department of Corrections (MNDOC), , 2007; see also Zandbergen et al., 2 2010). These studies have found little support that residence restrictions are effective at preventing the types of sex crimes they target. However, these studies have only considered sex crimes involving child victims. As such, no research has directly explored whether the incapacitative relocation inherent in these policies has affected sex crimes committed by RSOs involving adult victims. Furthermore, no research D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on has specifically explored whether residence restrictions may be effective at deterring sex crimes committed by individuals who are not the RSOs directly subject to these policies (i.e., exploring the general deterrence effect for non-RSOs). The potential for a general deterrence effect of non-RSOs is an important consideration for the evidence-based support of these policies, particularly if they do not show any incapacitative effects on the RSOs who are subject to them (see more generally Pawson, 2002, 2006). Additionally, given the numerous unintended consequences that these policies generate for the RSOs subject to the restrictions (see Bonnar-Kidd, 2010; Levenson, 2009; Soule & Earl, 2001), a general deterrence effect for non-RSOs seems plausible.3 This study addresses these research gaps by exploring the association between countylevel residence restrictions and sex crime arrest rates in New York counties between January 1998 and December 2009. In analyzing this association, this study provides important contributions to the research fields of sex offender policy, the effectiveness of criminal justice policies generally, and incapacitation and deterrence theories. Further, by providing policymakers with evidence suggesting that these policies either do or do not work as intended, this study can help promote evidence-based policy decisions, potentially reduce the difficulties that RSOs experience when attempting to reenter a community after incarceration, and help promote policies that successfully protect residents. 3 Residence Restrictions in New York While a comprehensive review of existing state, county, and local residence restrictions in the United States is outside the scope of the present study, the next section provides a more detailed overview of the history and current state of residence restrictions in New York, the focus of the present study.4 D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on New York does not have an existing state residence restriction, but it does contain county and local-level policies throughout the state (see New York State Office of Sex Offender Management (NYS OSOM), , 2010).5 These restrictions were passed about a decade after the statewide implementation of sex offender registration and community notification laws (see Sandler, Freeman, & Socia, 2008). On November 3, 2005, the village of Candor in Tioga County passed New York's first local-level residence restriction, and a few weeks later on November 30, Cayuga County passed the state’s first county-level policy (NYS OSOM, , 2010). Presently, while county-level residence restrictions in New York range from 500 to 2,000 feet in size, most are set at 1,000 feet and include some combination of schools, daycares, parks, and/or playgrounds in their scope. Appendix A identifies which New York counties had either a countylevel residence restriction or any local-level policies in place as of December 3, 2009.6 Existing Research on the Efficacy of Residence Restrictions Only two studies have examined whether residence restrictions are effective at reducing sex crimes.7 The earliest study analyzed the retroactive case histories of 224 recidivist RSOs in Minnesota from 1990 to 2006 (Duwe et al., 2008; MNDOC, , 2007). Although Minnesota did not have a residence restriction at the time of the study, researchers concluded that not one sex crime committed by previously convicted sex offenders in the 16 years of data would have potentially been stopped had a residence restriction been in place. 4 The most recent study found no evidence that Iowa’s statewide residence restriction had reduced sex crimes against minors (Blood et al., 2008). However, only three years of annual conviction rates were compared, and sex crimes committed by RSOs and non-RSOs were not separated. As a result, that study’s findings combined any unique specific and general deterrence effects the policy may have had. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on While this limited prior research indicates that residence restrictions may not be effective at reducing sex crimes, this research is still in its infancy. Additionally, some types of sex crimes have not yet been considered, such as those involving adult victims, even though many residence restrictions do not distinguish between convicted sex offenders who have child victims and those who have adult victims. When considering how a residence restriction might affect sex crimes via incapacitation or general deterrence, it is important to acknowledge that not all sex crimes are equivalent. In fact, a single residence restriction may affect different types of sex crimes in unique ways. In order to understand the various effects of a residence restriction, these different types of sex crimes must be identified and explained. Types of Sex Crimes Sex crimes may be roughly separated into four types based on the victim and the offender. The first type involves those crimes committed against child victims by individuals with prior convictions for sex crimes (i.e., current RSOs). The second type involves sex crimes committed against adult victims by current RSOs. These first two types can be considered recidivistic sex crimes, as these crimes involve the sexual recidivism of RSOs. The third type involves sex crimes committed against child victims by individuals without any prior sex crime convictions (i.e., non-RSOs). The fourth type involves sex crimes committed against adult victims also by non-RSOs. These latter two types can be considered non-recidivistic sex crimes, 5 as these crimes do not involve a current RSO sexually recidivating. As such, the term ‘sex crime’ is a multi-dimensional concept, and has important implications for how residence restrictions could potentially affect each type of sex crime. Potential Effects of Residence Restrictions Recidivistic Sex Crimes D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on Residence restrictions could have unique effects on the rate of recidivistic sex crimes. These include crime-reducing effects involving the incapacitation of RSOs, or crime-increasing effects involving the increase in hardships that RSOs are subject to. Incapacitation effects. Residence restrictions are meant to stop sex crimes committed against child victims by RSOs who live near child congregation locations, and who make initial contact with their victims around these same locations. This goal is presumed to be achieved through a form of incapacitation. That is, by relocating the residences of RSOs away from areas where children congregate, it is expected that these offenders will no longer be able to access these pools of potential child victims as easily (see Farkas & Stichman, 2002). However, this type of sex crime is extremely rare (Duwe et al., 2008; MNDOC, , 2007), due in part to the vast majority of crimes occurring between acquaintances or family members (Greenfield, 1997; Snyder, 2000), particularly for child victims. Therefore, any sizable effect that these policies have on the overall rate of recidivistic sex crimes, committed against either child or adult victims, would likely occur from relocating RSOs into different neighborhoods (see Chajewski & Mercado, 2009; Socia, 2011a; Youstin & Nobles, 2009; Zandbergen & Hart, 2006; Zgoba, Levenson, & McKee, 2009), rather than from increasing their distance from child congregation locations. This is because relocating RSOs into less populated neighborhoods may 6 reduce the ease with which RSOs become acquainted with and build relationships with future victims.8 Alternatively, a decrease in non-recidivistic sex crimes could be due to a decrease in reporting rates if victims were less likely to report a family member in order to ‘spare’ the offender from the consequences of a residence restriction. However, it seems unlikely that this D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on effect would occur with enough frequency to affect overall rates of non-recidivistic sex crimes committed against either child or adult victims (see also Prescott & Rockoff, 2011). Increased hardships. Residence restrictions may instead increase recidivistic sex crimes if relocating RSOs to certain neighborhoods limits their ability to access stable housing, suitable employment, supportive family members, and/or adequate treatment facilities (see Barnes, Dukes, Tewksbury, & De Troye, 2009; Casady, 2009; CSOM, 2007, 2008; Leon, 2011; Levenson, 2008; Levenson & Cotter, 2005; MNDOC, , 2003; Zandbergen & Hart, 2006). This is due in part to the reduction in stability and the increase in financial and emotional stress these RSOs might experience (Levenson, 2008; Levenson & Cotter, 2005; MNDOC, , 2003; Willis & Grace, 2008), which can ultimately hamper successful reentry and rehabilitation. This consequence seems particularly likely if residence restrictions force RSOs to become clustered in only a few neighborhoods that offer limited housing, rather than dispersing them throughout many other (less urban) neighborhoods (see Berenson & Appelbaum, in press; Socia, 2011a). No effects. Finally, it is possible that a residence restriction simply has no effect on the sex crimes committed by RSOs, other than displacing these individuals (and potentially their future crimes) from one neighborhood to another. This seems probable given the current lack of research supporting these policies (see Harris & Lurigio, 2010; Pawson, 2002, 2006; Socia & Stamatel, 2010; Tewksbury & Levenson, 2007). If this were the case, a countywide residence 7 restriction would be unlikely to affect recidivistic sex crimes unless the policy resulted in displacing many RSOs outside of the county. Non-Recidivistic Sex Crimes Residence restrictions could have unique effects on non-recidivistic sex crimes as well. These include crime-reducing effects involving the general deterrence of non-RSOs, or crime- D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on increasing effects involving an increase in the reporting and/or investigating of all sex crimes. General deterrence. Residence restrictions may deter sex crimes committed by individuals without prior sex crime convictions (and therefore who are not directly subject to these policies). Specifically, the threat of being subject to a residence restriction, and the unintended consequences that they can cause for RSOs (see Bagley, 2008; Levenson, 2008, 2009; Levenson & Hern, 2007; Levenson & Tewksbury, 2009; Tewksbury, 2007; Walker, 2007), may be enough to deter individuals from committing their first sex crime. Similarly, the media attention that comes with the passage of these policies may serve as a reminder to potential offenders of the entire ‘package’ of consequences that come with a sex crime conviction in addition to restrictions on residence (e.g., public shaming, incarceration, mandatory registration, community notification, electronic monitoring). In these ways, the passage of a residence restriction may deter certain individuals from committing sex crimes in the future in order to avoid potential punishment. If this were the case, then the presence of a county residence restriction should be associated with a decrease in the rate of non-recidivistic sex crime arrests, regardless of the policy’s effect on recidivistic sex crime arrests.9 Note that there would not be a general deterrence effect for current RSOs, as these individuals would already be subject to the threatened punishment(s) implied by these policies. 8 While no prior studies have directly examined the general deterrence effects of registry restrictions, there is some evidence that potential sex offenders can be deterred by the threat of punishment. Specifically, a study of NIBRS data by Prescott and Rockoff (2011) indicated that the presence of sex offender community notification policies may have deterred individuals who were not convicted sex offenders from committing future sex crimes. Thus, it is a feasible that D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on residence restrictions may have a similar effect. Increased reporting/investigation rates. The passage of residence restrictions is typically a highly political process. When passed at either the local or county level, media outlets may run headline stories about the new restrictions, politicians may hold press conferences to discuss the policy, and parents may make a greater effort to discuss with their young children the dangers of strangers and/or the importance of reporting unwanted sexual contact to an authority figure. As such, it seems plausible that the increased focus in the community on preventing sex crimes may increase the likelihood that residents will report sexual assaults, and may also increase the focus of law enforcement on investigating and solving such crimes. If these assumptions are correct, then the presence of a county residence restriction should result in increased non-recidivistic sex crime arrest rates in the county. This may also result in increased recidivistic sex crime arrest rates as well. Since it is unlikely that a residence restriction would actually cause an increase in the commission of sex crimes by individuals not yet subject to the policy, finding an increase in non-recidivistic sex crime arrest rates alone, or in both nonrecidivistic and recidivistic sex crime arrest rates, would help isolate the effects of increased sex crime reporting/investigating from the effects of increased RSO hardships, as the latter would likely result only in increased recidivistic sex crime arrest rates. The Current Study 9 In considering the unique mechanisms that could affect different types of sex crimes, and drawing on incapacitation and deterrence theories, this study determines whether county residence restrictions are associated with any of four different types of county sex crime arrest rates in the state of New York. Research Questions D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on This study considers four research questions: 1) Are county residence restrictions associated with rates of recidivistic sex crime arrests involving child victims? 2) Are county residence restrictions associated with rates of recidivistic sex crime arrests involving adult victims? 3) Are county residence restrictions associated with rates of non-recidivistic sex crime sex crime arrests involving child victims? 4) Are county residence restrictions associated with rates of non-recidivistic sex crime arrests involving adult victims?10 Given the conflicting ways in which these policies could influence the various rates of sex crime arrests, each research question assumes a null hypothesis that county residence restrictions are not associated with sex crime arrest rates. Data and Methodology Sample The study’s sample consisted of county-level data from all 62 counties in the state of New York. Data were collected on a monthly basis for each county between January 1998 and December 2009, resulting in 144 months of data for each county. As such, the unit of analysis was the county-month, yielding 8,928 cases (i.e., 144 months of data for each of the 62 counties in New York). These data came from NYS OSOM (2010) and the New York State Division of Criminal Justice Services’ Computerized Criminal History database (NYS DCJS CCH) (2010). New York provided a particularly useful setting for the study since it did not yet have an 10 ‘official’ statewide residence restriction that could have influenced county-level results (unlike most other states in the U.S.), its registration and community notification provisions were in place statewide about a decade before the first residence restriction was implemented (see Naomi J Freeman, Sandler, & Socia, 2009; Sandler et al., 2008), it contained counties with a wide variation of demographic, social, geographic, and crime characteristics, and it had a mixture of D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on counties with and without residence restrictions in place. Further, the majority of county and local-level laws in New York do not include specific grandfathering provisions, nor do they distinguish RSOs with child victims from those with adult victims, factors that could have otherwise influenced results in undesirable ways were they more common. Dependent Variables The analyses utilized four dependent variables, which were monthly measures of each of four different types of sex crime arrests. These four series measured the rates of arrests for crimes 1) committed against child victims by offenders with prior sex offense convictions (i.e., recidivistic sex crimes committed against children), 2) committed against adult victims by offenders with prior sex offense convictions (i.e., recidivistic sex crimes committed against adults), 3) committed against child victims by offenders without prior sex offense convictions (i.e., non-recidivistic sex crimes committed against children), and 4) committed against adult victims by offenders without prior sex offense convictions (i.e., non-recidivistic sex crimes committed against adults).11 12 As noted earlier, the first two rates measured arrests for recidivistic sex crimes potentially influenced by the direct consequences (incapacitative and otherwise) of a residence restriction on RSOs, while the latter two rates measured arrests for nonrecidivistic sex crimes potentially influenced by the effects of a residence restriction (deterrence or otherwise) on potential sex offenders who are not yet RSOs. 13 11 Independent Variables The main independent variable was a monthly dichotomous indicator of whether a county did (1) or did not (0) have a residence restriction in place. Two dummy variables were included in the model. The first identifies the county, while the second identifies the month-year of the observation to account for temporal influences. A third control variable was also included in the D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on model to account for county-specific temporal trends, such as a constant increase or decrease in sex crimes over time.14 Additionally, measures of two series of monthly crime arrest rates theoretically unrelated to sex crime arrests (i.e., robbery and burglary) were included to control for any general influences on the crime arrest rate that could have occurred within a county that was not specific to sex crimes, such as a crackdown on all violent crimes, a change in policing strategy, or a media campaign to increase crime reporting (see Prescott & Rockoff, 2011). Finally, a lag term measuring the rate of sex crime arrests (specific to the type of sex crimes being examined) in the preceding month was included in the model to control for any short-term autocorrelation within the series of the dependent variable. More detail on the measurements of these variables is available from the author upon request. Analytical Strategy The analyses utilized fixed-effects panel models to determine whether the presence of a county-level residence restriction was associated with any of the four different types of sex crime arrest rates, controlling for other demographic and social indicators. (Use of a model based on the sex crime arrest rate was preferred over using a model based on the count of sex crime arrests as the rate model accounted for year-to-year differences in the population of each county.) Four individual models were estimated, each relating to a specific type of sex crime. A fixed-effects panel helps to control for the relatively static between-county differences (and static within12 county characteristics) by comparing each county to itself over time.15 More information regarding the specifics of the fixed-effects panel model is available from the author upon request. Unfortunately, historical data on the number of RSOs living in each county in a given month or year were not available. Thus, the model could not control for changes in the recidivistic sex crime arrest rate due primarily to increases in the number of RSOs in the county. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on However, since the proportion of RSOs in the population was expected to be fairly static between counties, the fixed-effects panel model should have largely controlled for this influence. [Tables 1 and 2 About Here] Results Descriptive statistics are provided in Table 1. Analysis of variable correlations (available from the author upon request) indicated that multicollinearity is unlikely to be a problem for evaluating the relationship between county and local residence restrictions and various sex crime arrest rates at the county-month level. Results are provided in Table 2, with coefficients presented as odds-ratios. Results are examined in two sections based on whether the sex crime arrests involved recidivistic or non-recidivistic crimes. Effects on Recidivistic Sex Crimes Results in Table 2 indicated that the presence of a county residence restriction was not significantly associated with the rate of arrest for sex crimes committed against child victims by previously convicted RSOs. This supports the existing literature suggesting that these policies are unlikely to have a significant effect on child-victim sex crimes (e.g., Blood et al., 2008; CDPS, , 2004; Duwe et al., 2008; MNDOC, , 2003; 2007; Zandbergen et al., 2010). Effects on Non-Recidivistic Sex Crimes 13 The presence of county residence restrictions was not significantly associated with the rate of sex crimes committed against child victims by individuals without prior sexual convictions. However, the presence of a county residence restriction was associated with a 10 percent decrease in the rate of arrests for sex crimes committed against adult victims by individuals without prior sexual convictions (p < .01). This suggests that county residence D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on restrictions may have deterred at least some rapes committed against adults by ‘new’ offenders. This finding is also consistent with research indicating that the threat of formal sanctions can deter rape (Bachman, Paternoster, & Ward, 1992), and that post-release sex offender policies can have a general deterrence effect (Letourneau, Levenson, Bandyopadhyay, Armstrong, & Sinha, 2010; Prescott & Rockoff, 2011; but see Sandler et al., 2008).16 Discussion This study explored four research questions regarding the effects of county sex offender residence restrictions in New York. The first and second research questions concerned the association between residence restrictions and rates of arrest for recidivistic sex crimes committed against child and adult victims, respectively. The third and fourth research questions concerned the association between residence restrictions and rates of arrest for non-recidivistic sex crimes committed against child and adult victims, respectively. Because the existing research provides no support for the efficacy of residence restrictions to reduce sex crimes (e.g., Blood et al., 2008; Duwe et al., 2008; MNDOC, , 2007; Zandbergen et al., 2010), each research question had a related hypothesis that these policies were not associated with the specific type of sex crime under examination. Findings in Relation to the Existing Literature 14 No incapacitation effects. Overall, the present study supports the prior research (Blood et al., 2008; e.g., Duwe et al., 2008; MNDOC, , 2007; Zandbergen et al., 2010) in finding that residence restrictions are not effective at incapacitating recidivistic sex crimes committed against either children or adults. These findings are not surprising, considering that residence restrictions seek to increase the spatial distance between RSOs and pools of potential stranger victims, D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on whereas most sex crimes occur between acquaintances (Greenfield, 1997; Snyder, 2000). Thus, the relational distance between RSOs and potential victims is likely more important than the spatial distance in terms of preventing future sex crimes, and these results suggest that residence restrictions may not be affecting RSOs’ abilities to form these personal relationships with potential victims. On a related note, these findings also support research suggesting that RSOs’ residential proximity to schools, daycares, or other types of locations is unrelated to their propensity to sexually recidivate (see Zandbergen et al., 2010). This is, of course, assuming that these policies resulted in RSOs actually living farther away from such locations, despite research that finds compliance rates with residence restrictions to be around 50 percent or less in some areas (e.g., Berenson & Appelbaum, in press; Tewksbury & Mustaine, 2006; see also Youstin & Nobles, 2009). No increase in recidivistic sex crimes. The existing literature indicates that residence restrictions may hinder RSOs’ successful return to society by limiting their ability find suitable housing, access treatment centers, and/or live with supportive family members (see Barnes et al., 2009; Casady, 2009; CSOM, 2007, 2008; Leon, 2011; Levenson, 2008; Levenson & Cotter, 2005; MNDOC, , 2003; Zandbergen & Hart, 2006). However, most of these studies did not examine whether the supposed hindrances to reentry would result in increased arrests for sexual 15 recidivism. This study’s findings suggest that any hardships that RSOs are subjected to under county-level residence restrictions are not necessarily resulting in increased recidivistic sex crimes. This is not to say that RSOs are not being rearrested more frequently for any new crimes when subjected to a residence restriction, but rather that they are not more likely to be rearrested for new sex crimes. In fact, if RSOs were being rearrested more quickly for crimes other than sex D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on offenses, this may actually reduce the rate of recidivistic sex crimes since some of these RSOs would have been arrested (and possibly re-incarcerated) before they had a chance to sexually recidivate. Research on whether residence restrictions result in increased rates of arrest of RSOs for non-sexual crimes would be required to confirm such an outcome. Mixed general deterrence effects. The findings related to a general deterrence effect were also mixed. Specifically, results suggested that county residence restrictions were not associated with decreased arrests for non-recidivistic sex crimes committed against child victims. However, county residence restrictions were associated with fewer arrests for non-recidivistic sex crimes committed against adult victims. This may suggest that residence restrictions had a general deterrence effect on these types of sex crimes, which is consistent with the deterrence effects of notification policies found by Prescott and Rockoff (2011). This is important: it is the first time a study has found that residence restrictions may have had any type of crime-reducing effect. This finding also provides a new venue for future deterrence research. Specifically, analyzing the general deterrence effect of these laws on non-recidivistic sex crimes is an unexplored area that would greatly benefit from future longitudinal studies of other states, particularly given that the aggregate-level data explored in the present study does not allow for definite conclusions about why these policies had reduced arrests for such crimes. While general 16 deterrence was proposed as a possible explanation, it is unclear whether this finding is also influenced by other factors unrelated to either the implied punishment that comes with residence restrictions and/or the reminder of the other consequences that come with a sex crime conviction. Still, these mixed findings for a general deterrence effect are actually consistent with the larger body of research on deterrence. For example, conclusions regarding a general deterrence D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on effect from imprisonment risk on crime rates have similarly been mixed, with some research suggesting that increased imprisonment risk led to reductions in crime rates (e.g., Blumstein, Cohen, & Nagin, 1978; Nagin, 1998), and other research finding no such support (Pratt & Cullen, 2005). While it is notable the prior deterrence research has focused mainly on arrest and imprisonment, and less on post-conviction punishments, research specific to sex offender registration and community notification has also yielded mixed conclusions on whether these policies have deterred sex crimes (Letourneau et al., 2010; Sandler et al., 2008; Veysey, Zgoba, & Dalessandro, 2008; Washington State Institute for Public Policy, 2009; Welchans, 2005; Zgoba, Veysey, & Dalessandro, 2010). In line with the mixed findings of both the broader literature on deterrence and the research specific to sex offender policies, the current study supports that residence restrictions may be having a mixed general deterrence effect. Specifically, residence restrictions may only deter some non-recidivistic sex crimes committed against adults, and not those committed against children. Therefore, it is not surprising that the prior studies that have examined rates of sex crime without separating child victims from adult victims have found no significant results. In fact, the findings of this study suggest that unique mechanisms may be influencing certain types of sex crimes and not others. Future research on the effects of sex offender policies would be well served by separating crimes committed by offenders who are current RSOs from those 17 who are not, as well as separating sex crimes committed against child victims from those against adult victims. Explaining These Mixed Findings Unfortunately, it still remains unclear as to exactly why residence restrictions were associated with significantly decreased arrest rates for sex crimes involving adult victims and not D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on child victims. However, there are some potential explanations that would need to be confirmed through future research on the decision-making process of potential sex offenders. Differences between pedophiles and adult rapists. One such explanation may lie in the differences between pedophiles and adult rapists. Individuals are considered to be pedophiles when they are sexually attracted to children (typically 13 years or younger) at least 5 years younger than themselves and are at least 16 years old (American Psychiatric Association, 2000; Fagan, Wise, Schmidt, & Berlin, 2002). In attempting to satisfy such specialized sexual urges, a pedophile may be likely to commit some form of sex crime involving children (e.g., actual physical child molestation, viewing child pornography, voyeurism involving children), which would be illegal even if such a crime did not involve physical contact. 17 Conversely, nonpedophiles may be able to satisfy sexual urges through either legal (e.g., sex with a consenting adult, viewing legal pornography) or illegal means (e.g., patronizing a prostitute, voyeurism, rape). Thus, potential adult rapists may be more easily deterred simply because they have more legal ways in which to satisfy their sexual urges, at least compared to pedophiles.18 In any event, to confirm this hypothesis, data would be required on the individual decision-making processes of potential rapists and potential child molesters (i.e., those who do not yet have prior sexual convictions). While such data would be extremely useful for examining deterrence mechanisms, it would also be extremely difficult to obtain. 18 Grooming relationships. Another explanation for these mixed findings may be due to the length of time and effort child molesters put into grooming their victims. Specifically, many child molesters form relationships with a child (and sometimes the child’s family) in order to groom the child before molesting them, which may involve a lengthy process and result in continued victimized of the same child over a long period of time (see Naomi J. Freeman, 2007; D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on Pryor, 1999; Terry & Tallon, 2004). Thus, the passage of a residence restriction may occur midway through a potential child molester’s grooming process of a specific child. If this were to occur, a residence restriction may have less of a general deterrence effect than for individuals who were not involved in an active grooming relationship (which likely involves most individuals who could potentially rape an adult victim). Limitations Perhaps the most important limitation in this study is the inability to identify arrests for sex crimes involving strangers from those involving non-strangers. One might expect that crimes involving strangers may be more susceptible to general deterrence, and thus rates of arrests for stranger sex crimes may have experienced larger decreases than non-stranger sex crimes. Unfortunately, information on the relationship between the offenders and the victim exists only in individual case files. As noted earlier, given that the present study involved over 350,000 individual sex crime arrests across the state of New York over 12 years, it was simply not realistic to be able to obtain such data on this large a scale. A second important limitation involves the lack of historical/longitudinal data on RSO residences in each county. Unfortunately, as these historical data were unavailable, it remains unclear the extent that increases or decreases in the population of RSOs in each county had 19 influenced the rate of recidivistic sex crimes. Still, the findings suggest that residence restrictions had no significant incapacitation effect on recidivistic sex crime arrests involving child victims. The generalizability of this study’s findings is another limitation. As this study used data only from counties in New York, the generalizability of its findings are obviously limited to counties that have similar characteristics. However, this study was the first to examine the D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on efficacy of residence restrictions using longitudinal data at the county level, which represents a unique contribution to the existing literature. Future researchers should consider examining longitudinal data from other states, as their findings would provide useful comparisons to the conclusions of the current study as well as to the existing prior research. The measure of the presence of a county residence restriction may have not been precise enough to capture subtle differences in the types of such policies. Specifically, while the majority of county-level residence restrictions in New York were fairly similar in size and scope, there was still some variation between them. Additionally, the presence of local residence restrictions in a county could have also influenced results. However, conclusions about county residence restrictions did not change when alternative models were run that controlled for the presence of any local-level residence restrictions within the county (analysis not shown). Finally, the low base-rate of recidivistic sex crimes occurring at the county-month level may have been another limitation. The finding of no association between residence restrictions and arrests involving recidivistic sex crimes against either children or adults may have resulted from the relatively low rate of these crimes occurring at the county-month level. Specifically, the monthly county average for the rate of arrests for recidivistic crimes involving child victims was .11, and for recidivistic sex crimes involving adult victims was .31. This is compared to rates of arrest for non-recidivistic sex crimes involving child victims of 2.44, and for adult victims of 20 7.04. These low base-rates for recidivistic sex crime arrests could result in large standard errors and the increased potential for Type II (false negative) errors. Without either using better countylevel data than currently exists, or increasing the sample size through the inclusion of counties in other states, it is likely that this limitation will be a similar concern for other county-level studies analyzing recidivistic sex crime arrest rates. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on Future Research This study has provided a number of avenues for future research regarding the efficacy of residence restrictions. One such avenue involves replicating this study using data from other states and/or with longitudinal data on the rate of RSOs in each county. Such research would help to both support the present study’s findings and would help address the need for more evidence-based research regarding the effectiveness of residence restrictions. Another avenue for future research involves the examination of individual-level data regarding RSOs’ residences and recidivism following the implementation of a residence restriction. Such individual-level data could help confirm the whether 1) non-compliance, nonenforcement, or grandfathering affects the incapacitative effects of residence restrictions, 2) the hardships associated with a residence restriction are not influencing sexual recidivism, and/or 3) the differences between pedophiles and adult rapists are responsible for the mixed general deterrence effects. Policy Implications While there are many policy implications that stem from this study, perhaps the most important involves the continued passage and enforcement of residence restrictions. Specifically, the results of this study and the existing research suggest that these policies are not working as intended to protect children from recidivistic sex crimes. Further, these policies also do not 21 appear to be incapacitating existing RSOs from attacking adult victims. Thus, despite being popular with many politicians and residents, these policies are simply not effective at achieving the intended goals of protecting residents from RSOs in the community. On the other hand, this study’s findings suggest that residence restrictions may increase the safety of adult residents from sex crimes committed by individuals who are not yet RSOs. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on However, this supposed benefit may not legally or ethically justifiable for implementing policies that appear ineffective at their stated goals, and that subject broad categories of other individuals (i.e., RSOs) to restrictions on where they can live. This is particularly important given the evidence that these policies do not reduce sex crimes committed by RSOs, combined with the fairly extensive body of existing research showing the numerous unintended consequences that these policies can have on successful reentry and rehabilitation. Therefore, policymakers may wish to consider other policies that are either specifically targeted at generally deterring nonrecidivistic sex crimes, or that are also effective at reducing recidivistic sex crimes. Conclusions Overall, this study found that county residence restrictions were not associated with the rate of arrest for recidivistic sex crimes involving child victims, as intended, nor did their presence otherwise increase the safety of children from sexual assault by non-RSOs or the safety of adults from RSOs. However, the presence of a county residence restriction was associated with a decreased rate of arrest for non-recidivistic sex crimes involving adult victims, suggesting a possible general deterrence effect on some non-RSOs. Still, this result does not support the original intention of these laws in protecting children from sex crimes. Although potential explanations were provided as to why non-recidivistic sex crimes against adult victims could have been deterred, while other types of sex crimes were not, the data used in the current study 22 did not allow for further exploration and confirmation of why this difference exists. Still, the results of this study (and much of the prior literature) provide no support that residence restrictions are effective at protecting residents from future sexual assaults committed by RSOs. As such, policymakers are advised against the continued passage or enforcement of residence restrictions without more research on the residence restrictions that are already in place. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on 23 References American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) (4th ed.). Washington, DC: American Psychiatric Publishing, Inc. Bachman, R., Paternoster, R., & Ward, S. (1992). The Rationality of Sexual Offending: Testing a Deterrence/Rational Choice Conception of Sexual Assault. Law & Society Review, 26(2), 343-372. Bagley, A. L. (2008). An Era of Human Zoning: Banishing Sex Offenders from Communities Through Residence and Work Restrictions. Emory Law Journal, 57(5), 1347-1392. Barnes, J. C. (2011). Place a moratorium on the passage of sex offender residence restriction laws. Criminology & Public Policy, 10(2), 401-409. Barnes, J. C., Dukes, T., Tewksbury, R., & De Troye, T. M. (2009). Analyzing the Impact of a Statewide Residence Restriction Law on South Carolina Sex Offenders. Criminal Justice Policy Review, 20(1), 21-43. Berenson, J. A., & Appelbaum, P. S. (in press). A Geospatial Analysis of the Impact of Sex Offender Residency Restrictions in Two New York Counties. Law and Human Behavior. Blood, P., Watson, L., & Stageberg, P. (2008). State Legislation Monitoring Report: FY2007: Iowa Department of Human Rights. Blumstein, A., Cohen, J., & Nagin, D. S. (Eds.). (1978). Deterrence and Incapacitation: Estimating the Effects of Criminal Sanctions on Crime Rates. Washington, DC: National Academy of Sciences. Bonnar-Kidd, K. K. (2010). Sexual Offender Laws and Prevention of Sexual Violence or Recidivism. American Journal of Public Health, 100(3), 412-419. Casady, T. (2009). A Police Chief's Viewpoint: Geographic Aspects of Sex Offender Residency Restrictions. Criminal Justice Policy Review, 20(1), 16-20. Chajewski, M., & Mercado, C. C. (2009). An Evaluation of Sex Offender Residency Restriction Functioning in Town, County, and City-Wide Jurisdictions. Criminal Justice Policy Review, 20(1), 44-61. Cohen, M., & Jeglic, E. L. (2007). Sex Offender Legislation in the United States: What Do We Know? International Journal of Offender Therapy and Comparative Criminology, 51(4), 369-383. 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. Council of State Governments. (2007, August). Out of Bounds: States Reconsider Sex Offender Management Through Residency Restrictions. State Government News, 50, 33-36. CSOM. (2007). Managing the Challenges of Sex Offender Reentry. Washington, DC: Center for Sex Offender Management. CSOM. (2008). Legislative Trends in Sex Offender Management: Center for Sex Offender Management. Duwe, G., Donnay, W., & Tewksbury, R. (2008). Does Residential Proximity Matter? A Geographic Analysis of Sex Offense Recidivism. Criminal Justice and Behavior, 35(4), 484-504. Fagan, P. J., Wise, T. N., Schmidt, C. W., & Berlin, F. S. (2002). Pedophilia. JAMA: The Journal of the American Medical Association, 288(19), 2458-2465. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on 24 Farkas, M. A., & Stichman, A. (2002). Sex Offender Laws: Can Treatment, Punishment, Incapacitation, and Public Safety Be Reconciled? Criminal Justice Review, 27(2), 256283. Freeman, N. J. (2007). Predictors of Rearrest for Rapists and Child Molesters on Probation. Criminal Justice and Behavior, 34(6), 752-768. Freeman, N. J., Sandler, J. C., & Socia, K. M. (2009). A Time-Series Analysis on the Impact of Sex Offender Registration and Community Notification on Plea Bargaining Rates. Criminal Justice Studies: A Critical Journal of Crime, Law, and Society, 22(2), 153-165. Greenfield, L. A. (1997). Sex Offenses and Offenders: An Analysis of Data on Rape and Sexual Assault (No. NCJ-163392). Washington: National Institute of Justice. Harris, A. J., & Lurigio, A. J. (2010). Introduction to Special Issue on Sex Offenses and Offenders: Toward Evidence-Based Public Policy. Criminal Justice and Behavior, 37(5), 477-481. Leon, C. (2011). The contexts and politics of evidence-based sex offender policy. Criminology & Public Policy, 10(2), 421-430. Letourneau, E. J., Levenson, J. S., Bandyopadhyay, D., Armstrong, K. S., & Sinha, D. (2010). Effects of South Carolina's Sex Offender Registration and Notification Policy on Deterrence of Adult Sex Crimes. Criminal Justice and Behavior, 37(5), 537-552. Levenson, J. S. (2008). Collateral Consequences of Sex Offender Residence Restrictions. Criminal Justice Studies, 21(2), 153-166. Levenson, J. S. (2009). Sex Offender Residence Restrictions. In R. G. Wright (Ed.), Sex Offender Laws: Failed Policies, New Directions (pp. 267-290). New York: Springer Publishing Company. 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(2), 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, 13(8), 168-199. Levenson, J. S., & Hern, A. L. (2007). Sex Offender Residence Restrictions: Unintended Consequences and Community Reentry. Justice Research and Policy, 9(1), 59-73. Levenson, J. S., & Tewksbury, R. (2009). Collateral Damage: Family Members of Registered Sex Offenders. American Journal of Criminal Justice, 34(1-2), 54-68. Marvell, T. B., & Moody, C. E. (2008). Can and Should Criminology Research Influence Policy? Suggestions for Time-Series Cross-Section Studies. Criminology and Public Policy, 7(3), 359-365. Matson, S., & Lieb, R. (1997). Megan's Law: A Review of State and Federal Legislation (No. 9710-1101). Olympia, WA: Washington State Institute for Public Policy. Meloy, M. L., Miller, S., & Curtis, K. (2008). Making Sense out of Nonsense: The Deconstruction of State-Level Sex Offender Residence Restrictions. American Journal of Criminal Justice, 33(2), 209-222. Minnesota Department of Corrections. (2003). Level Three Sex Offenders Residential Placement Issues. St Paul, MN. Minnesota Department of Corrections. (2007). Residential Proximity & Sex Offender Recidivism in Minnesota. St. Paul, MN. Nagin, D. S. (1998). Criminal Deterrence Research at the Outset of the Twenty-First Century. Crime and Justice, 23, 1-42. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on 25 New York State Division of Criminal Justice Services. (2010). Computerized Criminal History Database. Albany, NY. New York State Executive Law, Article 12B § 259-c(14). New York State Office of Sex Offender Management. (2010, December 3, 2009). NYS Sex Offender Residency Restriction Laws. Retrieved January 26, 2010, from http://criminaljustice.state.ny.us/nsor/pdfdocs/residencylaws.pdf New York State Penal Law, Title E § 65.10(4-a). Pawson, R. (2002). Does Megan's Law Work? A Theory-Driven Systematic Review: ESRC UK Centre for Evidence Based Policy and Practice. Pawson, R. (2006). Evidence-Based Policy: A Realist Perspective. London: SAGE Publications. Pratt, T. C., & Cullen, F. T. (2005). Assessing Macro-Level Predictors and Theories of Crime: A Meta-Analysis. Crime and Justice, 32, 373-449. Prescott, J. J., & Rockoff, J. E. (2011). Do Sex Offender Registration and Notification Laws Affect Criminal Behavior? Journal of Law and Economics, 54(1), 161-206. Pryor, D. W. (1999). Unspeakable Acts: Why Men Sexually Abuse Children. New York, NY: NYU Press. Robertiello, G., & Terry, K. J. (2007). Can We Profile Sex Offenders? A Review of Sex Offender Typologies. Aggression and Violent Behavior, 12(5), 508-518. Sandler, J. C., Freeman, N. J., & Socia, K. M. (2008). Does a Watched Pot Boil? A Time-Series Analysis of New York State's Sex Offender Registration and Notification Law. Psychology, Public Policy, and Law, 14(4), 284-302. Snyder, H. N. (2000). Sexual Assault of Young Children as Reported to Law Enforcement: Victim, Incident, and Offender Characteristics. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics. Socia, K. M. (2011a). The Policy Implications of Residence Restrictions on Sex Offender Housing in Upstate, NY. Criminology & Public Policy, 10(2), 351-389. Socia, K. M. (2011b). Residence restriction legislation, sex crime rates, and the spatial distribution of sex offender residences. Unpublished Ph.D. Dissertation, University at Albany, SUNY, Albany, NY. Socia, K. M., & Stamatel, J. P. (2010). Assumptions and Evidence Behind Sex Offender Laws: Registration, Community Notification, and Residency Restrictions. Sociology Compass, 4(1), 1-20. Soule, S. A., & Earl, J. (2001). The Enactment of State-Level Hate Crime Law in the United States: Intrastate and Interstate Factors. Sociological Perspectives, 44(3), 281-305. Terry, K. J., & Tallon, J. (2004). Child Sexual Abuse: A Review of the Literature. In John Jay College Research Team, The Nature and Scope of Sexual Abuse of Minors by Catholic Priests and Deacons in the United States, 1950-2002. Washington, DC: United States Conference of Catholic Bishops. 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., & Levenson, J. S. (2007). When Evidence Is Ignored: Residential Restrictions for Sex Offenders. Corrections Today, 69(6), 54-57. Tewksbury, R., & Mustaine, E. E. (2006). Where to Find Sex Offenders: An Examination of Residential Locations and Neighborhood Conditions. Criminal Justice Studies, 19(1), 6175. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on 26 D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on 27 Tewksbury, R., Mustaine, E. E., & Stengel, K. M. (2008). Examining Rates of Sexual Offenses from a Routine Activities Perspective. Victims & Offenders, 3(1), 75-85. U.S. Census. (2010). Population Estimates Program: U.S. Department of Commerce. Veysey, B., Zgoba, K., & Dalessandro, M. (2008). A Preliminary Step Towards Evaluating The Impact of Megan's Law: A Trend Analysis of Sexual Offenses in New Jersey from 1985 to 2005. Justice Research and Policy, 10(2), 1-18. Walker, J. T. (2007). Eliminate Residency Restrictions for Sex Offenders. Criminology & Public Policy, 6(4), 863-870. Walker, J. T., Golden, J. W., & VanHouton, A. C. (2001). The Geographic Link Between Sex Offenders and Potential Victims: A Routine Activities Approach. Justice Research and Policy, 3(2), 15-33. Washington State Institute for Public Policy. (2009). Does Sex Offender Registration and Notification Reduce Crime? A Systematic Review of the Research Literature (pp. 16). Olympia, WA. Welchans, S. (2005). Megan's Law: Evaluations of Sexual Offender Registries. Criminal Justice Policy Review, 16(2), 123-140. Willis, G. M., & Grace, R. C. (2008). The Quality of Community Reintegration Planning for Child Molesters, Effects on Sexual Recidivism. Sexual Abuse: A Journal of Research and Treatment, 20(2), 218-240. Youstin, T. J., & Nobles, M. R. (2009). Residency Restrictions: A Geospatial Analysis of Sex Offender Movement Over Time. Crime Mapping: A Journal of Research and Practice, 1(1), 55-76. 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(2), 1-24. Zandbergen, P. A., Levenson, J. S., & Hart, T. C. (2010). Residential Proximity to Schools and Daycares: An Empirical Analysis of Sex Offense Recidivism. Criminal Justice and Behavior, 37(5), 482-502. Zgoba, K. M. (2011). Residence restriction buffer zones and the banishment of sex offenders. Criminology & Public Policy, 10(2), 391-400. Zgoba, K. M., Levenson, J. S., & McKee, T. (2009). Examining the Impact of Sex Offender Residence Restrictions on Housing Availability. Criminal Justice Policy Review, 20(1), 91-110. Zgoba, K. M., Veysey, B. M., & Dalessandro, M. (2010). An Analysis of the Effectiveness of Community Notification and Registration: Do the Best Intentions Predict the Best Practices? Justice Quarterly, 27(5), 667 - 691. Table 1 Descriptive Statistics Variable Sex crime arrest rate: child victims, prior conviction overall between within Mean .11 SD .37 .12 .35 Min 0 0 -.44 0 Max 5 .56 4.56 0 Observations N = 8,928 n = 62 T = 144 D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on Sex crime arrest rate: overall .31 .79 N = 8,928 n = 62 adult victims, prior conviction between within .52 .01 3.22 .59 -2.91 0 6.09 38 T = 144 Sex crime arrest rate: overall 2.44 3.83 N = 8,928 n = 62 child victims, no prior conviction between within 3.39 .06 16.10 1.84 -8.95 0 24.34 141 T = 144 Sex crime arrest rate: overall 7.04 14.11 N = 8,928 n = 62 adult victims, no prior conviction between Within 13.37 4.82 .14 67.47 -33.43 0.00 80.57 1.00 T = 144 County residence restriction overall .07 0.26 N = 8,928 n = 62 between within 0.11 0.00 0.34 0.24 .03 -0.27 0 0.98 .52 T = 144 Lagged sex crime arrest rate: child victims, prior conviction overall .01 N = 8,966 n = 62 between within .01 0 0.02 .53 .03 -.01 T = 143 See note at end of table. 28 Table 1, Continued Variable Lagged sex crime arrest rate: adult victims, prior conviction Lagged sex crime arrest overall between within overall .12 Mean .01 SD .06 .01 .06 .17 Min 0 <.01 -.02 0 Max 1.93 .03 1.92 3.84 .24 Observations N = 8,966 n = 62 T = 143 N = 8,966 n = 62 D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on rate: child victims, no prior conviction between within .05 .03 .16 -.12 0 3.85 T = 143 Lagged sex crime arrest overall .25 .23 3.86 .43 N = 8,966 n = 62 rate: adult victims, no prior conviction between within .07 .10 .23 -.13 0 3.83 T = 143 Burglary arrest rate overall .63 .51 9.12 N = 8,966 N = 62 between within .17 .23 1.03 .48 -.30 0 8.91 T = 144 Robbery arrest rate overall .22 .32 2.23 N = 8,966 N = 62 between within .28 0 1.31 .17 -.42 .49 2.31 T = 144 Resident population (in 10,000s) overall 30.81 52.30 256.71 N = 8,928 N = 62 between within 52.67 2.17 .51 246.84 40.68 8.57 T = 144 Note. Monthly rates of sex crime arrests (and their associated lag measures) are calculated through the use of the resident population exposure term for each county-month. 29 Table 2 Associations Between County Residence Restrictions and Monthly Rates of Arrest for Different Types of Sex Crimes in New York Counties Recidivistic Sex Crime Variable (Incapacitation) Non-Recidivistic Sex Crime (General Deterrence) D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on Child Victim Adult Victim Child Victim Adult Victim .93 .91 .97 .90** (.03) County residence restriction (.23) .31 (.14) .47 (.05) .99 Lag 1.23*** (.05) (.37) .99 (.31) 1.04 (.08) 1.02 Burglary arrest rate 1.06** (.02) .95 (.10) 1.10 (.07) 1.26 (.03) .94 Robbery arrest rate (.23) .04* (.16) .87 (.04) .78 (.03) .69* Constant (.06) (.72) (.25) (.12) N 8,437 8,866 8,866 8,866 Note. All models include variables to control for temporal and county-specific trends. Coefficients are presented as odds-ratios, with robust standard errors provided in parentheses. An exposure term was included in each model representing the population (in 10,000 residents) in each county as of July 1st of each year. º p < .10; * p < .05; ** p < .01; *** p < .001 (two-tailed) 30 Appendix A County and Local-Level Residence Restrictions in New York State County County law Any local law County County law Any local law a Albany X Niagara X X Allegany Oneida X X Bronx Onondaga X Broome X X Ontario X Cattaraugus X Orange X X Cayuga X X Orleans Chautauqua X Oswego X Chemung Otsego X Chenango X Putnam X Clinton X Queens Columbia Rensselaer Xa Cortland Richmond Delaware Rockland Xa Dutchess (b) Saratoga X X Erie X Schenectady Xa Essex X Schoharie Franklin X X Schuyler Xa Fulton Seneca X X Genesee St Lawrence X Greene Steuben Hamilton Suffolk X X Herkimer Sullivan Jefferson X X Tioga X Kings Tompkins Lewis Ulster X Livingston Warren X Madison Washington X Monroe X Wayne X Montgomery Westchester X Nassau X X Wyoming X New York Yates Note. These data were adapted from NYS OSOM (2010) and Socia (2011b), and were current as of December 3, 2009. a No longer valid. b Dutchess County’s residence restriction requires RSOs to complete and sign a Sex Offender Verification Form, under oath, but does not restrict the locations where RSOs can live. As such, for the purposes of this study Dutchess County is not considered to have a county-level residence restriction. D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on 31 Notes This also assumes that the ‘suitable targets’ of RSOs are children who gather around public locations such as schools and daycares. While this assumption does not appear to be supported by existing literature (e.g. Duwe, Donnay, & Tewksbury, 2008; Zandbergen, Levenson, & Hart, 2010), contradicting this assumption or the use of routine activities theory to justify these policies is outside the scope of the current study (see Barnes, 2011; Tewksbury, Mustaine, & Stengel, 2008; Walker, Golden, & VanHouton, 2001). 2 For the purposes of this study, “local-level” refers only to sub-county jurisdictions such as towns and cities. 3 Note that a general deterrence effect is not possible when considering current RSOs who are already subject to the residence restriction, as residence restrictions no longer represent a threat of additional punishment for future sex crimes. Rather, in cases involving current RSOs, it is the actual effects of the residence restriction that would or would not result in incapacitation (not deterrence) of future sex crimes. 4 For a more detailed review of state residence restriction laws, see Meloy, et al. (2008) and Council of State Governments (2007). 5 Technically New York does have a state residence restriction that applies only to RSOs on probation or parole that are either designated high risk (i.e., level 3) or whose victims were under 18 at the time of offense (see New York State Executive Law §259-c(14); New York State Penal Law §65.10(4-a)). However, this restriction only prohibits offenders from knowingly entering school grounds, and while the definition of school grounds includes public areas (parks, parking lots, stores, etc.) and/or vehicles located within 1,000 feet of the boundary line of a school, it does not appear to include private residential housing located within this ‘buffer zone.’ As such, the law does not appear to be an actual residence restriction, but rather a movement and/or loitering restriction. 6 NYS OSOM (2010a) declares that the compilation of local residence restriction laws is for information purposes only, and the accuracy cannot be guaranteed. Nevertheless, it represents the most complete and thorough compilation of county and local laws available for the state. 7 While not examining residence restrictions specifically, three additional studies indicated that proximity to schools and/or daycares was unrelated to sexual recidivism (e.g., Colorado Department of Public Safety (CDPS), , 2004; MNDOC, , 2003; Zandbergen et al., 2010), which is counter to the underlying assumptions of such laws. In addition to finding that proximity to schools and daycares did not influence recidivism, Zandbergen and colleagues (2010) also examined buffer zones of multiple sizes around schools and daycares and determined that recidivists were no more likely to live within these buffer zones than non-recidivists. 8 This assumes in part that offenders with prior sexual convictions against adult victims are also subject to the residence restriction. In some instances, residence restrictions apply only to those individuals with prior sexual convictions against child victims, only to the highest risk sex offenders, and/or only to individuals still subject to probation or parole monitoring. For the purposes of this study, it is assumed that all RSOs whose information is available on a public registry (which typically excludes low risk and juvenile offenders) are subject to a residence restriction applying to their county of residence. This is consistent with the wording of the majority of county and local-level residence restrictions in New York (see NYS OSOM 2010). 9 The author would like to thank Dr. Jeffrey Sandler for raising this issue. 10 Note that in this study, ‘recidivistic sex crime arrests’ are defined as the number of arrests for sex crimes committed by individuals with prior convictions for sexual offenses (i.e., RSOs), while ‘recidivistic sex crime arrest rates’ are simply the rate of such arrests per a given number of residents. This is conceptually different from recidivism counts or rates, which involve the number or rate of individuals with prior convictions who commit additional sex crimes (measured as an arrest), rather than the number or rate of the occurrence of such crimes/arrests. While an alternative measure would have been the recidivism rate of the actual RSOs living within a given area, these historical data were not available. Therefore, the rate of recidivistic sex crime arrests for a given population of residents is expected to provide a reasonable measure of the prevalence of sex crimes. 11 These rates were measured first as crime arrest counts in the month the crimes were committed (or the month of the arrests if the crime commission date was unavailable for a given crime), and were then converted into sex crime arrest rates using an exposure term in the model based on annual population estimates (as of July 1st of each year) that came from the U.S. Census (2010) Population Estimates Program. Note that the measures used in this study did not include crimes that were reported to law enforcement but did not result in an arrest, as it was not possible to determine whether the perpetrator of the crime was an RSO or not, and therefore whether the crime was recidivistic or non-recidivistic in nature. Still, using arrest rates was expected to represent a better proxy measure of the actual rate of sex crime incidents compared to either reporting or conviction rates. Further, data on unreported sex crimes 1 D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on 32 D U o np no u t c bli ite sh ed w ith M ou an t p us er cr m ipt is si on 33 (such as data gathered by the National Crime Victimization Survey) were not available at the county level for New York. 12 A conviction for a prior registerable sex crime was taken as a proxy for inclusion on the state sex offender registry and for being subject to a residence restriction if one was in place in the county of residence. 13 As noted earlier, the majority of sex crimes involve acquaintances, especially for sex crimes involving child victims (Greenfield, 1997). While it would have been useful to further separate these four series of sex crime arrest rates by the relationship between the perpetrator and their victim (i.e., stranger, non-stranger), these data are only consistently available at the state level within individual case files. Given that over 350,000 individual sex crime arrests were included in the dataset, it was not feasible to obtain reliable acquaintance status for the present study. This should be kept in mind when comparing any effects of county or local-level restrictions between the various types of sex crime arrests. 14 The use of these control variables are largely consistent with the advice given by Marvell and Moody (2008) regarding the cross-sectional time series methods that can account for autocorrelation among the observations. 15 The author would like to thank Dr. David McDowall for his helpful advice regarding the setup of the model. 16 Other than the constant, the only control variables to reach significance were the lag sex crime arrest rate and the burglary rate in the final model for non-recidivistic sex crimes committed against adults (p < .001 and p < .01, respectively). While significant, these findings do not hold much importance for interpreting the effects of a residence restriction. 17 Note that not all child molesters are pedophiles, just as not all pedophiles are child molesters. However, in terms of likely crimes involving direct contact with victims, pedophiles are more likely to molest children than to rape adults due to their specific age preferences. 18 This explanation, however, does not account for rapes that are committed not out of sexual frustration, but rather as a means of power and/or control over the victim (see Robertiello & Terry, 2007).