Community Protection Policies and Repeat Sexual Offenses in Florida

Abstract

The purpose of this study was to investigate the impact of sexual offender management policies on sex crime repeat arrest rates in Florida. Aggregate data for the period 1990 to 2010 were provided by the Florida Department of Law Enforcement. The repeat offense rate was defined as the proportion of arrests each year that were committed by individuals with a previous conviction in the same crime category. The average yearly repeat offense rate for sex crimes was 6.5%, which was consistently and significantly lower than rates for other crimes: 8.3% for non-sex assaults, 15.1% for robbery, 29.8% for drug offenses, and 11.6% for DUI. The average annual sexual repeat arrest rate prior to and after the implementation of sexual offender registration laws in 1997 was 4.9% and 7.5%, respectively, indicating a statistically significant increase. The average annual repeat arrest rates for non-sex assaults, robberies, drug crimes, and DUIs also increased after 1997. No significant differences were found when comparing the average annual percent change for sexual re-arrest (+3.47%) with non-sexual assault (+3.93%), robbery (−.73%), drug offenses (+1.59%), and DUI (+1.14). Sex crime repeat arrests in Florida do not appear to show a decline attributable to sex offender management policies implemented since 1997.

Keywords

sex offender policy re-arrest SORN sex offender registration

Sexual violence is a serious social problem affecting American society. It is estimated that 18% of females are sexually assaulted sometime in their lives (Black et al., 2011) and 4 out of 1,000 children in the U.S. population are sexually abused in a given year (Finkelhor, Hammer, & Sedlak, 2008). Sexual crimes, especially those against children, produce profound and enduring negative psychosocial effects for victims and their families (Briere & Elliot, 2003; Dong, Anda, Dube, Giles, & Felitti, 2003; Dube et al., 2005). As well, the impacts of these offenses reverberate throughout communities, and it has been suggested that the trauma of sexual violence should be considered a public health issue (Centers for Disease Control and Prevention, 2004). Specialized sex offender legislation has been passed over the past 20 years in an effort to protect potential victims from recidivistic sexual assault through public awareness of sex offenders living among us and close monitoring of them by law enforcement agents (Levenson & D’Amora, 2007).

In 1997, Florida began to implement a range of policies to manage sex offenders living in the community. Between 1997 and 2007, these policies included a public Internet registry, mandatory minimum sentencing, special conditions of sex offender probation, civil commitment, state and local residence restrictions prohibiting sex offenders from living within close proximity to places where children congregate, and electronic monitoring. The current exploratory study examines whether changes in sexual repeat arrest trends in Florida can be detected over time, and whether sex offender policies have contributed to a significant drop in repeat sex crimes compared with other crimes. It was hypothesized that sex offender management policies would result in fewer repeat sex crimes committed by known sex offenders. It was expected that this research can inform policy development and promote evidence-based practices in sexual offender management.

Contemporary Sex Offender Policies

In 1994, the U.S. Congress passed the Jacob Wetterling Act, which required all 50 states to implement procedures to register the whereabouts of convicted sex offenders with law enforcement agencies. In 1996, the “Megan’s Law” amendment allowed states to make registry data available to the public. In 2006, Congress passed the Adam Walsh Act (AWA), which enhanced sex offender registration and notification (SORN) requirements and durations and provided guidelines to decrease variability in these laws across the nation. Florida was one of the first states to become compliant with the AWA in 2007.

According to the National Center for Missing and Exploited Children (NCMEC), in June 2014, approximately 774,600 registered sex offenders (RSOs) resided in the United States, with 63,544 in Florida. Nationally, about one third of RSOs are not found on public registries, presumably because they have been assessed by their state’s sex offender management procedure to pose a low risk for future offending (Ackerman, Harris, Levenson, & Zgoba, 2011). Among RSOs on public registries, about 14% nationally have been designated by states as high risk, predator, or sexually violent (Ackerman et al., 2011) and nearly 15% have more than one sex offense conviction. Approximately 90% of RSOs have had a minor victim, and about 33% have had victims less than 10 years old. Most (87%-89%) victims (adult and minor) are females (Ackerman et al., 2011; Finkelhor et al., 2008).

Sex offender laws have been motivated by the belief that the recidivism rates are exceedingly high (Fortney, Levenson, Brannon, & Baker, 2007; Meloy, Boatwright, & Curtis, 2012; Sample & Kadleck, 2008; Zgoba, 2004). Research, however, suggests that sex offense recidivism rates are lower than commonly believed (Bureau of Justice Statistics, 2003; Hanson & Morton-Bourgon, 2005; Harris & Hanson, 2004), and lower than recidivism rates in other crime categories (Bureau of Justice Statistics, 2002; Sample & Bray, 2006). In fact, recent longitudinal research indicates that sex offender recidivism risk decreases substantially over time as individuals live in the community sex offense–free; for example, after 16.5 years in the community without a new sex crime arrest, even high-risk sexual offenders are no more likely to be arrested for a new sex offense than non-sexual criminals (Hanson, Harris, Helmus, & Thornton, 2014; Harris & Hanson, 2012).

Despite widespread fear of sex offenders, the reality is that sex crime rates, like all crimes, have decreased over the past several decades (Finkelhor & Jones, 2006; Uniform Crime Report [UCR], 2012). Although the exact causes for the drop in crime are difficult to identify, experts suggest that a number of societal factors have contributed to the change. These include longer prison sentences, more aggressive policing, decreased opportunities for crime due to enhancements in technology and surveillance, the economic stability of the 1990s and early 2000s, changing demographics with the aging of the U.S. population, and other social dynamics related to norms, values, and informal social controls (Finkelhor & Jones, 2006; Uggen & McElrath, 2013; Zimring, 2006). Nonetheless, concerns about recidivistic sexual violence have driven the popularity of sex offender policies.

The purpose of SORN is ostensibly to improve public awareness about sex offenders living among us so that concerned citizens and parents can take preventive action to protect themselves and their children from sexual victimization (Anderson & Sample, 2008; Kernsmith, Comartin, Craun, & Kernsmith, 2009). In addition, sex offender registration systems provide a mechanism by which law enforcement agencies can track the whereabouts of these criminals and potentially identify a pool of local suspects when new crimes are committed. As attention to sex offenders living in our communities has grown, other policies aimed at preventing repeat victimization have been implemented, such as residential restrictions, enhanced sex offender supervision, and mandatory minimum sentencing (LaFond, 2005; Lamade, Gabriel, & Prentky, 2011; Levenson & D’Amora, 2007; Tabachnick & Klein, 2011; Zgoba, 2004). A central question is whether sex offender policies in general, and SORN in particular, have been successful in reducing sex offense recidivism.

Impact of Sex Offender Management Policies on Recidivism

At this time, the empirical research is still in a nascent stage due to the relatively recent implementation of these laws, and also due to methodological challenges faced by researchers when conducting sex crime policy analysis. For example, low base rates, the multiple criminal justice policies enacted within short time frames, challenges obtaining reliable recidivism data, and the need for long follow-up periods all contribute to the complexity of determining the impact of these laws (Levenson & D’Amora, 2007). Furthermore, each state’s SORN policy is idiosyncratic, subjecting different types of offenders to a variety of registration and notification requirements and complicating efforts to conduct national sex offender policy research (A. J. Harris, 2011).

Two studies that have detected reductions in sex crime recidivism as a result of SORN were conducted in Minnesota and Washington (Duwe & Donnay, 2008; Washington State Institute for Public Policy, 2005). In Minnesota, the recidivism rates of the notification group were significantly lower than both the pre-notification group (those matched on risk but released before the law went into effect) and the non-notification group (lower risk offenders not subject to disclosure; Duwe & Donnay, 2008). After controlling for generally decreasing crime trends, researchers in Washington found a significant decrease in sex offense recidivism (from 5% to 1%) after 1997 when SORN laws went into effect in that state (Washington State Institute for Public Policy, 2005). Although the authors acknowledged they were unable to account for other possible explanations for this reduction (e.g., more severe sentencing guidelines, or improved probationary supervision), they concluded that community notification has likely contributed to reductions in sexual reoffending. Noteworthy is that both states use empirically derived risk assessment procedures to classify offenders and tailor community management strategies to those who pose the greatest threat to community safety.

Most investigations, however, have not detected significant decreases in sex crime rates that can be attributed to SORN laws and policies. In South Carolina, 6,064 sex offenders convicted between 1990 and 2004 were tracked while controlling for time at risk, and registration status did not predict sexual recidivism in any model (Letourneau, Levenson, Bandyopadhyay, Sinha, & Armstrong, 2010). A trend analysis in New York cast doubt on SORN’s ability to reduce sexual offending, and no declines were found for child molester, rapists, recidivists, or first-time sex offenders (Sandler, Freeman, & Socia, 2008). In New Jersey, researchers compared 250 sex offenders released before Megan’s Law went into effect with 300 sex offenders released after the passage of the law, and found no significant differences in sex offense recidivism, the time it took for sex offenders to reoffend, or the number of victims (Zgoba, Veysey, & Dalessandro, 2010). The authors followed up with a trend analysis, and though they found a significant decrease in sexual offense recidivism aggregated across counties, they cautioned that variations in jurisdiction rates precluded a conclusion that reductions were attributable to Megan’s Law implementation (Veysey, Zgoba, & Dalessandro, 2008). Although SORN status was not a significant predictor of sexual recidivism in New Jersey, high-risk sex offenders were more likely to commit future criminal offenses, including sex offenses, and to do so fairly quickly following release, suggesting that empirically based risk assessment is a valuable component of sex offender management (Tewksbury, Jennings, & Zgoba, 2012).

Two recent reports raised questions about the utility and validity of the AWA tier categories in Florida, New Jersey, Minnesota, and South Carolina (Zgoba et al., 2012) and New York (Freeman & Sandler, 2010). Both studies concluded that the AWA offense-based classification system did a poor job of identifying potential recidivists and that empirically grounded procedures were better for screening sex offenders into relative risk categories to establish monitoring requirements.

Multistate studies have produced mixed findings because variability in both research methodologies and SORN policy characteristics can contribute to differing results reported across studies. A time-series analysis investigated the impact of SORN laws on sexual assault rates in 10 states (Vasquez, Maddan, & Walker, 2008), finding that California had a significant increase in rape rates following implementation of registration, whereas Hawaii, Idaho, and Ohio had significant decreases in rape rates; the remaining 6 states (Arkansas, Connecticut, Nebraska, Nevada, Oklahoma, & West Virginia) showed non-significant trends. The authors concluded that SORN policies did not appear to systematically reduce sex crime rates. A large-scale analysis examining more than 300,000 sex offenses in 15 states found that whereas registration appeared to diminish the rate of recidivistic sex offenses, public notification did not (Prescott & Rockoff, 2011). Using UCR data from 1985 to 2003, Agan (2011) did not find a significant decrease in arrest rates of rape or sexual abuse after registration was implemented or after giving the public access to the registry via the Internet. Agan also examined BJS data that tracked individual sex offenders after their release in 1994, and determined that having to register as a sex offender did not lead to significant reductions in sex offense recidivism (Agan, 2011). Using UCR data for the years 1970 to 2002, Ackerman, Sacks, and Greenberg (2012) investigated the effect of SORN legislation and reported that these laws have not resulted in dramatic declines in forcible rapes. Several scholars have concurred that the accumulation of empirical evidence suggests that the costs outweigh the benefits of SORN laws (Ackerman et al., 2012; Zgoba, Witt, Dalessandro, & Veysey, 2009).

As knowledge about the whereabouts of sex offenders has become widespread, citizens have also demanded legislation restricting sex offenders from living within close proximity to locations such as schools, parks, or day care centers. Residence restrictions are now found in at least 30 states (Meloy, Miller, & Curtis, 2008) and thousands of municipalities, despite evidence suggesting that such laws are not associated with prevention of repeat sexual violence (Colombino, Mercado, Levenson, & Jeglic, 2011; Nobles, Levenson, & Youstin, 2012; Zandbergen, Levenson, & Hart, 2010). In fact, zoning sex offenders out of densely populated areas results in greatly diminished housing options (Zandbergen & Hart, 2006, 2009; Zgoba, Levenson, & McKee, 2009) and increases the likelihood of transience and homelessness (Levenson, Ackerman, Socia, & Harris, 2014). SORN laws and residence restrictions impede successful reintegration by creating barriers to employment, stable housing, and social support; they can even exacerbate the psychosocial stressors that increase dynamic risk for recidivism (Levenson & Cotter, 2005a, 2005b; Mercado, Alvarez, & Levenson, 2008; Tewksbury & Lees, 2006; Tewksbury & Mustaine, 2009).

In summary, the abundance of evidence related to sex offender policy effectiveness does not point to reduced recidivism as a result of these laws, and the few studies finding reductions attributable to SORN were conducted in states using empirically derived risk classification systems and targeted notification strategies. The stigma of sex offender registration can also create significant obstacles to successful community re-entry. On the other hand, some scholars have opined that although sex offender policies may not demonstrate shining success in accomplishing explicit objectives such as reduced recidivism, they do accomplish other important symbolic goals by sending a clear message that sexual victimization will not be tolerated and that politicians are willing to address public concerns (Sample, Evans, & Anderson, 2011; Sample & Kadleck, 2008). It is important, however, for policy makers and stakeholders to understand the effects of these laws so that decisions can be informed by evidence. The purpose of this study was to explore whether sex offender management policies were associated with changes in sexual repeat arrest trends over time in Florida.

Method

Data

The data utilized for these analyses included annual aggregated crime counts for five offenses in Florida for each year from 1990 through 2010. These data included arrests for felony sex offenses, non-sexual assaults, robberies, drug offenses, and DUI (Driving under the Influence) offenses. The comparative crimes were included because factors other than sex offender–specific policies may be responsible for any potential changes. Although it was not possible to control for other variables based on our data, the comparison crimes allowed for exploration of trends that might be unique to sex offenders. Non-sexual assault and robbery were chosen because they are serious felonies that can cause severe physical and emotional harm to victims. Drug and DUI crimes were chosen because the “War on Drugs” is somewhat analogous to the increased public attention and sanctioning of sex crimes over the past several decades.

Traditional recidivism studies entail tracking individual offenders over time and obtaining their criminal records to determine new arrests and risk factors. Limited resources precluded this type of study. As an alternative, aggregate yearly arrest counts were requested from the Florida Department of Law Enforcement (FDLE) via a specified statute list with crime codes. FDLE provided official yearly aggregate reports of crime utilizing the state’s Computerized Criminal Histories (CCH), which are a fingerprint-based record of crime. FDLE receives these criminal records from contributing law enforcement agencies. CCH data provided counts of arrests per year by statute, and when multiple charges existed for one arrest event, the most serious was selected by FDLE. All data were de-identified by FDLE and were provided and analyzed in aggregate form.

Because the data were requested and provided in aggregate form, we were unable to conduct a traditional recidivism study that tracked a specific cohort of offenders for a determined period of time after their release. Therefore, we defined our “recidivism” rate in a unique fashion: FDLE provided, for each crime and each year, a count of how many of the arrests were attributable to an offender who had a previous conviction for an offense within the same crime category. These were divided by the annual arrest count totals in each category and that proportion was defined as the aggregate repeat arrest rate.

Variables and Data Analysis

The variables provided by the FDLE were comprised of total yearly counts of arrests and repeat arrests as defined above for each felony crime category for the time frame 1990 to 2010. Additional variables were then created for each crime category: indicators of pre- and post-policy periods, repeat arrest rates, and annual percentage changes for each crime category. As described above, repeat arrest rates were calculated by determining the proportion of arrests each year committed by individuals with a previous conviction in the same crime category. Because low base rate crimes can produce statistical significance for small changes, the data were examined using multiple techniques.

The annual percentage change in repeat arrest rates across the time frame for the five crimes was calculated. The percentage change in crime is commonly used to measure incremental changes over time by comparing data from one time period (e.g., 1 year) with data reported in a prior equivalent period (UCR, 2012). The annual percent change is computed by subtracting one yearly count from the following yearly count, then dividing the difference by the count from the earlier year and multiplying the quotient by 100. For each crime, these yearly changes were then added up and divided by the number of year-to-year changes (20) to calculate the average annual percent change for each crime category.

The “intervention point” was determined to be 1997, the year Florida implemented its public sex offender registry as well as its special conditions of sex offender probation. Bivariate analyses were conducted to compare differences between the pre- and post-policy periods, and to compare trends between crime categories. The pre-implementation time period was 1990 to 1997, and for some analyses the post-implementation period was truncated to 8 years (1998-2005) to produce an equivalent time frame. The full time frame extending the post-implementation period to 2010 was also examined.

Results

Florida Arrest Trends

Florida’s violent crime UCR reports from 1993 to 2010 for forcible sex offenses, robbery, and aggravated assault were publicly available and were downloaded and examined for general arrest trends over the time period (Florida Statistical Analysis Center, 2014). UCR data reflect broader categories of crime and therefore approximate, but do not match exactly, the CCH data provided by FDLE. The annual count in each category was divided by Florida’s population and then multiplied by 100,000 to determine the rate of each crime type per 100,000 people in the population. This technique controls for population growth. As seen in Figure 1, all three violent crime rates decreased by about half over the time period, and these declining trends began prior to the implementation of sex offender management policies in 1997.

Figure 1.

Florida UCR rates per 100,000 population 1993-2010.

Annual raw counts of arrests for sex crimes, non-sexual assaults, robberies, drug offenses, and DUIs were provided by FDLE. While the rate of crime in the population decreased over the time period, the raw number of arrests in each category increased each year with the exception of robbery (see Figure 2). So, crime declined in relation to population growth, even as police were arresting more individuals. Between 1990 and 2010, the yearly number of sex crime arrests in Florida increased from 2,377 to 3,946, with the highest level (5,012 arrests) in 2005. The yearly number of non-sexual assault arrests increased from 15,716 to 32,848. The number of robbery arrests decreased from 11,922 to 11,416, and the number of drug arrests increased from 37,620 to 64,295 with the highest peak in 2007. The number of DUI arrests increased from 47,016 to 54,571.

Figure 2.

Annual raw counts of arrests in each crime category 1990-2010 (FDLE CCH data).

Florida Repeat Arrest Trends, 1990-2010

Figure 3 illustrates the repeat arrest rates for the crime categories included in this analysis between 1990 and 2010. Repeat offender rates for all crimes were calculated by determining the proportion of arrests attributed to unique individuals previously convicted of a crime in the same category. Average yearly sex crime repeat arrest rates were consistently lower than other crimes. There were 78,649 sex crime arrests over the full period, and the average yearly sexual reoffense rate for the full period was 6.5%. In other words, on average, each year, 6.5% of the sex crime arrests were perpetrated by an individual with a previous conviction for a felony sex crime.

Figure 3.

Florida repeat offender rates: 1990-2010 (FDLE CCH data).

Over the full follow-up period, there were a total of 565,325 felony non-sex assaults, with an average repeat arrest rate of 8.3%. There were 230,876 felony robberies, with an average repeat arrest rate of 15.1%. There were more than 1.2 million drug arrests, with an average repeat arrest rate of 29.8%, and approximately 1.2 million DUI arrests, with an average repeat arrest rate of 11.6%.

The average yearly repeat arrest rates over the time period are also listed in Table 1. Drug crimes had the highest rates of repeat arrest, with a level almost five times as high as sexual reoffending. Felony sex crime arrests represented the least prevalent crime and the lowest repeat arrest rate among the five types of crime included in this analysis. The sex crime repeat arrest rate was statistically significantly lower than the other types of crime included in the analyses (p < .001).

Table 1.

Annual Average Repeat Offender Rates Pre- and Post 1997.

Crime category	Average annual repeat offender rate (1990-2010)	Repeat rate pre-1997	Repeat rate post 1997^a	Mean difference pre- and post 1997 (truncated) (t test)	Mean difference pre- and post 1997 (not truncated) (t test)
Sex offense	6.5	4.9	7.5	+2.5%***	+2.5***
Non-sexual assault	8.3	6	9	+3.0%***	+3.7***
Robbery	15.1	15.3	16.1	+0.75%	+0.35
Drug	29.8	27	31	+3.9%***	+4.4***
DUI	11.6	10.2	12.5	+2.3%***	+2.3***

Time period was truncated (1998-2005) to render the pre– and post–time frames equivalent. The non-truncated period was 1998-2010.

***

p < .001.

To determine whether sex offender community protection policies affected sexual reoffending, the average repeat arrest rates for each crime category during the pre and post (after 1997) intervention periods were compared using t tests. To calculate statistical significance, the post-policy period was truncated and analyzed from 1998 to 2005 so that the pre- and post-periods were equivalent. Because multiple t tests can lead to spurious findings, the Bonferroni correction technique was used to test the statistical significance of multiple comparisons. This method involves adjusting the significance threshold for rejecting the null hypothesis and is calculated by dividing the alpha level by the number of comparisons (Vogt, 2005). In this case, the significance level, p < .05, was divided by 5, which reflected the number of items that were tested for each group comparison (.05 / 5 = .01). This more conservative threshold was used to determine whether changes were statistically significant (see Table 1).

Average repeat arrest rates significantly increased in all categories after 1997, with the exception of robbery, which displayed a slight increase that was not statistically significant. These analyses were also conducted for the non-truncated post-policy period (1998-2010) and very similar repeat arrest rates were noted. A slightly greater average increase in rates was seen in the full post-policy period for non-sexual assault (3.7%) and drug offenses (4.4%), whereas robbery rates declined slightly (0.35%). Again, all of the changes in the non-truncated period were statistically significant except for robbery (see Table 1).

Annual Average Percent Change

In addition to the comparisons of average yearly repeat arrest rates, the percentage change in these rates from year to year was examined to determine whether trends in the non-sexual crime categories were significantly different from sexual crime repeat arrest patterns. Because crime counts can go up or down in a given year for a variety of reasons while still showing a particular trend over the longitudinal time frame, the average annual percentage change was calculated in addition to the mean differences in rates described above. The computational strategy described in the “Method” section is illustrated using the following example: In 1990, the data provided by FDLE reported 2,377 arrests for sex crimes, 102 of which involved an individual with a prior conviction for a sex crime (4.3%). In 1991, out of 2,480 sex crime arrests, 121 were repeat offenders (4.9%). This change amounts to a raw increase in repeat offenders of 0.6% from 1990 to 1991 (4.9 − 4.3). However, the change in the proportion of repeat offenders represents a 14% increase ([4.9 − 4.3] / 4.3 × 100). The first percent change calculation reflects the change from 1990-1991, so 1991 becomes the first year in the pre-intervention time period.

The average annual percent change in repeat arrests for each crime category was compared for the pre- (1991-1997) and post- (1998-2010) periods (see Table 2). Repeat arrests continued to increase each year in the post-intervention period in all crime categories except for robbery, but were more modest than in the pre-intervention period; a two-tailed t test was used to compare annual average percentage changes in the pre- and post-periods for each crime category and only drug crime was statistically significant (+.74). When the post-period was truncated (1998-2005), no statistically significant differences were found.

Table 2.

Average Annual Percent Change in Repeat Arrests Pre- and Post-SORN.

		M % change	SD	t	Significance
Sex offense	PRE	6.4	20.3	0.545	.601
Sex offense	POST 1997	1.9	11.2	0.545	.601
Non-sexual assault	PRE	4.0	6.9	0.018	.986
Non-sexual assault	POST 1997	3.9	3.9	0.018	.986
Robbery	PRE	−3.3	5.3	−1.201	.245
Robbery	POST 1997	0.67	7.8	−1.201	.245
Drug	PRE	3.2	3.4	2.146	.046
Drug	POST 1997	0.74	1.7	2.146	.046
DUI	PRE	2.6	6.1	0.934	.380
DUI	POST 1997	0.34	2.8	0.934	.380

Note. The pre– and post–time frames were (1991-1997) and (1998-2010), respectively. SORN = sex offender registration and notification.

Finally, the annual average percent change in sex crime repeat arrests over the entire time frame (1991-2010), an increase of 3.47%, was compared with the annual average percent change in the other crime categories (see Table 3). Again, using a two-tailed t test, no statistically significant differences in trends were noted.

Table 3.

Annual Average Percent Change in Sex Crime Repeat Arrests Compared With Other Crimes (1991-2010).

Sex crime (+3.47%) compared with:	M % change	t	df	Significance
Non-sexual assault	+3.93	−1.684	19	.109
Robbery	−0.73	0.764	19	.455
Drug	+1.59	−1.062	19	.301
DUI	+1.14	−0.268	19	.792

Discussion

The hypothesis that sex offender management policies in Florida would result in declining sexual re-arrest rates was not supported. There are three primary observations that can be seen in these data. First, sex crime repeat arrest rates (the proportion of all sex crime arrests that belong to individuals with a prior conviction for a sex offense) are substantially lower than commonly believed, hovering on average around 6.5% over the 20-year period. On average, in Florida, less than 7% of sex crime arrests each year were committed by repeat sex offenders. Second, sexual repeat arrest rates were consistently lower than those of other felony crimes that also result in serious injury to victims and which have high fiscal and social costs for communities. Third, we did not find evidence that the panoply of policies passed since 1997 to track, monitor, and restrict known sex offenders (e.g., registration and public notification, residential restrictions) is associated with a reduced rate of sexual re-arrest compared with other crimes. In fact, the trend indicates a modest but statistically significant increase in sex crime average annual re-arrest rates over time. Because the trend is in the same direction as non-sexual assault, drug, and DUI crimes, it is likely attributable to something other than sex offender–specific policies.

When conducting analyses to determine the impact of a set of policies on sexual reoffending, there are a number of possible findings that may be observed as well as a variety of potential explanations for those observations (Agan, 2011). First, we might not observe any significant change in trends at all, suggesting that the intervention has had no effect on criminal behavior and concluding that the policies have not achieved their explicit goal. Second, we might observe a decrease in reoffending, suggesting that there has indeed been a change, but the specific explanation for the decrease would be difficult to ascertain. For instance, because multiple community protection policies were implemented over the time span, it would be difficult to tease out the effect of any or each single policy, or even to determine the effects of the combination of policies. Further analyses would be required to attribute the outcome to a particular policy while controlling for a number of other factors, such as social or cultural variables or systemic changes in the procedures of law enforcement agencies or courts. Finally, we might observe an increase in reoffending, suggesting that a change has occurred in the opposite direction of what was expected, and calling for some conjecture about the factors that have led to this result. In the current study, it is this final outcome that was observed.

According to the FBI, crime rates in general, adjusted to population changes, have declined over the past two decades, with violent crime decreasing by about 49%, although violent crimes increased slightly in 2012 (UCR, 2012). It is interesting that though this general trend is seen in Florida, repeat arrest proportions in Florida increased across categories over the study period. Thus, although the rate of crime adjusted for population growth in all five categories went down, the percentage of arrests comprised of repeat offenders went up.

We speculate that it is probable that this finding does not reflect a true increase in reoffending, but rather an increased likelihood of re-arrest, due to enhanced detection, reporting, and/or arresting practices. Criminal offenders in Florida are serving increasingly longer probationary periods, and technology has provided supervising officers with improved surveillance techniques. As well, law enforcement strategies have been enhanced over the past two decades, resulting in improved investigative methods, and on a societal level, there is less tolerance for criminal behavior. Thus, we might expect that when known offenders (especially those on probation) do engage in criminal activities, they are more likely to be reported, caught, arrested, and punished. While crime rates in the overall population may be declining in Florida, repeat offenders might be more likely to be detected and held accountable.

Although all repeat arrest rates increased over time, sex offenders—who are monitored even more closely than other offenders and beyond the end of formal probationary supervision—may be more likely to be reported to authorities and arrested when allegations of new sex crimes are suspected. Indeed, some might interpret the increase in sexual re-arrest rates to be a desirable outcome and a signal of success of SORN laws, indicating that sex offenders who repeat their crimes are more likely to be detected than they used to be. Furthermore, one could argue that SORN has had a general deterrent effect because the percentage of arrestees who committed a first-time sex offense declined after 1997. In other words, as the proportion of repeat arrests increased, the proportion of first-time arrests dropped accordingly, perhaps suggesting a small general deterrence impact.

The percentage change in repeat arrest rates over the post-policy period was higher for sex offenders than for robbers, drug offenders, and DUI offenders (but not statistically significant). A possible explanation is that the re-entry obstacles faced by sexual offender registration increase dynamic risk for reoffense. It is well established that the stigma of registration creates barriers to meaningful employment, stable housing, and social support systems (Jeglic, Mercado, & Levenson, 2011; Levenson & Cotter, 2005a; Mercado et al., 2008; Tewksbury & Mustaine, 2009). Some scholars have cautioned that these obstacles might actually increase risk that sex offenders will resume a life of crime or reoffend sexually (Kruttschnitt, Uggen, & Shelton, 2000; Tabachnick & Klein, 2011; Tewksbury & Mustaine, 2009; Willis & Grace, 2009). Criminal offenders in general are most likely to successfully reintegrate when they receive support in the community and are able to engage in prosocial activities that promote conformity to the values of society (Maruna, 2001; Maruna, LeBel, Mitchell, & Naples, 2004; Petersilia, 2003; Uggen, Manza, & Behrens, 2004). Thus, a higher level of scrutiny compounded by the psychosocial stressors of the collateral consequences of registration may increase the chance of re-arrest for known sexual offenders in a way that is distinct from other offenders.

Implications for Policy

The purposes of SORN are twofold: to increase public awareness of sex offenders’ whereabouts and to allow law enforcement agencies to track and monitor these offenders more closely. The ultimate goal of this rigorous monitoring is to prevent recidivistic sex crimes by decreasing opportunities for victimization. Florida has one of the most comprehensive and strict registration systems in the country, adhering to the federal guidelines of the AWA and publicly displaying all sex offenders via an online registry for 25 years to life. It does not appear that sex offender–specific surveillance policies resulted in reduction of recidivism, even though they may have increased the likelihood that sexual reoffenders will come to the attention of authorities.

In the absence of clear evidence that these policies reduce recidivism, other alternatives for sex offender management might be considered and pursued. For instance, sex offense recidivism risk significantly declines as RSOs spend time in the community offense-free (Hanson et al., 2014). Since a small minority of sex crimes is recidivistic, lifetime or 25-year registration durations may be unnecessary for most registrants. Sex offender management strategies may be more effective and cost-efficient when utilizing empirically derived risk assessments to target those at highest risk to reoffend. In this way, resources might be better directed toward more dangerous sex offenders, and obstacles to reintegration can be minimized for lower risk offenders.

This study, while unique in its design, yields results similar to other published research suggesting that reduced sexual recidivism is not a significant effect of sex crime legislation (Ackerman et al., 2012; Agan, 2011; Letourneau et al., 2010; Prescott & Rockoff, 2011; Vasquez et al., 2008; Zgoba et al., 2010). Sample et al. (2011) cautioned that in the minds of politicians and the public, the symbolic expression of zero tolerance for sexual violence will likely outweigh fiscal considerations, concerns for offender rights, and empirical testing. In the end, however, policies should be informed by scientific evidence.

Limitations

One limitation of using official arrest data in research is that undetected or unreported offenses cannot be accounted for. Research suggests that many victims of sexual assault are reluctant to report sex crimes for a variety of reasons (BJS, 2010) and that about 30% of child sexual abuse incidents are reported to authorities (Finkelhor et al., 2008). On the other hand, because they are so closely scrutinized, RSOs may be more likely than first timers who commit sex crimes to be discovered or reported. Although we cannot account for the precise number of offenses that go unreported, we might assume that the proportion of undetected offenses has remained fairly constant over time (or decreased with increased public awareness about sexual assault). Thus, when examining trends, unreported crimes should not affect our ability to detect changes in repeat arrest rates over time.

Our method of defining repeat arrest was admittedly unconventional and differed from the typical approach of studying recidivism by following a criminal cohort over time to track new arrests. We were also unable to account for “pseudo-recidivism” in which offenders are arrested after an index offense for a crime that actually occurred before the index offense (usually due to delayed reporting by victims). Thus, our reoffending estimates should not be construed as true recidivism rates and were devised simply to offer an examination of average yearly repeat arrest proportions over time and to compare the proportion of repeat offenders in various crime categories. Because the data were provided and analyzed in aggregate form, we had no way of distinguishing individuals with specific offense characteristics, such as predators versus offenders, child abusers versus those with adult victims, offenders with male versus female victims, or low versus high risk.

Summary and Conclusion

In conclusion, sex crime repeat arrest rates in Florida did not appear to show a decline after SORN laws or other policies were implemented around 1997. Repeat arrest rates increased in the post-policy period for all crime categories. It is possible that the enhanced supervision of criminal offenders in general and sex offenders specifically has made it more likely that their recidivistic offenses will be discovered, resulting in increases in repeat offender arrests. Absent a statistically significant reduction in repeat sex crime arrests after the introduction of sex offender management laws, we cannot conclude that sex offender policies have achieved the goal of preventing sexual reoffending in Florida.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Seed funding was provided by an attorney for the purpose of obtaining initial data, assessing the feasibility of a study, conducting preliminary analyses, and drafting a confidential work product report. Thereafter, the data were re-analyzed for the purpose of developing and preparing this article, and these activities were unfunded. The opinions or points of view expressed are solely those of the authors.

References

Ackerman

A. R.

Harris

A. J.

Levenson

J. S.

Zgoba

(2011). Who are the people in your neighborhood? A descriptive analysis of individuals on public sex offender registries. International Journal of Law and Psychiatry, 34, 149-159.

Ackerman

A. R.

Sacks

Greenberg

D. F.

(2012). Legislation targeting sex offenders: Are recent policies effective in reducing rape? Justice Quarterly, 29, 858-887.

Agan

A. Y.

(2011). Sex offender registries: Fear without function? Journal of Law & Economics, 54, 207-239.

Anderson

A. L.

Sample

(2008). Public awareness and action resulting from sex offender community notification laws. Criminal Justice Policy Review, 19, 371-396.

Black

M. C.

Basile

K. C.

Breiding

M. J.

Smith

S. G.

Walters

M. L.

Merrick

M. T.

. . . Stevens

M. R.

(2011). National intimate partner and sexual violence survey. Atlanta, GA: Centers for Disease Control and Prevention.

Briere

Elliot

D. M.

(2003). Prevalence and psychological sequelae of self-reported childhood physical and sexual abuse in a general population sample of men and women. Child Abuse & Neglect, 27, 1205-1222.

Bureau of Justice Statistics. (2002). Recidivism of prisoners released in 1994. Washington, DC: U.S. Department of Justice.

Bureau of Justice Statistics. (2003). Recidivism of sex offenders released from prison in 1994. Washington, DC: U.S. Department of Justice.

Bureau of Justice Statistics. (2010). National Crime Victimization Survey. Washington, DC: U.S. Department of Justice.

10.

Centers for Disease Control and Prevention. (2004). Sexual violence prevention: Beginning the dialogue. Atlanta, GA: Author.

11.

Colombino

Mercado

C. C.

Levenson

J. S.

Jeglic

E. L.

(2011). Preventing sexual violence: Can examination of offense location inform sex crime policy? International Journal of Law and Psychiatry, 34(3), 160-167.

12.

Dong

Anda

R. F.

Dube

S. R.

Giles

W. H.

Felitti

V. J.

(2003). The relationship of exposure to childhood sexual abuse to other forms of abuse, neglect, and household dysfunction during childhood. Child Abuse & Neglect, 27, 625-639.

13.

Dube

S. R.

Anda

R. F.

Whitfield

C. L.

Brown

D. W.

Felitti

V. J.

Dong

Giles

W. H.

(2005). Long-term consequences of childhood sexual abuse by gender of victim. American journal of preventive medicine, 28, 430-438.

14.

Duwe

Donnay

(2008). The impact of Megan’s Law on sex offender recidivism: The Minnesota experience. Criminology, 46, 411-446.

15.

Finkelhor

Hammer

Sedlak

A. J.

(2008). Sexually assaulted children: National estimates and characteristics. Washington, DC: U.S. Department of Justice, Office of Juvenile Justice and Delinquency Prevention.

16.

Finkelhor

Jones

(2006). Why have child maltreatment and child victimization declined? Journal of Social Issues, 62, 685-716.

17.

Florida Statistical Analysis Center. (2014). Crime in Florida: Florida uniform crime report (1993-2013). Retrieved from http://www.fdle.state.fl.us/Content/FSAC/Menu/Crime-Trends/Total-Index-Crimes.aspx

18.

Fortney

Levenson

J. S.

Brannon

Baker

(2007). Myths and facts about sex offenders: Implications for practice and public policy. Sex Offender Treatment, 2(1), 1-17.

19.

Freeman

N. J.

Sandler

J. C.

(2010). The Adam Walsh Act: A false sense of security or an effective public policy initiative? Criminal Justice Policy Review, 21(1), 31-49.

20.

Hanson

R. K.

Harris

A. J.

Helmus

Thornton

(2014). High-risk sex offenders may not be high risk forever. Journal of Interpersonal Violence. doi:0886260514526062

21.

Hanson

R. K.

Morton-Bourgon

(2005). The characteristics of persistent sexual offenders: A meta-analysis of recidivism studies. Journal of Consulting and Clinical Psychology, 73, 1154-1163.

22.

Harris

A. J.

(2011). SORNA in the post-deadline era: What’s the next move? Sex Offender Law Report, 12(6), 81-86.

23.

Harris

A. J. R.

Hanson

R. K.

(2004). Sex offender recidivism: A simple question. Ottawa, Ontario: Public Safety and Emergency Preparedness Canada.

24.

Harris

A. J. R.

Hanson

R. K.

(2012, October). When is a sex offender no longer a sex offender? Paper presented at the 31st Annual Research and Treatment Conference of the Association for the Treatment of Sexual Abusers, Denver, CO.

25.

Jeglic

Mercado

C. C.

Levenson

J. S.

(2011). The prevalence and correlates of depression and hopelessness among sex offenders subject to community notification and residence restriction legislation. Journal of Criminal Justice, 37(1), 46-59.

26.

Kernsmith

P. D.

Comartin

Craun

S. W.

Kernsmith

R. M.

(2009). The relationship between sex offender registry utilization and awareness. Sexual Abuse: A Journal of Research and Treatment, 21, 181-193.

27.

Kruttschnitt

Uggen

Shelton

(2000). Predictors of desistance among sex offenders: The interaction of formal and informal social controls. Justice Quarterly, 17(1), 61-88.

28.

LaFond

J. Q.

(2005). Preventing sexual violence: How society should cope with sex offenders. Washington, DC: American Psychological Association.

29.

Lamade

Gabriel

Prentky

(2011). Optimizing risk mitigation in management of sexual offenders: A structural model. International Journal of Law and Psychiatry, 34, 217-225.

30.

Letourneau

Levenson

J. S.

Bandyopadhyay

Sinha

Armstrong

(2010). Effects of South Carolina’s sex offender registration and notification policy on adult recidivism. Criminal Justice Policy Review, 21, 435-458.

31.

Levenson

J. S.

Ackerman

A. R.

Socia

K. M.

Harris

A. J.

(2014). Transient sex offenders and residence restrictions. Criminal Justice Policy Review. doi:10.1177/0887403413512326

32.

Levenson

J. S.

Cotter

L. P.

(2005a). The effect of Megan’s Law on sex offender reintegration. Journal of Contemporary Criminal Justice, 21(1), 49-66.

33.

Levenson

J. S.

Cotter

L. P.

(2005b). 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.

34.

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.

35.

Maruna

(2001). Making good: How ex-convicts reform and rebuild their lives. Washington DC: American Psychological Association.

36.

Maruna

LeBel

T. P.

Mitchell

Naples

(2004). Pygmalion in the reintegration process: Desistance from crime through the looking glass. Psychology, Crime & Law, 10, 271-281.

37.

Meloy

Boatwright

Curtis

(2012). Views from the top and bottom: Lawmakers and practitioners discuss sex offender laws. American Journal of Criminal Justice, 1-23.

38.

Meloy

Miller

S. L.

Curtis

K. M.

(2008). Making sense out of nonsense: The deconstruction of state-level sex offender residence restrictions. American Journal of Criminal Justice, 33, 209-222.

39.

Mercado

C. C.

Alvarez

Levenson

J. S.

(2008). The impact of specialized sex offender legislation on community re-entry. Sexual Abuse: A Journal of Research & Treatment, 20, 188-205.

40.

Nobles

M. R.

Levenson

J. S.

Youstin

T. J.

(2012). Effectiveness of residence restrictions in preventing sex offense recidivism. Crime & Delinquency, 58, 491-513.

41.

Petersilia

(2003). When prisoners come home: Parole and prisoner reentry. New York, NY: Oxford University Press.

42.

Prescott

J. J.

Rockoff

J. E.

(2011). Do sex offender registration and notification laws affect criminal behavior? Journal of Law & Economics, 54, 161-206.

43.

Sample

L. L.

Bray

T. M.

(2006). Are sex offenders different? An examination of rearrest patterns. Criminal Justice Policy Review, 17(1), 83-102.

44.

Sample

L. L.

Evans

M. K.

Anderson

A. L.

(2011). Sex offender community notification laws: Are their effects symbolic or instrumental in nature? Criminal Justice Policy Review, 22(1), 27-49.

45.

Sample

L. L.

Kadleck

(2008). Sex offender laws: Legislators’ accounts of the need for policy. Criminal Justice Policy Review, 19(1), 40-62.

46.

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, 284-302.

47.

Tabachnick

Klein

(2011). A reasoned approach: Reshaping sex offender policy to prevent child sexual abuse. Beaverton, OR: Association for the Treatment of Sexual Abusers.

48.

Tewksbury

Jennings

W. G.

Zgoba

K. M.

(2012). A longitudinal examination of sex offender recidivism prior to and following the implementation of SORN. Behavioral Sciences & the Law, 30, 308-328.

49.

Tewksbury

Lees

(2006). Consequences of sex offender registration: Collateral consequences and community experiences. Sociological Spectrum, 26, 309-334.

50.

Tewksbury

Mustaine

(2009). Stress and collateral consequences for registered sex offenders. Journal of Public Management & Social Policy, Fall, 215-239.

51.

Uggen

Manza

Behrens

(2004). Less than the average citizen: Stigma, role transition and the civic reintegration of convicted felons. In Maruna

Immarigeon

(Eds.), After crime and punishment: Pathways to offender reintegration (pp. 261-293). Devon, UK: Willan.

52.

Uggen

McElrath

(2013). Six social sources of the U.S. crime drop. The Society Pages: Social Science That Matters. Retrieved from http://thesocietypages.org/papers/crime-drop/

53.

Uniform Crime Report. (2012). Crime in the United States. Washington, DC: Federal Bureau of Investigation.

54.

Vasquez

B. E.

Maddan

Walker

J. T.

(2008). The influence of sex offender registration and notification laws in the United States. Crime & Delinquency, 54, 175-192.

55.

Veysey

Zgoba

Dalessandro

(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.

56.

Vogt

W. P.

(2005). Dictionary of statistics and methodology: A nontechnical guide for the social sciences (3rd ed.). Thousand Oaks, CA: Sage.

57.

Washington State Institute for Public Policy. (2005). Sex offender sentencing in Washington State: Has community notification influence recidivism? Olympia: Author.

58.

Willis

Grace

(2009). Assessment of community reintegration planning for sex offenders: Poor planning predicts recidivism. Criminal Justice and Behavior, 36, 494-512.

59.

Zandbergen

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.

60.

Zandbergen

Hart

T. C.

(2009). Availability and spatial distribution of affordable housing in Miami-Dade county and implications of residency restriction zones for registered sex offenders. Retrieved from http://www.aclufl.org/pdfs/SORRStudy.pdf

61.

Zandbergen

Levenson

J. S.

Hart

(2010). Residential proximity to schools and daycares: An empirical analysis of sex offense recidivism. Criminal Justice and Behavior, 37, 482-502.

62.

Zgoba

(2004). Spin doctors and moral crusaders: The moral panic behind child safety legislation. Criminal Justice Studies, 17, 385-404.

63.

Zgoba

Levenson

J. S.

McKee

(2009). Examining the impact of sex offender residence restrictions on housing availability. Criminal Justice Policy Review, 20(1), 91-110.

64.

Zgoba

Miner

Knight

R. A.

Letourneau

Levenson

J. S.

Thornton

(2012). A multi-state recidivism study using Static99R and Static2002 risk scores and tier guidelines from the Adam Walsh Act. Retrieved from https://www.ncjrs.gov/pdffiles1/nij/grants/240099.pdf

65.

Zgoba

Veysey

Dalessandro

(2010). An analysis of the effectiveness of community notification and registration: Do the best intentions predict best practices? Justice Quarterly, 27, 667-691.

66.

Zgoba

Witt

Dalessandro

Veysey

(2009). Megan’s Law: Assessing the practical and monetary efficacy. Retrieved from http://www.ncjrs.gov/pdffiles1/nij/grants/225370.pdf

67.

Zimring

F. E.

(2006). The great American crime decline. New York, NY: Oxford University Press.