SORRY, THIS BUS IS NOT IN SERVICE: PUBLIC TRANSPORTATION
ACCESSIBILITY AS A SOURCE OF RECIDIVISM RISK IN RURAL AREAS
By
Alison R. Colby

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

Criminal Justice - Master Of Science

2013

ABSTRACT
SORRY, THIS BUS IS NOT IN SERVICE: PUBLIC TRANSPORTATION
ACCESSIBILITY AS A SOURCE OF RECIDIVISM RISK IN RURAL AREAS

By

Alison R. Colby

While criminologists have successfully identified some risks associated with recidivism,
the literature has largely been from an urban perspective. The current study aims to address this
gap by examining probabilities for recidivism among offenders returning to rural counties. One
challenge to prisoner reentry that remains understudied is transportation. Therefore, a further
contribution of this study is the recognition of transportation as an obstacle to reentry and in turn,
a source of risk for recidivism, especially among rural offenders with little access to public
transportation services. Additionally, the current study examines for differences between two
types of recidivism: reconviction and technical violation. Utilizing a dataset of 3,296 men
released from prison in 1998, this data was combined with new dataset created specifically to
capture different measures of rurality at the county level. Results indicate that when compared to
offenders returning to urban counties, offenders returning to rural counties are more likely to
recidivate, and more quickly, as these individuals experienced shortened times to failure.
Furthermore, results indicate large variances in probabilities for recidivism and estimation of
time to failure between offenders who fail by reconviction and those who fail by technical
violation, warranting further analysis. Finally, access to transportation proved statistically
significant across some models, but results were inconsistent for causal inference.

This work is dedicated to my mother, Catherine, as well as my partner, Jason. Thank you very
much for the love, patience, and support you each constantly bestow upon me. I would not be the
person I am today without either of you, and I would not have been able to meet this
accomplishment without you both by my side.

iii

ACKNOWLEDGEMENTS

Additionally, I would like to properly acknowledge and thank several individuals for
their unrelenting help through this process. First, I would like to thank Dr. Beth Huebner at
University of Missouri at St. Louis for granting me the permission to use her dataset. Next, I
would like to thank the members of my thesis committee (Dr. Christina DeJong, Dr. Jennifer
Cobbina, and Dr. Jesenia Pizarro) for their guidance in preparing this manuscript. Specifically, I
would like the thank Dr. DeJong for personally taking the time and dedication in providing many
thoughtful revisions, and of course, numerous crash courses (or remedial lessons) on statistical
analysis. Without her help and constant guidance, I would surely be lost. Furthermore, I would
like to thank Dr. Pizarro for always challenging me and pushing me to do my best, and Dr.
Cobbina for asking the important questions and helping me see more clearly through my own
work.

iv

TABLE OF CONTENTS

LIST OF TABLES

vi

Introduction

1

Literature Review
Prisoner Reentry and Recidivism
Demographic Characteristics
Legal Characteristics
Environmental Characteristics
Rurality and the Risk for Recidivism
Transportation as a Challenge to Prisoner Reentry and Source of Recidivism Risk

5
5
7
8
9
12
14

Theoretical Justification: Social Disorganization Among Rural Communities

19

Purpose of Study

24

Data and Methodology
Dependent Variable
Independent Variables

27
31
32

Results and Findings

40

Discussion and Limitations

49

Conclusion

61

APPENDIX

63

REFERENCES

75

v

LIST OF TABLES

Table 1: Comparing Variable Characteristics between Missouri and the United States

64

Table 2: Description of Variables Measured in Study

65

Table 3: Descriptive Statistics for Study Sample

67

Table 4: Logistic Regression Models of Recidivism for
Reconviction and Technical Violations

69

Table 5: Logistic Regression Models of Recidivism for Reconviction Only

70

Table 6: Logistic Regression Models of Recidivism for Technical Violations Only

71

Table 7: Survival Analysis of Time to Failure to Recidivism for
Reconviction and Technical Violation

72

Table 8: Survival Analysis of Time to Failure to Recidivism for Reconviction Only

73

Table 9: Survival Analysis of Time to Failure to Recidivism for Technical Violation Only 74

vi

Introduction
Over the last three decades, U.S. prison populations have grown at an unprecedented rate
(Garland, Wodahl, & Mayfield, 2011). In 1980, for every 100,000 people, only 140 were
incarcerated (Lynch &Sabol, 2001). By 2011, this rate increased to 492 per 100,000 (Carson
&Sabol, 2012). Other estimates have gauged that 2.4 million, or 1 in 100 adults, are currently
serving time in prison or jail (Pew Center on the States, 2008).
An increase in the number of offenders behind bars thus translates to an increase in the
number of offenders that are released, as roughly 95 percent of all inmates eventually return to
their communities, and usually within three years (Hughes, Wilson, & Beck, 2001). For instance,
during 1980 approximately 170,000 offenders were released from prison (Lynch &Sabol, 2001)
but by 2008, this number more than quadrupled to nearly 735,000 (Carson &Sabol, 2012).
Roughly 2,000 inmates are released daily in the United States, which now exceeds predicated
estimates of 1,600 inmates per day (Petersilia, 2001; Travis & Lawrence, 2002; Travis &
Petersilia, 2001).
Upon their return, offenders can directly affect the public safety of their communities, as
well as social, economic and public health institutions (Lynch &Sabol, 2001; Petersilia, 2001;
Travis et al., 2001). However, the process of reentry can also directly affect the offender as they
face numerous obstacles that can further increase their risk for recidivism, or the likelihood of
being returned to prison. And while the literature on prisoner reentry and recidivism has also
grown rapidly over the last thirty years, it has largely been from an urban perspective (Osgood &
Chambers, 2000; Wodahl, 2006) examining large metropolitan cities such as Baltimore (Taylor,
1997; Visher et al., 2004), Chicago (La Vigne, Fisher, & Castro, 2004; Sampson et al., 1997;
Shaw & McKay, 1942), Boston (Brooks et al., 2005), Seattle (Miethe & McDowell, 1993), and

1

other major metropolitan areas (Nelson, Deess, & Allen, 1999; O’Brien, 2001; Richie, 2001;
Solomon, Gouvis, &Waul, 2001). However, similar to those returning to densely populated
urban cities, offenders returning to small towns or rural areas affect these communities (Wodahl,
2006). Therefore, it is important to acknowledge and empirically determine whether these effects
are significant or not. By juxtaposing our understanding of the environments in which rural and
urban crime occur, it may be possible to improve our general understanding of crime (Weisheit
& Donnermeyer, 2000).
Although there have been a handful of studies specifically examining crime in rural areas
(Freudenburg& Jones, 1991; Jobes, 1999; O’Connor & Gray, 1989; Petee & Kowalski, 1993;
Wilkinson, 1984; Wilkinson et al., 1982, 1984), some continue to argue that there has not been
enough focus on the rural perspective, especially when it comes to prisoner reentry in these
particular environments (Bottoms, 1994; Weisheit & Donnermeyer, 2000; Wodahl, 2006).
Addressing some of these scholarship deficiencies, this study examines paroled offenders and
determines if there is variation in recidivism outcomes based on whether parolees returned to an
urban or rural environment.
One obstacle to prisoner reentry (which is also understudied) is transportation and access
to it (Garland, Wodahl, & Mayfield, 2011). Though some offenders may have their own personal
vehicle or other reliable forms of private transportation, more often than not the recently released
rely on others for travel (Garland, Wodahl, & Mayfield, 2011). Since transportation systems
form the basis by which economic development can occur, as well as the means by which society
interacts, an ineffective transport system can limit economic and social opportunities (Murray et
al., 1998). However, simply having an effective transportation system may not be enough. For
example, in large cities, mass public transportation outlets are often available to all at an

2

affordable price making it relatively easy to get from “point A” to “point B,” but previous studies
have pointed out that these offenders (and especially women) have found transportation to be an
obstacle to their reentry experience (Hattery& Smith, Richie, 2001; Scroggins &Malley, 2010).
If urban offenders are reporting problems with transportation, it is expected that rural offenders,
especially those dependent upon public transportation, will experience greater challenges to
reentry as mass public transit systems are limited in these areas (Gatersleben&Uzzell, 2007).
Expanding upon an earlier analysis conducted by Huebner and Berg (2011), this study
aims to discover if varying levels of accessibility to public transportation affect the probability of
recidivism and if so, whether these effects are more heavily concentrated across rural
environments or urban ones. If a relationship is found, this analysis will not only be an
innovative addition to the study of prisoner reentry and recidivism, but also to ecological studies
of crime. Furthermore, differences in the type of recidivism are parsed out in this study and
include separate analyses based on if the offender failed on a reconviction or if they failed on a
technical violation, providing more detail to the literature. First, this study provides a review of
the literature discussing what is generally known thus far on prisoner reentry and recidivism,
how these generalities differ between urban and rural environments, and the various
demographic, legal, and environmental covariates that are influential to the risk of recidivism.
Notably in this review, transportation is recognized as a challenge to prisoner reentry, and
therefore a potential source of recidivism risk. The author is unaware of other quantitative
studies examining public transportation and recidivism risk.
Social disorganization theory (Shaw & McKay, 1942) serves as theoretical justification
for explaining recidivism among rural offenders, despite it being traditionally applied to large,
urban metropolitan cities. Since this study is more concerned with how certain places affect

3

crime and the potential risks for recidivism, an ecological theory of crime is an appropriate
foundation. Furthermore, there has been adequate scholarly support for applying the principle of
social disorganization to rural communities (Barnett & Mencken, 2002; Bouffard & Muftić,
2006; Freudenburg& Jones, 1991; Huggins, 2009; Jobes, 1999; Lee et al., 2003; O’Connor &
Gray, 1989; Osgood & Chambers, 2000; Petee & Kowalski, 1993; Reisig &Cancino, 2004;
Rephann, 1999; Spano& Nagy, 2005; Wells & Weisheit, 2004).
The purpose of study will then be discussed, including the statement of research
questions and hypotheses. Subsequently, a detailed outline of the data and methodology used is
presented, including information on the research site and descriptive statistics of the study
sample. The last sections interpret and report the statistical results of the study and relate its
findings. Finally, strengths and limitations of the study will be acknowledged and discussed, and
recommendations for future research are suggested.

4

Literature Review
Prisoner Reentry and Recidivism
Though the number of prisoners under state and federal jurisdiction declined by 0.9
percent from 2010 to 2011, roughly 1,600,000 individuals currently remain incarcerated in U.S.
prisons (Carson &Sabol, 2012). As a majority of inmates eventually return to their community, a
total of 840,676 adults were reported to be under some form of state or federal parole supervision
in 2010 (Glaze &Bonczar, 2011) and the failure rate is high: nearly two-thirds will be rearrested
and sent back to prison in three years time (Austin, 2001). In a criminal justice context,
recidivism can be defined as re-involvement of an individual to criminal behavior after they have
been convicted of a prior offense, sentenced and (presumably) corrected (Maltz, 1994).
Recidivism “results from the concatenation of failures” (Maltz, 1994, p. 1) and occurs
frequently. Approximately 40 percent of all prison admissions are parole violations that fall into
the following two categories: offenders who are being returned for technical violations and those
who have been convicted of a new crime and will now serve an additional sentence (Austin,
2001). Technical violations can occur for many reasons: failing drug or alcohol tests, failing to
report to the parole officer, not notifying the parole officer of an address or employment change,
failing to follow a treatment plan, or consorting with known felons (Hughes, Wilson, & Beck,
2001). In other words, technical violations are the failure to abide by the rules and conditions of
supervisory parole and do not necessarily have to be illegal or criminal acts. Traditionally, the
general public and members of state and federal parole boards seemed to be less tolerant of
failure in the past, sending parolees back to prison on more technical violations for an
undetermined length of time (Petersilia, 2001; Travis & Lawrence, 2002). For instance, the

5

number of parolees returned for technical violations increased from 32 to 49 percent between
1985 and 2001 (Blumstein & Beck, 2005).
Since the 2008 economic recession, however, states have been more concerned how to
maintain overcrowded correctional facilities on a dwindling budget. While the prison population
did decrease in 27 states between 2008 and 2009, it continued to grow in 23 others and the
federal prison population increased by nearly 7 percent (Gottschalk, 2010). Facing a $24 billion
budge shortfall in 2009, the state of California was forced to release 16,000 prisoners back to
their communities. With some states granting nonviolent offenders early release, this leaves even
more adults under some form of correctional supervision, and therefore, at risk for recidivating.
Notably, Gottschalk (2010) argues that economic justification ignores the fact that successful
reentry and the decarceration of offenders will cost money as these individuals need significant
educational, vocational, housing, health, and economic support once released. These needs and
services, and the process of trying to obtain them, can also influence the risk for recidivism.
The ability to predict future criminal behavior is a central part of the criminal justice
system. From it stems community safety, prevention, treatment, ethics, and justice (Andrews
&Bonta, (1994) 2003, pg. 225). If predicting future criminal behavior is one possibility,
predicting future risks of recidivism is another. Developed by Andrews and Bonta (1994, 2003),
the Level of Service Inventory-Revised (LSI-R) is a theoretically based assessment tool which
samples 54 risk and need items that have been found to be associated with criminal conduct.
Some of the items included are employment history, housing, criminal history, drug use and
social relationships.
There is adequate support for LSR-I scores being able to predict recidivism within
different types of offender populations, as well as different environmental settings (Cumberland

6

& Boyle, 1997; Lowenkamp, Holsinger, &Latessa, 2001; O’Keefe, Klebe, &Hromas, 1998;
Raynor et al., 2000). Others have tested the abilities of the LSR-I and concluded it to be just as
good, if not better, than other risk assessment instruments used to predict recidivism (Gendreau
et al., 1996; Kroner & Mills, 2001; Loza&Loza-Fanous, 2001; Motiuk, Bonta, & Andrews, 1986;
Raynor et al., 2000). For the purposes of this study, previously identified risk and need items
have been included as independent variable characteristics and will be discussed in further detail.
Demographic Characteristics
One of the leading predictors of desistance from crime is age of the offender (Bushway et
al., 2001; Gottfredson, 1990; Sampson &Laub, 1993). It is traditionally agreed upon that as the
age of the offender increases, the probability of recidivism decreases. For example, when
examining recidivism rates nationwide, Langan and Levin (2002) found that the younger the
prisoner when released, the higher the rate of recidivism: 80 percent (82.1%) of offenders under
the age of 18 were rearrested within three years, compared to 45.3 percent of those who were 45
years of age and older. Huebner and Berg (2011) discovered that young men, especially
minorities, recidivated more quickly and were least likely to desist from crime. Also supporting
the age-crime curve, similar research has found older offenders less likely to recidivate following
their release from prison (Sampson &Laub, 1993; Shover& Thompson, 1992). Desistance from
crime, however, is not “merely a function of an individual’s chronological age” (Loeber& Le
Blanc, 1990, p. 452). Many other demographic factors and personal characteristics of the
offender affect the risk of recidivism as well including relationship status, number of dependent
children, employment history, educational attainment, and mental health status.
Individuals with few social bonds to society are also more likely to recidivate following a
period of incarceration (DeJong, 1997), as marriage, having children, and employment stimulate

7

and sustain the offender’s perception of identity. This may give them the perception that they
have something to lose (Braithwaite, 1989; Maruna, 2001). In fact, research has found that those
offenders who were able to secure a stable job, even despite lengthy criminal history, were less
likely to reoffend (Laub& Sampson, 2003; Western, Kling, &Weiman, 2001). In connection with
employment, level of educational attainment is vital to reentry since uneducated offenders are
more likely to return to prison shortly after their release (Allen, 1988; Batiuk, 1997; Harlow,
2003; Vacca, 2004). Finally, including personal characteristics such as mental health is important
as previous studies have found antisocial attitudes and problems with mental health positively
related to recidivism (Berg & Huebner, 2010), especially among rural offenders identified as
mentally ill(Bucks County CCAP Committee, 2012; Myers & Associates, Inc., 2012).
Legal Characteristics
The length of an offender’s criminal record and the crime for which they were sentenced
to are both strong predictors of recidivism (Gottfredson & Gottfredson, 1994). Knowing the
number of prior convictions an offender has can be particularly useful as those with a lengthier
criminal history are more likely to recidivate (Langan& Levin, 2002; Gendreau et al., 1996). The
type of offense committed by the offender is also useful in determining future recidivism
(Burden, 2009). For example, Langan and Levin (2002) found 73.8 percent of all property
offenders were rearrested within three years, compared to 66.7 percent of drug offenders, 61.7
percent of violent offenders, and 62.2 percent of public order offenders that were rearrested.
The length of time served in prison can also affect future recidivism risk (Burden, 2009),
as the longer time spent in prison translates the greater chance of failure once released. For
example, research has found that increased time spent in prison weakens employment
opportunities (Pager, 2003), chances for steady employment (Crutchfield &Pitchford, 1997;

8

Western & Beckett, 1999), and prospects for higher lifetime earnings (Needles, 1996).
Researchers have also documented a positive association between offenders receiving
institutional misconduct violations and future recidivism (Huebner & Berg, 2011) and salient
factor scores indicating offenders as “high risk” have generally proven to validly predict
recidivism (Gendreau, Goggin, & Smith, 2002; Gendreau, Little, &Goggin, 1996; Vose, Cullen,
& Smith, 2008).
Environmental Characteristics
Drug dependence is another key factor in understanding patterns of recidivism, as there
are several causal roles between substance abuse and criminal behavior (Huebner & Berg, 2011).
For example, certain drugs have been linked to erratic behavior (Fagan, 1990), and drug use has
been known to provoke criminal employment opportunities (Uggen& Thompson, 2003; Wright
& Decker, 1997) in these release settings. More importantly, the association between drug
dependence and recidivism is significant given the prevalence of drug use among prison
populations: Fifty-seven percent of state prisoners have reported drug involvement in the month
prior to their offense, while 33 percent reportedly committed their crime while under the
influence of drugs (Laub& Sampson, 2003; Mumola&Karberg, 2006). Drug use is also known to
increase the risk of recidivism due to the consequences it can have in sustaining employment and
family relationships (Gottfredson &Gottfreson, 1994; Laub& Sampson, 2003; Sampson &Laub,
1993). Offenders incarcerated for drug related crimes have some of the highest rates of
recidivism (Chaiken&Chaiken, 1990; Harrison, 2001; Langan& Levin, 2002; Spohn&Holleran,
2002), and those classified as drug dependent upon release tend to fail more quickly (Huebner,
Varano, & Bynum, 2007).
Positive relationships and a strong social network can also be influential in offender

9

populations as they provide informal social control against future criminal behavior (Sampson,
1992). Typically, those who are married are less likely to reoffend (Fagan, 1989; Farrington &
West, 1995; Horney et al., 1995; Laub, Nagin, & Sampson, 1998; Laub& Sampson, 2001;
Ouimet& Le Blanc, 1996; Sampson &Laub, 1993; Warr, 1998) but fewer than half of all
prisoners are married (Burden, 2009). Across the general prison population, just 17 percent of
state prisoners and 30 percent of federal prisoners are married (Petersilia, 2003). Returning to a
sustained marriage, or intimate partner, upon release can provide parolees with a place to live.
Research has suggested that by having a stable housing situation, or place to live, is another step
in the right direction in reentry process and can help lead the offender finding legitimate work
(Solomon, Gouvis&Waul, 2001). However, by having serving time in prison, former inmates
have a difficult time forging social bonds necessary for marriage (Laub, Nagin, & Sampson,
1998; Western &McLanahan, 2000). It may come to no surprise then that offenders with the
lowest marriage rates tend to reside within urban, inner-city neighborhoods (Waite &Spitze,
1981).
Since most offenders are single, separated, or divorced (Laub, Nagin, & Sampson, 1998),
the ability of the offender to reconnect with family and community members is vital (MallikKane &Visher, 2008). Family and kin-based social networks can provide more than just
emotional support for parolees; they can also provide stable housing, financial support, and help
with transportation. Consequently, if ties with friends and family are fractured, the offender may
gravitate towards potentially hazardous and socially deviant networks, increasing the risk of
recidivism (Shover, 1996). Therefore knowing the offender’s relationship status and living
situation they will be returning to upon release can be useful in predicting future risks of
recidivism.

10

Parole supervision also plays an important role in risk of recidivism. Most parolees fail
immediately, with rearrest being most common during the first six months of supervision (Maltz,
1984; Schmidt & Witte, 1988). Although Lerner (1977) previously discovered that parole
supervision reduces criminal behavior for those released from local correctional institutions, and
that overall, supervision is effective in the reduction of recidivism, it is not a realistic pathway
for most offenders. Parole supervision costs remain about $2,200 per parolee, per year, yet most
jurisdictions do not have the necessary funds, resources, or correctional personnel for adequate
supervision. With more prisoners being released on a daily basis, parole officers are now
responsible for more clients. During the 1970s, parole officers were ordinarily assigned 45
th

parolees; at the turn of the 20 century, caseloads of 70 parolees per officer were common
(Petersilia, 2001). Today, caseloads have only continued to increase; in 2009, the average parole
officer was found to have an average caseload of 141 clients (Kerbs, Jolley, & Jones, 2009).
Increased caseloads, in turn, can affect the level and quality of supervision maintained by parole
officers, as well as officer discretion, thus influencing the probability for recidivism.
Relatedly, the type of community in which parole officers operate can also affect the
probability of recidivism (Burden, 2009). The behavior of legal stakeholders varies across
different types of communities such that police officers in urban areas are much more likely to
take a professional, or legalistic, approach to processing crimes (Wilson, 1968). Law
enforcement and other legal stakeholders in rural areas may choose to operate differently.
Although research on parole supervision in rural areas is limited, it has previously been found
that most officers rely on informal means of social control (Wilson, 1968; Weisheit et al., 2006).
This can have a direct effect on the decision-making process and level of discretion of the parole
officer, and determining whether the offender will return to prison on a technical violation or not.

11

Exactly where offenders return to in the community can provide important context to the
reentry process, as the places they return to, including the physical geography surrounding the
offender, can also serve as a source of recidivism risk. For instance, disadvantaged
neighborhoods that have been associated with decreased labor markets tend to offer fewer
opportunities to develop social bonds that are vital to the reentry process (Clear, Rose, Warring,
& Scully, 2003; Sullivan, 1989; Wilson, 1997). Levels of concentrated disadvantage are
important to incorporate into studies of recidivism since offenders who reintegrate into more
disadvantaged, crime-ridden environments have shown greater chances for failure (Braithwaite,
1989; Huggins, 2009; Shover, 1996).
Research on neighborhood context and reentry outcomes, however, is mixed. For
instance, Kubrin and Stewart (2006) found offenders returning to neighborhoods with higher
levels of concentrated disadvantage were more likely to recidivate, while Burden (2009) reported
concentrated disadvantage not being significantly related to recidivism. This study hopes to shed
light on this argument, but more importantly, it is suggested that since factors associated with
concentrated disadvantage are known to increase the risk of recidivism, and because these factors
manifest themselves differently within rural communities (Wodahl, 2006), it is necessary to
closely examine the effect rurality may have on offender recidivism.
Rurality and the Risk for Recidivism
Generally, states require most inmates return to the county in which they were originally
sentenced (Austin, 2001) and it has been recently noted that recently released offenders may
have more difficulty transitioning to “life on the outside” in rural settings (Wodahl, 2006). Rural
communities have their own culture and tend to offer less social privacy despite increased
geographic privacy (Weishet& Wells, 1996). Traditionally, rural communities tend to be

12

distrustful of outsiders, stress self-reliance, are less supportive of public programs, and often
resist government involvement in their lives (Beltrame, 1978; Coward, DeWeaver, Schmidt, &
Jackson, 1983; Keller & Murray, 1982; Kenkel, 1986; Littrell&Littrell, 1991; Robertson &
Donnermeyer, 1997; Seroka&Subramaniam, 1991; Swanson, Cohen, & Swanson, 1979; Warner
&Leukefeld, 2001; Weisheit & Donnermeyer, 2000).
Due to these differences in social culture and privacy, offenders returning to rural
communities arguably have different reentry experiences than urban offenders in terms of the
types of social networks formed, overall availability of jobs and housing, and the quantity (and
quality) of reentry resources and services. For example, previous research has suggested that
rural offenders may become more reliant on family members and their own resources to solve
problems (Conger, 1997), or may choose not to address their problems at all. When examining
rural substance abuse and treatment programs, Rost and colleagues (1993) noticed that recently
released offenders avoided approaching mental health and substance abuse treatment providers
out of fear of stigmatization. However, should a rural offender wish to participate in these or
similar reentry services, because of their geographic isolation, one of the most obvious barriers to
reentry in rural communities is simply the lack of reentry services in the first place (Fitchen,
1991; Leukefeld et al., 2002; Robertson, 1997).
Employment and housing opportunities can vary based on community size (Garland,
Wodahl, & Mayfield, 2011). Rural communities tend to have less diversity in employment
opportunities (Ghelfi&McGranahan, 2004), and the jobs that do exist are often service oriented
with limited economic benefits (Gibbs, Kusmin, &Cromartie, 2004). Those employed in rural
areas typically earn less to comparable positions in the city, and this pattern holds true for
offenders that eventually gain steady employment as well (Wodahl, 2006). This an important

13

distinction to make as a majority of offenders returning to rural areas will be unable to make a
decent wage through legitimate work, thus increasing the likelihood they may return to criminal
acts to supplement their income (Wodahl, 2006). It also does not help that, due to the strictly
imposed rural social culture, these communities may be more resistant to hiring ex-offenders,
leaving the offender with even fewer options for legitimate work (Holzer, Raphael, & Stoll,
2001). When it comes to safe and affordable housing, rental properties are scarce in rural areas
(Housing Assistance Council, 2003) as a vast majority of housing units are already owned and
occupied. Poverty rates are also relatively high across rural communities (Economic Resource
Center, 2004), thus making the process of becoming a homeowner that much more difficult,
especially for the returning offender.
Combining what is currently known about recidivism with existing research on rural
economics, culture, and society, it is possible to obtain a better understanding of how the
challenges of prisoner reentry can potentially influence recidivism in rural settings (Wodahl,
2006). To date, not many studies have specifically examined recidivism rates among rural
offenders (Burden, 2009; Garland, et al., 2011). However, the few studies that have been
conducted have discovered that offenders returning to rural areas were significantly more likely
to recidivate (Burden, 2009). These results were unexpected as Burden (2009) hypothesized that
urban parolees would return to prison more quickly. With rural offenders having greater
disadvantages and limitations to successful reentry due to their physical location, similar patterns
of recidivism are expected among the current study sample.
Transportation as a Challenge of Prisoner Reentry and Source of Recidivism Risk
The two largest obstacles for rural residents are physical distance and isolation (Weisheit
& Donnermeyer, 2000); thus, transportation serves an important role in rural America. By
offering increased mobility and access to rural residents, public transportation functions as an

14

essential cornerstone of rural development (Rubel, 1990). It is important to note, however, that
existing rural transportation networks have fallen into a state of disrepair with many states losing
access to service through federal deregulation during the 1990s (Due et al., 1990; U.S.
Department of Agriculture, 1991). Now more than twenty years later, a physical overhaul of the
infrastructure, as well as numerous technological advances, are still needed in order to bring rural
public transportation systems up to date for 2013.
Patterns of transportation in rural communities differ from urban communities (Stern,
1982). Due to variations in demand, accessibility, and level of service, regular daily service is
hard to provide. Additionally, with little to no access to public transportation and large distances
between houses, engaging in community life is often physically difficult in rural areas (Due et
al., 1990; Economic Resource Center, 2005; Weisheit & Donnermeyer, 2000). It is difficult for
ex-offenders as well. For instance, the typical rural substance abuser must travel further distances
for treatment services, and sometimes without the availability of public transportation
(Schoeneberger et al., 2006).
As with studies of prisoner reentry and recidivism, the majority of studies on public
transportation have also primarily been from an urban perspective (Garrett & Taylor, 1999;
Nutley, 2005). Studies specifically addressing rural bus service have found that maintaining
higher levels of service in such areas can be economically infeasible due to insufficient demand
(Rees &Wragg, 1975; White, 1976) and high cost of operations (Brown, 1996). Other obstacles
that can hinder the development of public transport in rural areas include low population density,
long distances separating rural transit stops that raise the cost of transportation, and limited
monetary resources of local governments (Brown, 2008; Gruidl, Merrett, & Connelly, 2008;
O’Sullivan, 2007).

15

With limited or no alternative means of transportation, individuals in rural areas have
been compelled to rely on personal vehicles for accessibility, speed, and as a way to retain
control over their lives (Gatersleben&Uzzell, 2007). This may explain why 97 percent of trips in
rural areas occur in cars (Pucher&Renne, 2005). Because offenders returning to rural areas are
unlikely to have their own vehicle or private transportation available to them (Leukefeld et al.,
2002), they tend to not be very transient (Oser et al., 2006). But, there still are rural residents
who require and utilize public transportation on a regular basis, even if this only represents a
small percentage of the population. For example, a past report on transportation service to rural
residents of small Iowa communities identified intercity bus service as the primary form of
transportation with more than 300 communities being served by at least one bus a day (Due et
al., 1990). In a federal report addressing the implementation of the Bus Regulatory Reform Act
of 1982, the Motor Carrier Ratemaking Study Commission (1983) found that 69 percent of Iowa
residents live anywhere from 1 to 36 miles from the nearest bus stop, and at least 11 to 20 miles
away from the second closest stop.
Transportation as a challenge to prisoner reentry is at least starting to be recognized in the
literature. A recent Housing Needs Study for the York County (Pennsylvania) Criminal Justice
Advisory Board reported the lack of public transportation as a general barrier to securing
affordable housing and services for those offenders with serious mental illness (Myers &
Associates, Inc., 2012). Specifically, study participants trying to obtain reentry services reported
a “lack of transportation outside the city, which impedes access to certain services suggesting the
need for additional resources to be located outside the City of York” (Myers & Associates, Inc.,
2012, p. 14). Furthermore, mentally ill offenders participating in the Forensic Re-Entry
Enhancement (FREE) program in Bucks County (also located in Pennsylvania) reported housing,

16

employment and transportation as their main challenges to reentry (Bucks County CCAP
Committee, 2012). In fact, nearly 70 percent of all participants reported that access to
medications and treatment was hampered by either the lack of transportation or the inability to
pay for it (Bucks County CCAP Committee, 2012).
In their recent study examining self-reported obstacles and strains of reentry to small
metropolitan communities, Garland, Wodahl, and Mayfield (2011) also report transportation as
an obstacle to successful reentry. Though a challenge for only about 15 percent of respondents,
no other study examining offender recidivism to date has recorded transportation as an obstacle
to reentry. Parolees often did not have reliable transportation of their own within three months
after their release from prison. Instead, they relied on others for travel or reached their
destination by bicycle. Similar to the studies previously mentioned (Bucks County CCAP
Committee, 2012; Myers & Associates, Inc., 2012), being able to fulfill educational and program
requirements was cited several times as a hardship, largely in part due to problems with
transportation and employment schedules. More importantly, some participants stated their
transportation issues always made them fearful of missing work and therefore consequently
receiving a parole violation, increasing the possibility of being returned to prison (Garland,
Wodahl, & Mayfield, 2011). Because of this recent recognition, access to transportation warrants
scholarly attention in examining whether it might also be related to risk for recidivism. With a
reported lack of public transportation service in rural areas and evidence of most offenders not
owning their own vehicle, the probability of recidivism may very well be greater for rural
offenders due to these circumstances.

17

Theoretical Justification: Social Disorganization Among Rural Communities
Developed out of the urban streets of Chicago, Shaw and McKay’s (1942) theory of
social disorganization contends that crime rates can be explained by the structural characteristics
of a community. Specifically, they point to three characteristics of a community (economic
disadvantage, racial and ethnic heterogeneity, and residential instability) that impact its ability to
regulate the behavior of its members through informal social control (Bouffard & Muftić, 2006),
as the concept of social networks and social bonds are a prominent theme in determining levels
of social disorganization (Kornhauser, 1978; Sampson & Groves, 1989; Shaw & McKay, 1942).
In other words, neighborhood crime rates are actually a function of the community’s ability to
establish formal and informal connections among its members so as to realize common values
and collectively come together to work toward solving and preventing social problems (Bellair,
1997; Bursik, 1988; Bursik&Grasmick, 1993; Crutchfield et al., 1982; Land, McCall, & Cohen,
1990; Osgood & Chambers, 2000; Stark et al., 1983). However, this could prove challenging
among rural communities.
A highly recognized and supported ecological model, social disorganization theory has
previously only been applied to metropolitan cities due to the fact that rural crime was thought to
be different from urban crime (Clinard 1942; Kowalski & Duffield, 1990). Others (Glenn & Hill
1977; Palen 1977) contend that a majority of this perceived difference has largely dissipated with
advances in the standardization of education, transportation, and communication, thus casting
some doubt about the inappropriateness of the social disorganization model for rural settings
(Petee & Kowalski, 1993). In fact, Osgood and Chambers (2000) argue that though much of their
analysis on rural youth violence is borrowed from urban sociology, the systems of social
relationships involved are just as relevant to crime and delinquency in small towns and rural

18

communities.
One of the first studies to examine social disorganization among rural communities was
conducted by Wilkinson and colleagues (1984). Though their results did not support the
application of social disorganization, it was one of the first steps in comparative research on
patterns of violent crime across different environments (Wilkinson et al., 1984). Fifteen years
later, Osgood and Chambers (2000) contended that the systematic processes of social control are,
in fact, supported in rural settings. When examining juvenile violence in nonmetropolitan
counties, they found that the themes of social disorganization theory did generalize well to rural
communities (Osgood & Chambers, 2000). More recently, social disorganization has been tested
in nonmetro environments (Barnett & Mencken, 2002; Bouffard & Muftić, 2006; Burden, 2009;
Garland, Mayfield, & Wodahl, 2011; Reisig &Cancino, 2004) with some interesting results. For
instance, Burden (2009) discovered that probabilities for recidivism were higher among rural
offenders, while Garland and colleagues (2011) suggest that core obstacles and challenges to
reentry, such as financial issues, employment and transportation, are consistent across
geographical locations.
Also finding social disorganization models useful for rural communities, Barnett and
Mencken (2002) argue variations in population change can have a direct affect on crime rates.
When these areas lose population, the level of social organization is transformed in important
ways: the network structures of community surveillance through which social control is enforced
are lost, thereby potentially also losing a part of the structural support that allows people to
survive (Barnett & Mencken, 2002). For some individuals, this can then lead to more desperate
economic situations, and ultimately, acts of crime and violence. Their conclusion is consistent
with existing research (Jobes, 1999; Kposowa&Breault, 1993; Kposowa et al., 1995; Osgood

19

&Chambers, 2000; Petee & Kowalski, 1993) and state that models of crime in nonmetropolitan
counties should consider the contextual effects of concepts from social disorganization theory
(Barnett & Mencken, 2002).
Rural residents, like their urban counterparts, form and maintain social networks.
However, it is likely that the nature of these networks differ between each group (Reisig
&Cancino, 2004). For example, Fischer (1982) found that residents of small towns were more
often members of kin-based social networks, and others have found similar results (Beggs,
Haines, &Hurlbert, 1996; Hofferth& Iceland, 1998). Consequently, rural residents are more
likely to exchange social support with family members than nonrelated peers (Reisig &Cancino,
2004). Earlier research conducted by Amato (1993) noted three potential explanations for urbanrural differences in exchange relationships: 1) urban residents rely less frequently on family
members for support because they are able to draw from a greater variety of formal
organizational sources for support; 2) exchange relationships in more rural areas are based on a
sense of obligation, whereas more on a matter of choice in urban settings; and 3) urban residents
are more likely to live farther away from immediate family members, so they may be more
reliant on friends for social support. Despite these differences, Reisig and Cancino (2004)
propose that the social capital gained from these different types of relationships continues to flow
across urban and rural communities, and their results also support the notion that social
disorganization theory can be generalized beyond the urban setting (Reisig &Cancino, 2004).
With a majority of rural social networks being kin-based, this can directly affect the level
of involvement and type of support provided to offenders returning home from prison, including
transportation needs. Typically, family members are more likely to help out with basic reentry
needs such as housing, finances, childcare, and transportation. These patterns of kin

20

involvement, however, have been found to differ across White/Caucasian and Black/AfricanAmerican families (Sarkisian&Gerstel, 2004). For example, members of White families reported
greater involvement in financial and emotional kin support, whereas Black family members were
more involved in practical support, including offering transportation when needed. Additionally,
with Garland and colleagues (2011) discovery that offenders returning home from prison often
relied on others for travel, the likelihood of asking for a ride or for help would most likely
decrease among rural offenders with few friends and nonkin-based social networks. Therefore,
especially across rural communities where public transportation services are lacking
(Gatersleben&Uzzell, 2007), transportation and the availability of it can directly be impacted by
level of social disorganization and type of social networks available in the area. In turn, this can
have an affect on the probability that recidivism will occur.
Levels of population density can affect the social disorganization of a neighborhood
(Sampson et al., 1997; Shaw & McKay, 1942). These levels, however, can vary widely between
rural and urban areas (Burden, 2009). Wirth (1938) was one of first scholars to discover that as
population density increased, social relationships began to break down. In their comprehensive
review on rural studies published to date, Weisheit and colleagues (2006) discovered in 21 of 23
recorded studies, communities with rapid population growth experienced a vast increase in crime
rates, with some even at three or four times the rate of the population growth.
With consistent lower levels of population density, researchers more concerned with
urban ecology have typically overlooked rural communities. This oversight is detrimental,
however, as some rural areas experience dramatic surges in population due to suburb expansion
and local economic growth (Burden, 2009). By incorporating a measure of population density in

21

this study, perhaps a more clear distinction of recidivism patterns can be discerned across urban
and rural counties.

22

Purpose of Study
While there is budding research examining the sources of recidivism risk for rural
offenders (Burden, 2009; Garland, Wodahl, & Mayfield, 2011), much remains to be discovered.
Expanding upon an earlier exploration of offender recidivism (Huebner & Berg, 2011), this
study aims to answer the following questions using the same dataset:
1a. Are offenders who return to rural counties more or less likely to recidivate
compared to offenders who return to urban counties?
1b. Do offenders returning to rural counties differ in their timing of recidivism
compared to offenders returning to urban counties?
2a. Are there differences in probabilities for recidivism between offenders who
fail on a reconviction compared to those offenders who fail on a technical
violation?
2b. Are there differences in the estimating the timing of recidivism between
offenders who fail on a reconviction compared to those offenders who fail
on a technical violation?
3a. Does having access to public transportation show a statistically significant
effect on the probability for recidivism?
3b. Does having access to public transportation show a statistically significant
effect on the time to failure?
Based on previous literature, it is therefore hypothesized that:
1a. Offenders who return to rural counties are more likely to recidivate compared
to offenders who return to urban counties.

23

1b. Offenders who return to rural counties will experience shortened times to
failure compared to offenders who return to urban counties.
2a. Probabilities of recidivism will vary depending on if the offender failed by
reconviction or failed by technical violation.
2b. Estimations on the time to failure will vary depending on if the offender failed
by reconviction or failed by technical violation.
3a. Having access to public transportation will show a statistically significant
effect on the probability of recidivism. This study aims to discover the
direction of this relationship through statistical analysis.
3b. Having access to public transportation will show a statistically significant
effect on the estimation of time to failure. This study aims to discover the
direction of this relationship through statistical analysis.
Traditionally, urban environments have been viewed as criminogenic because of their
fragile, impersonal social fabric and weakened social bonds (Bottoms, 1994). The dense and
highly populated inner city has long sparked researchers’ interests, and due to higher levels of
recording and reporting official crime rates, data have always been easily accessible from urban
metropolitan cities (Hogg & Carrington, 1998). Additionally, since a majority of the prison
population primarily returns to large metropolitan cities (Lynch &Sabol, 2001), it is only
expected that these environments would be ideal for study.
The United States is arguably an industrial nation rather than an agricultural one (Sagarin,
Donnermeyer, & Carter, 1982).With the urban population currently at 71 percent (U.S. Bureau of
Census, 2000), very few other countries have such a large portion of their population residing in
metropolitan areas. Yet, most places in the United States are identified as rural (Weisheit

24

&Donnermeyer, 2000). With a total land area of nearly 2.3 billion acres, recent reports continue
to show our nation continues to be overwhelmingly utilized for agriculture and farming
(Vesterby et al., 2006).
More importantly, while most industrial countries have some agriculture, and most
largely agricultural countries have some industry, seldom does one country have such a large
concentration of both (Sagarin, Donnermeyer, & Carter, 1982).This makes our nation a unique
case study, as our population experiences and is affected by crime across urban and rural
environments (Rephann, 1999; Scheer, Borden, & Donnermeyer, 2000). With noted differences
in their social norms and culture (Weisheit & Donnermeyer, 2000; Weisheit & Wells, 1996),
including criminogenic needs and criminality (Burden, 2009; Freudenburg& Jones, 1991; Jobes,
1999; O’Connor & Gray, 1989; Petee & Kowalski, 1993; Reisig &Cancino, 2004; Weisheit,
Falcone & Wells, 1995, 2006; Wilkinson, 1984; Wilkinson et al., 1982, 1984) rural areas are still
in need of closer examination.
And while there is budding research examining the sources of risk for recidivism among
rural offenders (Burden, 2009; Garland, Wodahl, & Mayfield, 2011), much remains to be
discovered. Specifically, little is known about how community-level factors influence recidivism
(Burden, 2009). Retaining focus on the element of rurality, this study improves upon prior works
by identifying transportation as a challenge to prisoner reentry and creating a quantitative
measure to determine if varying levels of accessibility can affect the probability of recidivism.

25

Data and Methodology
Data for this study were obtained from the Missouri Department of Corrections, as the
state maintains a comprehensive database on offender demographic characteristics, risk levels,
prison misconduct, service provision, and post-release convictions, including technical violations
(Huebner & Berg, 2011). After accounting for missing cases, a total of 3,296 men released from
prison in 1998 were included in the analysis group. Follow-up data originally collected through
May 2006 remains included in the study, as well as data from the 2000. The author has expanded
the original dataset by including variable information obtained from the two most recent U.S.
Decennial Censuses (2000; 2010) (see Table 2 in Appendix for more detail). Data on public
transportation available in the state was compiled from the Missouri Department of
Transportation, the Missouri Public Transportation Association, and the American Public
Transportation Association.
Missouri continues to serve as the research site due to the fact this is where data was
originally collected. However, this particular state, like our nation, also has a significant
concentration of urban and rural counties. Although Missouri is home to large metropolitan cities
like St. Louis and Kansas City, a majority of counties in the sample (n = 81 out of 115) have
been deemed “nonmetro” according to current Rural Urban Continuum Codes (Economic
Research Service, 2013). Additionally, Missouri incarceration rates have also increased. This has
led to a substantial rise in the number of offenders annually released in the state. For instance, in
1975, only 1,120 offenders were released from Missouri prisons but by 2010, this number
skyrocketed to 19,421 offenders released (Carson &Sabol, 2011). With a large increase in the
number of adults under parole supervision, and a vast majority of the state being nonmetro,
Missouri can serve as an ideal environment for examining recidivism from the rural perspective.

26

Overall, Missouri appears to be comparable to national averages (see Table 1 in Appendix for
more detail) but there are a few distinctions that make this area unique for comparing urban and
rural environments. Slightly higher in population density than the national average (81.2
residents per square mile versus 79.6), most Missourians tend to reside in, or at least nearby, one
of the nine major metropolitan cities. One noticeable distinction that merits discussion is the
racial and ethnic composition of the state. While the nation’s population is reportedly 72.4
percent White/Caucasian, this racial group is overwhelmingly represented in the state population
(82.8%). Though the percentage of Black/African-American residents in Missouri is close to the
national average of 12.6 percent (11.6%), there is a large discrepancy in the percentage of
Hispanic/Latino residents. Nationally, this racial group represents a little over 16 percent of the
population (16.3%). In Missouri, however, the Hispanic/Latino population is hardly represented
at all, with only 3.5 percent comprising the state population. Notably, only 5.1 percent of
Missouri residents speak a second language, compared to nearly 18 percent (17.9%) of the
country. This low figure may make more sense when observing the percent of native born in the
state: 97.3 percent of Missouri residents are native born compared to 87.7 percent of U.S.
residents.
Another distinction between Missouri and the United States are rates of crime. For the
most part, violent crime occurs roughly at the same rate as national averages, but the aggravated
assault rate in Missouri is higher (312.7 per 100,000 residents versus 241.1). Missouri also has
higher rates for certain property crimes. For example, the most recent burglary rate reported to be
745.7 (compared to 702.2, nationally), the larceny-theft rate to be 2,308.3 (compared to 1,976.9)
and the motor vehicle theft rate to be 254.8 (compared to 229.6) (FBI, 2011). In attempting to
explain such elevated rates, some have argued that urban and rural crime trends are now

27

mirroring one another through the advancement of modern communication and transportation
(Weisheit, Falcone, & Wells, 2006). Opportunities for crime have increased in rural areas due to
increasing inventories of farming equipment and other operational inputs; the interstate highway
system and improvements in transportation; and the overall growth of the rural population due to
relocation of industry associated with energy development, construction of suburbs and the
migration of retired households to small town, open-country environments (Saragin,
Donnermeyer, & Carter, 1982).
The current study expands upon the original data provided by including additional
environmental variables specific to rural populations. First, a variable identifying if participants
returned to an urban or rural county within the research site was created. This was used by
obtaining the most recent Rural Urban Continuum Code (RUCC) for each county in the sample,
available from the Economic Research Service located within the U.S. Department of
Agriculture (2013). Originally developed in 1974, the RUCC forms a classification scheme that
distinguishes metropolitan counties by the population size of their metro area, and
nonmetropolitan counties by degree of urbanization and adjacency to a metro area (Economic
Research Service, 2013).
Specifically, the official Office of Management and Budget (OMB) metro and nonmetro
categories are subdivided further, and each county in the U.S. is assigned one of the nine codes.
A value of 1 represents counties in metro areas with populations of one million of more, while a
value of 9 represents counties in nonmetro areas that are completely rural, or an urban population
less than 2,500, and not adjacent to a metro area (Economic Research Service, 2013). This
method allows researchers to break county data into finer residential groups, particularly for the
analysis of trends in nonmetro areas. For the purposes of this study, a county was considered

28

rural if it had an RUCC code of 4, 5, 6, 7, 8, or 9. Examining rural offenders with substance
abuse issues, Oser and colleagues (2009) used a similar measure for rurality, but instead labeled
counties rural if they had an RUCC code of 6, 7, 8, or 9. Due a higher prevalence of nonmetro
counties within the study site, counties with an RUCC code of 4 or 5 were identified as rural.
Next, a measure of population density (number of individuals per square mile) has been
added to the dataset using information from the 2000 U.S. Decennial Census. Examining
population density is worthwhile when examining rural environments, as recent studies have
shown it to be a better indicator of rurality than overall population counts (Bouffard & Muftić,
2006). Social disorganization theorists have also long recognized population density as a useful
variable for predicting crime and deviant behavior (Bursick &Grasmick, 1993; Lee et al., 2003;
Sampson, 1985; Sampson & Groves, 1989; Sampson et al., 1997; Shaw & McKay, 1942) due to
the observation that as population density increases, social relationships begin to break down
(Wirth, 1938).
As transportation has at least been mentioned to be a challenge for some returning
offenders (Bucks County CCAP Commission, 2012; Due et al., 1990; Garland, Wodahl, &
Mayfield, 2011; Myers & Associates, Inc., 2012), this analysis aims to determine whether access
to public transportation is related to recidivism, particularly among rural offenders as their
communities tend to lack these services (Economic Resource Center, 2005). Variables measuring
the accessibility and frequency of various public bus services offered at the county-level have
been included and coded for each participant.
This study proceeds into three stages. First, data on a sample of men released from prison
are used to estimate how demographic, legal and environmental characteristics, including levels
of rurality, affect their probability of recidivism. To begin, univariate analysis will be conducted

29

in order to gather descriptive statistics and look for potential outliers and errors in the data. Next,
bivariate analysis will examine the distribution of the sample of those returning to rural counties
and those returning to urban. Second, multivariate logistic regression was used to determine
which variables affected the probability of recidivism in the sample. Before the third stage,
however, an ordinary linear regression was performed to check for collinearity. It is important to
note that no extreme or conflating collinearity diagnostics were found. Finally, a Cox
proportional hazard model was aided in determining which independent variables were
significant in estimating shortened or delayed times to failure, and if offenders returning to rural
counties experienced shortened times to failure when compared to urban offenders.
Dependent Variable
It is first necessary to clearly define how recidivism will be measured, as strong debate
over its definition and measurement continues (Bushway et al., 2001; Maltz, 1984; National
Research Council, 2008; Ritter, 1997). The most commonly used measures of recidivism include
rearrest, reconviction, and return to prison through revocation of parole supervision (Burden,
2009). Though Langan and Levin (2002) have suggested that parolee rearrests and reconvictions
may overestimate rates of recidivism, this study follows in line with Huebner and Berg’s (2011)
original measure. Dichotomous in nature, their measure of recidivism includes offenders who
have sustained either 1) a new conviction following release from prison (reconviction), or 2)
those returned to prison through revocation of parole (technical violation). By including
technical violations in the measure of recidivism, possible sample bias may be reduced while
also allowing for a more accurate sample (Huebner & Berg, 2011). They continue to explain that
if revocations due to technical violations were not included in the measure of recidivism, their

30

absence could then lower estimates of failure and bias models of analysis by omitting what could
be measured and quantified as recidivism (Hubener& Berg, 2011).
Overall, 2,200 men (66.7%) recidivated during the original study period; 1,143 were
reconvicted, with the remaining 1,057 participants receiving a technical violation. This rate of
recidivism is consistent with other research documenting a 59 percent failure rate among
offenders using similar measures (Yahner&Visher, 2008). Interestingly enough, although only
840 study participants returned to rural areas, this group showed a higher probability of
recidivism (70.47%) when compared to the 2,456 parolees returning to urban counties (65.47%).
Reconvictions were more common among those returning to rural counties (49.18%) compared
to technical violations (29.64%), and similar percentages were found in the urban subgroup
(46.87% returned for reconviction and 32.90% for technical violation).
Independent Variables
This study incorporates three broad clusters of independent variables that have been
associated with recidivism. Demographic characteristics were assessed at the original time of
offender intake and include age at release (in years), if the offender is a minority (Black = 1), and
if they are a high school graduate(yes = 1) or not. More minority participants returned to urban
counties (40.59%) than those identified as rural (11.55%). Also included in the dataset is if the
offender reported a stable work history (yes = 1) or not. This variable was measured prior to their
incarceration and defined as having no noted employment deficits and good reports from
employers. Men with sporadic or no employment or training experience prior to incarceration
served as the reference category (Huebner & Berg, 2011). Only a little over 25 percent (25.79%)
of the sample met these requirements and reported a stable work history. Relatedly, stable mental
health (yes = 1) is also included and was assessed by the state department of corrections at the

31

original time of offender intake. A majority of the sample (93.11%) reported to be mentally
sound, and this applied across rural and urban counties. As family and social networks are vital
to prisoner reentry, the number of dependent children (yes = 1) is one measure of pre-prison
social relationships. Offenders returning to rural counties were slightly more likely to be a father
to one or more dependent children (34.40%) when compared those returning to urban counties
(32.90%).
Measures of past criminal behavior and other legal characteristics are also critical to
understanding future criminal involvement. For this study, criminal history is measured as
indicated by the number of prior convictions, for any crime, prior to the offenders’ current
offense. Using offense classifications from Langan and Levin (2002), current offenses included
are property (burglary, larceny, malicious destruction of property), drug offenses (drug use,
possession, sales, and trafficking), and personal offenses(rape, robbery, assault, murder, and
sexual offenses). Some men in the sample, however, had committed offenses not so easily
categorized. Therefore, other offenses committed by offenders in the sample have been
recognized and include domestic crimes, obstructing or fleeing from law enforcement, disturbing
the peace, and weapon and traffic offenses. Most study participants were serving time for
committing a property crime (40.84%), with personal offenses coming in second (23.82%).
Participants were least likely to be serving a sentence for a drug offense (19.60%) or other
miscellaneous offenses (15.75%). However, when examining the distribution of offense type
across rural and urban counties, offenders returning to rural counties committed slightly more
property crime (44.88% compared to 39.35%) and other offenses (20.00% compared to 14.29%).
Although institutional misconduct is often omitted from existing empirical studies of
recidivism (Huebner & Berg, 2011), some have found some significant predictive power in

32

measure of criminal history and institutional misconduct. (DeLisi, 2003; Matsueda, Gartner,
Piliavin, &Polakowski, 1992). Therefore, a measure of the total number of misconduct
violations, for any offense or infraction, sustained between entrance to prison and date of release
is included. Misconduct was prevalent among the original sample with an average of 6.84
misconduct tickets prior to release. However, misconduct violations were more frequent among
those returning to urban counties (7.34) compared to those returning to rural counties (5.46).
Time served in prison, in days, is also included as a control measure. Overall, men averaged
approximately two years (768.74 days) in prison before release (Huebner & Berg, 2011).
Offenders returning to urban areas served almost one year longer than those returning to rural
(829.68 days compared to 590.53). Using a composite risk factor for predicting recidivism, this
study also incorporates the offender’s salient factor score, based on three criteria: conviction and
confinement measures (prior incarceration, prior convictions and total sentence length), stability
(age at first imprisonment, drug and alcohol use, and offending patterns), and performance and
behavior measures(failure on community supervision, escape attempt, and prior conviction for
burglary). Salient factor scores can range from 0 (highest risk) to 11 (lowest risk). Though
measures of prior convictions are already included in the model, the salient factor score is seen
as a global policy tool for offender risk assessment. Despite differences in time served and
misconduct violations, offenders returning to urban (7.37) and rural counties (7.51) showed
similar salient factor scores, though rural offenders appear to be slightly less risky.
The dataset also includes seven measures of post-release context addressing the
offender’s release setting and living situation. First, a dichotomous measure of drug dependence
(yes = 1) is included as assessed 90 days post-releaseusing the Texas Christian University Drug
Screen II (Knight, Simpson, &Moyey, 2002). Nearly 22 percent (21.84%) of participants were

33

identified as drug-dependentupon release, and those returning to rural counties appear to be just
slightly more drug dependent (22.86%) than urban offenders (21.5%). Next, each participant’s
living situation was captured to include the following measures: whether the offender returned to
a sustained marriage (yes = 1), intimate partner (yes = 1), if they returned to family (yes = 1)or
transitional housing (yes = 1), if they were placed under enhanced supervision, and the level of
concentrated disadvantage in their returning county.
Designed to capture stability in social bonds, only 15.24 percent of offenders returning to
rural counties reported being in a sustained marriage, however, this rate is still higher than urban
parolees (12.05%). The measure of transitional housing includes men who lived in a communitybased halfway house, as well as those who returned to a release center managed by the
department of corrections (Huebner & Berg, 2011). With public housing options more prevalent
in urban cities, nearly 20 percent (19.91%) of participants returning to urban counties were able
to utilize a halfway house, while only 8.21% of rural parolees returned to transitional housing.
As discussed, conditions of parole supervision can shape reentry and the risk for
recidivism. Though almost a quarter of the sample (23.5%) reported being on some form of
enhanced parole supervision upon release, this form of supervision was utilized more in rural
counties: 32.86 percent of rural parolees compared to only 20.24 percent of those returning to
urban counties. Notably, as enhanced parole supervision is often ordered for men deemed at
higher risk for recidivism, it has also been associated with greater chance to receive a technical
violation (Petersilia & Turner, 1991). Therefore, it is anticipated that individuals receiving
enhanced parole supervision will fail more quickly.
In addition, levels of concentrated disadvantage are also included in the study. Originally
constructed by Huebner and Berg (2011), this measure uses data on the first post-release address

34

merged with data from the 2000 U.S. Decennial Census, geocoded in GIS ArcView and linked to
a census tract. The five-item factor score of concentrated disadvantage includes the proportion of
individuals living on public assistance, below the poverty level, unemployed, Black, and living
in female-headed households. The difference in levels of concentrated disadvantage across metro
and nonmetro counties is striking: those returning to urban counties reported a disadvantage level
of 0.14 while parolees returning to rural counties reported levels of -0.41. Also striking in
difference are levels of population density. Rural parolees were spread out with only 62.71
individuals per square mile while urban parolees returned to far more densely populated
counties: 1,398.39 individuals per square mile.
In order to get a sense of how far some offenders may need to travel to report to their
parole officer, and also to control for county parole resources, a categorical variable of parole
office location was included in the dataset by identifying the county location for each of the 53
parole offices across the state. Once identified and linked to a census tract, each county was
coded into the following categories: Has contiguous county parole office = 1; Has a county
office = 2; and Has more than one county office = 3. When examining the distribution of the
sample, parolees returning to rural counties seem to experience far less administrative coverage
than those returning to urban counties. For example, only 31.92 percent of rural parolees
returned to a county with its own parole office, leaving 35.36 percent of rural parolees access to
a parole office contiguously located nearby. Though parole office location and coverage reported
to be higher for urban parolees, still only 61.69 percent of those returning to urban counties had
access to more than one parole office in their entire county.
The Missouri Public Transportation Association and Department of Transportation are
responsible for providing service to all 115 counties, including those identified as rural. Agencies

35

providing public bus service to these areas utilize two regional systems: OATS, Inc., which
serves a majority (85 counties) of the state and Southeast Missouri Transportation System, Inc.
(SMTS), which serves the remaining 20 counties (Missouri Public Transportation Association,
2013). Additionally, there are nine county systems, 19 city systems, and two university systems
available for general public use. There are also a few nonprofit organization and agencies that
offer some form of enhanced transportation service, whether by city bus, taxi coupon programs,
or intercity bus assistance (Missouri Public Transportation Association, 2013). In order to
examine the types of public transportation available in each county, locally identified bus
services are coded as follows: Has an OATS system (yes = 1); Has an SMTS system (yes = 1);
Has a county wide system (yes = 1); Has a university system for public use (yes = 1); and Has a
nonprofit or some other agency transportation service (yes = 1).
Identifying an earlier measure of access to public transit from Murray and colleagues
(1998), the level of access to service can be defined as simply having the opportunity for use.
Sometimes, however, having the opportunity alone is not enough if there is low frequency of
service. An earlier study examining network connectivity and frequency of rural bus services in
Israel, Stern (1980, 1982) previously measured frequency of service by observing a range in the
number of bus departures available each day. Recognizing the special transportation needs
among rural residents, the state of Missouri claims they will “find a way to help you get where
you need to go,” (Missouri Rural Public Transportation Providers, 2013) by providing every
county with at least some public bus service.
Although transportation services are available to anyone regardless of age, income,
disability, race, gender, religion, or national origin (OATS, 2013; SMTS, 2013), there are few
designated bus stops, or terminals, where riders may gather and wait to be picked up. Rural bus

36

services in Missouri typically fall within two categories: regular route service with deviations
(limited to three-quarters of a mile off the fixed route) and demand response service by
appointment. The latter requires the rider to call at least 24 or 48 hours in advance and schedule
their pick-up, with some transportation agencies recommending at least seven days in advance
(City of Joplin MAPS Transit, 2010).
In order to partially control for the varying levels of frequency of service, types of public
bus service available in each county were identified in each county and if they operate under
regular route service, on demand pick-up service, or both. Counties limited to on demand pickup service (those only receiving OATS and SMTS services) have been identified and coded at
the lowest level of accessibility (no regular coverage = 1). Counties offering at least some regular
service either through a city within the county, a public university owned system within the
county, or a county operated transit system have been identified and coded at the middle level of
accessibility (partial coverage = 2). And finally, counties offering bus service with daily
departures available throughout the week have been identified and coded at the highest level of
access (regular coverage = 3). Not surprisingly, 72.62 percent of parolees retuning to rural
counties had no regular coverage and only 27.38 percent having some access. A majority of
those returning to urban counties reported having some access available to them (70.48%), but
all in all, only 19.14 percent of the sample had regular, daily access to bus service within their
returning county.
High energy, impulsiveness, low self-control, and sensation-seeking behavior are the
“Big Four” correlates that broadly define the antisocial personality pattern that can strongly
influence criminal behavior (Andrews &Bonta, (1994) 2003). Not surprisingly, these personality
traits are common in adult offenders (Andrews &Bonta, (1994) 2003). Less capable of

37

recognizing personal responsibilities, making and sticking to a daily or weekly schedule, these
individuals tend to lack adequate personal organizational and life management skills. Thus,
parolees in need of a ride in order to report to their parole officer, attend a substance abuse
counseling session, or show up for a job interview, may find scheduling a ride through one of
these rural transportation providers challenging. It is commendable that the state has these
services in place for rural residents, but it is solely up to them (including the rural offender) to
remember how and when to utilize the service. It is expected that although a vigorous effort to
adequately provide public transportation services across the state has been made, offenders
returning to rural counties will experience difficulty in access to these services, thus potentially
affecting the risk for recidivism.

38

Results and Findings
This section details the results of logistic regression and survival models designed to
predict the probability of recidivism among adult male parolees, controlling for the independent
variables described in the study. Recidivism serves as the dependent variable for logistic models,
but has been divided into the following categories for more fine-tuned analysis: total failure, or
those that either received a reconviction or technical violation; failure only by reconviction; and
failure only by technical violation. The logistic regression models and results displayed in Tables
4, 5, and 6 (see Appendix) are designed to estimate the probability that an individual will
recidivate during the time of original study. Model 1 displays probabilities for recidivism based
on the level of rurality, while Model 2 denotes probabilities for recidivism based on having the
availability of regular public transportation coverage. Finally, Model 3 reports the probabilities
for recidivism based on the availability of more than one parole office in the county. Positive
coefficients delineate a greater chance for recidivism, while negative coefficients coincide with a
lesser probability of recidivism (Long, 1997).
The results displayed in Tables 7, 8, and 9 include coefficients from Cox proportional
hazard models, or survival models, and estimate the probability of failure at time t (Cox, 1972;
Singer & Willett, 2003). A positive coefficient for each variable characteristic indicates the
increased probability of hazard for that particular event; therefore decreasing predicted offender
survival duration. In other words, positive coefficients delineate a shorted time to failure. A
negative coefficient indicates the decreased hazard probabilities, resulting in increased survival
times, or delayed time to failure. Mirroring logistic regression models, Model 1 displays
estimated time to failure based on the level of rurality, estimated time to failure based on having
the availability of regular public transportation coverage is shown in Model 2, and the estimated

39

time to failure based on the availability of more than one parole office in the county is displayed
in Model 3.
Beginning with the results from logistic regression models for total recidivism displayed
in Table 4, these initially suggest agreement with existing predictive models of recidivism. As
anticipated, indicators of age, race, offense type, criminal history, drug dependence and returning
living situation proved to be central correlates in the chance that recidivism will occur. Results
for the first coefficient of age are positive and suggest that aging decreases the chance of the
offender recidivating. The log odds ratio presented in Models 1, 2, and 3 signify that for each
yearly increase in age, male participants were .037 times less likely to recidivate. Additionally,
negative coefficients for the minority variable suggest that Black/African-American men were
.488 times more likely to recidivate if returned to a rural county; .435 times more likely to
recidivate if returned to a county with regular public transportation coverage; and .501 times
more like to recidivate should they return to county with more than parole office.
Previously committing a drug offense also proved to a significant predictor, and parolees
in this group were overall less likely to recidivate compared to other types of offenses (B = .260). The negative coefficient for the institutional risk score indicates that men with higher
salient factor scores (lower risk) were less likely to recidivate, while the positive coefficient for
total number of misconduct violations indicate with each additional violation, men are .034 times
more likely to recidivate.
Examining results for environmental characteristics, drug dependence and marriage were
found to be significantly strong predictors. For participants identified as drug dependent, these
individuals were extremely more likely to recidivate compared to offenders who were not
identified as drug dependent, reporting increased probabilities of .784. Furthermore, higher

40

probabilities of failure for drug dependent offenders were reported across each of the three
models and types of recidivism (for comparison see Tables 4, 5, 6, 7, 8, and 9 in Appendix). As
well were men returning to transitional housing; the log odds ratio indicate that men are .333
times more likely to recidivate if they return to a transitional housing compared to those
offenders who did not return to transitional housing; .339 times more likely to recidivate if
returning to county with regular public transportation coverage; and .342 times more likely to
recidivate if returning to a county with more than one parole office. Conversely, parolees
returning to their spouses upon release were significantly less likely to recidivate (B = -.970)
compared to those who did not return to a sustainable marriage.
More importantly, results in Model 1 suggest that those returning to rural counties show
greater odds for recidivating when compared to parolees returning to urban counties. Those
returning to rural counties upon release reported an increase in probability of recidivating by
.348. In other words, offenders returning to rural counties are .348 times more likely to fail.
When examining the probability of recidivism based on parole office location, Model 3 indicates
that men returning to counties with more than one parole office are .314 times less likely to
recidivate (B = -.314). Notably, the probability for recidivism of those returning to counties with
access to regular public transportation, as displayed in Model 2, showed these offenders
experienced a 1.465 (B = 1.465) increase in the likelihood of recidivating.
Table 5 denotes results logistic regression models of recidivism based on those
participants who failed specifically by receiving a reconviction. Though similar to the results
previously discussed for total recidivism, some interesting details emerge across this model.
Although many indicators remain central correlates in the likelihood that recidivism will occur
(age, race, offense type, criminal history, drug dependence and returning living situation), there

41

are distinct differences worthy of discussion. First, having dependent children now emerges as a
significant predictor, as men who are the guardian of a minor child were .510 times less likely to
fail on a reconviction. This was also true for Models 2 (B = -.503) and 3 (B = -.499). Stable work
history also proved to be a significant predictor, but not as anticipated. Results in Table 5 show
that men with stable work history were .238 times more likely to fail on a reconviction. These
results prove different from previous research that has indicated offenders able to secure stable
work are less likely to reoffend (Laub& Sampson, 2003; Western, Kling, &Weiman, 2001).
However, it is important to note that this particular measure of work history does not capture if
the offender obtained and sustained employment upon release, only if they have sustained a
previous stable work history prior to incarceration.
In this particular model, the number of prior convictions now reports to be statistically
significant. For men in this particular sample, the log odds ratios report that each additional prior
conviction increases the chance for recidivism, though only slightly (B = .120). But another
interesting fact stands out in the results for those offenders who failed on a reconviction: the
influence of returning to transitional housing. Remaining a significant predictor in this model, for
those parolees eventually returned on a reconviction, transitional housing appears to decrease the
odds of recidivating (B = -.293). Also remaining a significant predictor is the level of rurality, as
men continued to show greater odds of recidivating when returning to a rural county (B = .395).
Furthermore, this model predicts for those returning to a county with more than one parole office
are significantly less likely to fail on a reconviction (B = -.409), while those returning to a county
with access to regular transportation service seem to experience an increased likelihood of failure
(B = 1.619).

42

The third logistic regression model analyzing the probability of recidivism for parolees
receiving technical violations are displayed in Table 6 (see Appendix). It is here where especially
striking differences in probabilities were noted. Age, race and environmental characteristics
remain central correlates in the model, but show a weaker influence. The first indicator that is
noticeably different is that offenders with dependent children were more .587 times more likely
to fail by technical violation – the opposite of what we saw with offenders who failed by
reconviction. Additionally, stable work history is no longer statistically significant for this group.
Another indicator of difference is the predictive ability legal characteristics and criminal
history. For each additional prior convictions sustained, this now reportedly lowers the odds for
recidivism on a technical violation (B = -.106). Furthermore, the type of offense committed is
also different for offenders in this group. Under this model, drug offenses were no longer
statistically significant, but property offenses were and increased the odds for recidivism by .379.
Variables of criminal history including salient factor score, number of misconduct violations, and
the number of days served in prison were also no longer statistically significant in their
predictive abilities for offenders who failed on a technical violation.
Reviewing environmental characteristics, drug dependence, marriage and returning to
transitional housing remained centrally correlated as well. Similar to previous regression models,
men identified as drug dependent were more likely to recidivate (B = .360), and those that
returned to a sustained marriage were less likely (B = -.447). However, men returning to
transitional housing were significantly more likely to fail on a technical violation by .542 times.
These results were the complete opposite from the reconviction group. Most important of all,
across all three measure of rurality (rural county; regular access to public transportation; and
county with more than one parole office) were no longer statistically significant under this

43

regression model, and therefore can no longer serve as adequate predictors of recidivism for
offenders failing on a technical violation.
When examining survival analysis of estimates on how long it takes parolees to
recidivate, or their estimated time to failure, patterns emerged similar to logistic regression
models but again with some notable differences. Reviewing estimated odds of the hazard of the
event of failure, results in Table 7 (see Appendix) show age, race, criminal history, and certain
environmental characteristics congruent with earlier results. As men increased each year in age,
their hazard was decreased by .024, or in other words, experienced a 2.4 percent delay to failure.
Whereas minority participants displayed a significant hazard increase, overall quickening their
time to failure (B = .213). What especially stood out in this particular model is the offender
having dependent children. This variable proved to be statistically significant, but indicates that
men who are fathers experience a shortened time to failure (B = .144)
Following logistic regression models, participants who previously committed drug
offenses, who were already less likely to recidivate, report noticeable delays in their time to
failure based on survival analysis (B = -.233). The same held true for those offenders who
committed other offenses. This group also experienced delayed times to failure (B = -.204).
Criminal history patterns are also congruent with logistic regression models and suggest that
those with higher salient factor scores (lower risk) take longer to fail (B = -.078); and those with
more misconduct violations fail more quickly (B = .020), as well as men that served more days in
in prison (B = .000)
Although participants who previously committed drug offenses reported delayed time to
failure, men identified as drug dependent in the sample experienced a significant increase in
hazard, thus decreasing their survival time while simultaneously quickening their time to failure

44

(B = .369). Similar results were obtained for those who returned to transitional housing (B =
.240). These participants experienced an increase in hazard of that particular event, thus
decreasing their odds of survival. Conversely, men who returned to their spouses upon their
release were congruent with recidivism patterns; this group showed a significant delay in their
time to failure (B = -.582).
Finally, and as anticipated, rurality proved to be a significant indicator in estimating
times to failure. Men returning to rural counties upon release increase their hazard by 14.9
percent (B = .149), thus indicating these participants experienced shortened time to failure by
returning to these areas. This would make sense as logistic regression models predicted that
returning to a rural county increased the offender’s chances for recidivating. Results in Model 2
show access to public transportation is no longer significant in estimating times to failure. When
examining results in Model 3, however, it appears that parolees returning to counties with more
than one parole office decrease their hazard by 17.5 percent. (B = -.175), thus delaying their time
to failure.
The results from survival analysis of time to failure to recidivism on a reconviction are
shown in Table 8 (see Appendix). Generally these covariates showed similar results to logistic
regression and survival analysis for overall recidivism, however, some proved to be stronger
estimates in time to failure. For instance, minority participants experienced a further increase in
hazard (B = .282), and those with drug offenses doubled their decrease in hazard, thus
significantly delaying time to failure (B = -.448). Having dependent children seemed to be
particularly beneficial to this group, as each additional child appears decrease the odds of hazard
by 22.9 percent (B = -.229). Similar to logistic regression results for this particular group, stable
work history interacts differently with those who were eventually returned to prison on a

45

reconviction. Now statistically significant, these participants with a stable work history report a
15.4 percent (B = .154) increase in hazard and therefore tend to fail more quickly.
Reviewing legal characteristics, the number of prior convictions proves to be statistically
significant for this group and men who eventually fail on reconvictions increase their hazard by
7.3 percent (B = .032) for each additional prior conviction they have sustained. Those who
previously committed drug offenses continued to experience significant delays in time to failure
(B = -.448), while again, drug dependent men in the sample reported a 30.3 percent (B = .303)
hazard increase, thus quickening their time to failure.
Marriage remains a central correlate and continued delay in time to failure. Notably,
transitional housing for offenders who fail on a reconviction is no longer statistically significant
in estimating time to failure. Finally, parolees returning to rural counties that eventually fail on a
reconviction report a 28.2 percent (B = .282) hazard increase, while those returning to counties
with more than one parole office experience a 31.9 percent (B = -.319) decrease in their odds of
hazard, thus delaying their ultimate failure. Notably, parolees returning to counties with regular
transportation coverage reported shortened times to failure on a reconviction by 98.2% (B =
.982).
For study participants who recidivate on technical violations, results for the survival
analysis of time to failure are displayed in Table 9 (see Appendix). Again, age of the offender
remains a central correlate, decreasing by .025 hazard odds each year as the offender gets older,
however, the minority variable and stable work history are no longer statistically significant in
this particular model. Having dependent children significantly increases hazard odds (B = .489),
thus shortening time to failure for those recidivating on a technical violation. The number of
prior convictions remains significant from the previous model depicted in Table 8, but for those

46

failing on a technical violation, each additional prior conviction decreases hazard odds by 6.6
percent (B = .066), conversely extending their time to failure. Again, criminal history patterns are
also congruent with logistic regression models and suggest that those with higher salient factor
scores (lower risk) take longer to fail (B = -.067); and those with more misconduct violations fail
more quickly (B = .013).
When it comes to environmental variables, drug dependent men who eventually failed on
a technical violation report .436 increased hazard odds, thus shortening their overall time to
failure by 43.6 percent. This was also true for those returning to transitional housing and who
failed on a technical violation (B = .506). What is more notable in this particular model is the
measure of returning to family. Not statistically significant in any other model in this study, men
who returned to family upon their release, and who eventually failed on a technical violation,
slightly increased their hazard odds by 14.1 percent (B = .141) and failed more quickly compared
to those who did not return to member of their family. Finally, all measures of rurality were
found to be no longer statistically significant in estimating the time to failure for those offenders
receiving technical violations. These results suggest that estimates of time to failure for a
technical violation could not be as accurately captured as estimates of time to failure for a
reconviction

47

Discussion and Limitations
This study was designed to explore the probabilities of recidivism among paroles who
returned to rural counties compared to those returning to urban, as well as the possible influence
public transportation may or may not have as a source of recidivism risk. The current study’s
goal was threefold. First, the difference between covariates and related concepts of recidivism
risk were identified and compared across the rural-urban continuum. Next, these same covariates
were further analyzed for differences between two measures of recidivism: those who fail by
reconviction and those who fail by technical violation. Finally, this study was also concerned
with determining if access to public transportation affects the probability for recidivism.
Overall, this analysis is congruent with existing models of recidivism, but more
importantly, confirms the first set of hypotheses put forth and also supports more recent findings
among rural offenders (Burden, 2009; Garland, Wodahl, & Mayfield, 2011). As anticipated, the
probability (and therefore, risk) of recidivism is higher for parolees residing in rural areas
(Hypothesis 1a), and participants in this group also experienced a shortened time to failure as
well (Hypothesis 1b). Following previous models of recidivism, measures of age and race
continued to be significant predictors in the likelihood of failure with younger men and men of
color being more likely to recidivate, and at a more quickened pace. Like Huebner and Berg
(2011) discovered in their original analysis, social relationships and drug dependent remain
important predictors for the current study as well. Men who reportedly returned to a sustained
marriage were significantly less likely to recidivate, and were delayed in their failure, while
participants who claimed to be drug dependent were more likely to fail and in shorter periods of
time. When looking at the offender’s release setting and related environmental characteristics,
results suggest that men who return to transitional housing experience higher probabilities for

48

recidivism and fail more quickly than those returning to family. These results, however, can only
be applied when examining total failure of the sample, as differences existed based on if the
offender failed by reconviction or by technical violation.
By separating probabilities of recidivism for reconviction versus probabilities of
recidivism for technical violations, some interesting differences emerged. These unique
distinctions help lend support for the second set of hypotheses put forth: that probabilities for
recidivism will vary between reconvictions and technical violations (Hypothesis 2a) and
estimations on the time to failure will also vary between these groups (Hypothesis 2b). For
instance, transitional housing remained to be a strong predictor in the probability for recidivism
overall, but these outcomes varied depending on whether the offender failed on a reconviction or
technical violation. Parolees returning to transitional housing after the release were less likely to
fail on a reconviction, but this living situation actually increased likelihood of failure on a
technical violation, and shortened time to failure by half. Huebner and Berg (2011) point out that
offenders are often placed in transitional housing when they pose a significant risk to the
community or when they lack social support and capital. If transitional housing is a proxy for
weak networks of social support and capital, then this could possibly explain why the measure of
transitional housing is significantly related to risk of recidivism (Huebner & Berg, 2011). It is
suggested that further examination is needed in order to parse out differences between offenders
who fail by reconviction and those who fail by technical violation and the influence returning to
transitional housing upon release has on these differences.
Continuing to look at measures of post-release context, social relationships also proved to
be significant in their predictions of recidivism overall, but again, these measures depended the
type of failure. As a majority of participants returned to live with family, this particular living

49

situation was found to be a significant predictor, but only in the estimate of time to failure
(recidivism) for a technical violation. Surprisingly, this particular group of men displayed
slightly shortened times to failure by returning to family. Huebner and Berg (2011) reported
similar findings in their original analysis and have suggested that these results may imply those
who return to family might be more at risk to re-offend since they are less equipped with the
personal capital that is needed to successfully navigate the post-release period (Huebner & Berg,
2011). The fact that the current study yielded similar results for this measure calls for increased
examination to the assumption that offenders who return to social networks, especially one that
is kin-based, have relatively better chances of abstaining from crime after their release (Visher&
Travis, 2003). Although beyond the scope of this study, it is suggested that future studies
consider the quality of familial relationships, as well as the levels of strain and stigma returning
individuals can have on their families, in order to better understand the reentry processes
(Huebner & Berg, 2011).
Having dependent children also proved significant predictor in the risk for recidivism,
and these probabilities, again, are dependent on the type of failure. Being a parent significantly
decreased the probability for recidivism for the reconviction group, and these men were also
delayed in their time to failure. Conversely, the complete opposite holds true for offenders
eventually failing on a technical violation. These individuals experienced a much greater chance
for recidivism after release, and on top of that, a shortened time to failure. These findings lend
support to Huebner and Berg’s (2011) notion that potential bonds to children could be part of
long-term offender change and desistance. But being an offender and a parent at the same time
can have negative effects as well and previous research has highlighted some of the difficulties
and stress that can materialize from parent-child relationships (Braman, 2005; Slocum, Simpson,

50

& Smith, 2005). Generally, less than half of men live with their children prior to incarceration
(Glaze &Maruschak, 2008) but a majority of men are court-ordered to pay child support.
Therefore, it is reasonable to assume that this requirement can place further economic pressure
on men who are on parole supervision, ultimately increasing their odds of failure (Visher, La
Vigne, & Travis, 2004), especially for technical violations. It is a common knowledge that
parenting a child comes with great responsibilities. Perhaps parents whom recidivate on a
technical violation have a different relationship, bond and level of responsibility with their
children compared to those who fail on a reconviction. With such disparity in the probability of
recidivism and estimates of time to failure for this particular measure, more multilevel modeling
is needed to better assess how the quality and nature of parent-child relationships affect the
reentry process. And since nearly one third of parolees returning to rural counties reported being
a parent to at least one dependent child during time of the study, it is further suggested that future
models leave room for the rural perspective as well.
The number of parole offices available in a county also emerged as a significant predictor
for the risk of recidivism, at least in the present research site. It is important to note that Missouri
currently has 53 district parole offices spread across 115 counties. There are various satellite,
district, IPO, and sub office domiciles available in several counties, but it is unclear if any of
these locations are used for parole supervision and reporting services. The level of reentry
services can vary across different environments. Because the current study is more concerned
with rural environments where there typically is a lack of reentry services (Fitchen, 1991;
Leukefeld et al., 2002; Robertson, 1997) simply having more than one parole office in a county
proved to be an overall predictor for recidivism.

51

Burden (2009) previously examined location of parole office in her study of rural
parolees, however, results were not statistically significant for this measure. Creating a similar
measure to Burden, but one that is categorical in nature and not assessed spatially, results from
the current study do show some significance from this new measure. Having more than one
parole office within a single county (at least for the state of Missouri) decreased the likelihood
for overall recidivism, as well as for those who failed by receiving a new reconviction. However,
this measure was no longer statistically significant across prediction models for recidivism on a
technical violation. Like Burden (2009), this study also did not include more meaningful parole
supervision variables such as manpower and officer workload (Klinger, 1997), but it is agreed
upon that future studies should make it a point to include better measures of both parole and
police officers in rural communities. This may help to lend insight into the relationship between
supervision and recidivism in these particular areas.
Although this study attempted to measure access to public transportation by examining
the varying levels of bus service coverage, the results analyzing access to transportation and
level of service coverage were unpredictable in their ability to estimate probabilities of
recidivism and time to failure. Some models proved statistically significant, but with such
inconsistency, the current recommends that future analyses find more accurate measures of
access to public transportation. Therefore, the third and last set of hypotheses remains
unconfirmed. However, it is worthy to point out that while a majority of parolees (72.62%)
returning to rural counties experienced a complete lack of regular bus service coverage, only a
small amount of the sample (19.14%) had regular, daily bus service available to them. With
qualitative examinations and official commissions finding offender needs and challenges
associated with transportation (Bucks County CCAP Commission, 2012; Due et al., 1990;

52

Garland, Wodahl, & Mayfield, 2011; Myers & Associates, Inc., 2012), it is highly suggested that
future research on recidivism and prisoner reentry should focus on this usually-overlooked
service, as it is one that the public, including offenders, depend on.
More importantly, rurality was found to be statistically significant in predicting the
probability for overall recidivism, with results being particularly stronger for offenders failing on
a reconviction. Adding support to Burden’s (2009) previous findings, parolees who live in rural
areas were significantly more likely to recidivate than parolees who live in urban areas.
However, rurality failed to be statistically significant across logistic and survival models for
those failing on a technical violation. Previous research has suggested that rural communities
tend to have deeper friendship networks (Weisheit, Falcone, & Wells, 2006) and be less tolerant
of deviance (Wilkinson, 1991). It is therefore plausible that parolees who return to rural, less
populated areas “would find that the community they left not only remembered their past crimes,
but also would be quick to report any current involvements in deviant or criminal behavior”
(Burden, 2009: pg. 117). Again, rural social culture is different than social culture in urban areas
and these differences play out in distinct ways in regards to forming social networks and gaining
social capital (Weisheit & Wells, 1996). These differences also apply to returning offenders and
may influence probabilities for recidivism in various ways.
Also, criminal justice officers tend to respond to crime and delinquency differently
depending on the type of community they operate in (Burden, 2009; Stark, 1987; Klinger &
Bridges, 1997). Parolees can be returned to prison for a wide array of offenses ranging from
serious (e.g., committing a new burglary, assault, rape) to the more trivial, such as violations of
their parole conditions (e.g., drinking alcohol, spending time with known felons) (Burden, 2009).
Criminal justice officers operating in rural communities, including parole officers, have been

53

known to use more informal means of social control (Wilson, 1968; Weisheit, Falcone, & Wells,
2006) rather than a professional approach as typically seen in urban officers (Wilson, 1968). Due
to possible increased levels of discretion, and that parolees may be apprehended for a wider array
of offenses than the average individual (Burden, 2009), the current study insists for the need to
closely examine parole client-officer relationships in rural environments.
While these findings contribute to the literature in innovative ways, current limitations of
the study should be noted and discussed. First, the data sample examined is limited to adult
males incarcerated in a single state; therefore, one should be cautious in interpreting results as
generalizable. The particular research site may be an anomaly in levels of rurality, as well as
urbanism; in the level of access and frequency of public transportation; and in the number and
location of parole offices. Each of these variables can all drastically vary when compared to
other states or countries. Future research continuing to focus on the rural perspective should
examine similarly situated U.S. states for comparison, particularly other rural areas in the
Midwest, West, Southern, and Appalachian regions.
Relatedly, minority variables in the current study were unable to account for other racial
and ethnic groups, and future studies should therefore include more participants of color,
especially those who identify as being Hispanic/Latino, Asian, or Native American. Since the
current study sample reported very few Hispanic, Asian, and Native American offenders within
the sample population, the author chose to only included offenders who were racially identified
and coded as White/Caucasian or Black/African-American. Missouri currently reports a
statewide Hispanic/Latino population of 3 percent (U.S. Census Bureau, 2010c), and these
figures are even less for Asian and Native American residents. Thus, the racial and ethnic

54

1

makeup for this particular sample is limited . It is highly recommended that forthcoming
examinations of prisoner reentry include more diversity in their measures of race and ethnicity.
Finally, as only adult males were represented in this study, sources of risk for recidivism
among rural female offenders should also be examined. It is likely that women parolees returning
to rural communities have different experiences in their reentry process compared to women
returning to the city, and these differences can shape future recidivism outcomes. As it has been
touted by feminist criminologists that women enter gendered “pathways” to crime (Daly, 1992,
1994; Daly & Chesney-Lind, 1988) and therefore differ in offending experiences than men
(Davidson & Chesney-Lind, 2009), rural female offenders should also be closely examined for
their gender specific challenges and obstacles to reentry compared to rural male offenders.
Specifically, as women with dependent children may need to utilize public transportation
services in order to take their children to a doctor’s appointment, meet with social service agents,
and complete weekly errands such as grocery shopping. As Scroggins and Malley (2010) report
specific transportation problems as a challenge to reentry for women offenders in urban areas,
there is dire need for these types of transportation patterns to be examined across more rural
environments.
Another limitation of importance involves the use of official records as data on technical
violations and parole revocations can be highly subjective and vary substantially from state to
1

It is important to note that data from institutional records maintained by the department of
corrections included measures of race and ethnicity, however Huebner and Berg (2011) are apt to
point out the small population of Hispanic inmates present in their original study sample. As the
current study is using the same dataset, the remaining minute presence of men of color in the
study population restricted analysis outcomes by ethnicity. Before assuming for missing cases,
nearly all men were originally identified as of minority race in the sample are black (n = 1,214),
but the minority measure also includes 15 Native American and seven Asian men (Huebner &
Berg, 2011). Additionally, 60 men indicated that they were of Hispanic origin; 52 were
Hispanic-white, six Hispanic-black, one Hispanic-Native American, and one Hispanic-Asian
(Huebner & Berg, 2011).
55

state (Huebner & Berg, 2011). The current study is aware of this limitation and contends that as
future examinations expand to other rural areas, it is necessary to conduct in-depth qualitative
interviews with rural offenders in order to capture patterns of recidivism that may go unreported
or undetected by criminal justice officials. Furthermore, like Huebner and Berg (2011), these
data still do not reflect the full complexities of the incarceration and reentry experience. This
study failed to address measures of the quality of offenders’ social relationships, levels of postrelease employment, the treatment they may have received while incarcerated, and self-report
data on offenders’ overall incarceration experience.
Though some studies have found that personalized treatment programs available in prison
can positively affect recidivism (Cullen, 2002; Gaes, Flanagan, Motiuk, & Stewart, 1999), and
that these effects are further intensified when coupled with extensive aftercare (Hiller, Knight, &
Simpson, 1999), but these effects may not apply to rural offenders. Although treatment and
reentry services are generally lacking among rural communities (Fitchen, 1991; Leukefeld et al.,
2002; Robertson, 1997), even if these services were available, research has suggested that rural
offenders may be hesitant to utilize them due to the negative stigmas attached to their use by
community members (Rost, et al., 1993). Again, this is where in-depth qualitative research
would be able to fill some of these remaining gaps as to if and how rural offenders obtain
treatment and rehabilitative services once released, the quality of these services, the location of
these services, and the overall availability, compared to similar services provided in urban areas.
Unlike Huebner and Berg (2011), the current study did not specify any clear timing of
recidivism. Instead, timing was based simply on survival analysis estimates if time to failure was
shortened or delayed based on three different measures of rurality, as well as type of recidivism.
It is important to delineate between early, mid, and late timings in failure as it has been

56

suggested that desistance can best be viewed as an ongoing process (Bushway, Thornberry,
&Krohn, 2003; Laub& Sampson, 2001; Maruna, 2001; Uggen, Manza, & Behrens, 2003).
Timing was not in the scope of the current study, but scholars knowledgeable of the reentry
process would surely benefit from examining the timing of recidivism of rural offenders and
2

learning of their particular experiences in the desistance process .
A further limitation of importance is the inability to determine the exact address to where
study participants originally returned. Having a street address or coordinated return location
would allow for more advanced analysis utilizing Geographic Information Systems (GIS), which
could prove to be extremely beneficial to spatial and ecological studies on patterns of recidivism
and prisoner reentry. Addresses would also be able to serve as a reference check, as this study
was unable to determine how long participants stayed in their return locations, if they moved, or
if they still reside in the same location since the original time of study. Since rural offenders tend
not to be as transient as those residing in urban areas (Oser et al., 2006), it would be interesting
to see how many rural parolees have continued to reside in the same living situation they
originally returned to. With more information and time, a detailed multimethod and ecologically
based study of offenders returning to rural areas could be performed. Furthermore, the current
study utilizes measures on the county level, which offers its own difficulties in operationalization
and generalizability. Measures on this level may lead to a false sense of similarity within these

2

Huebner and Berg (2011) previously examined the timing of failure in their study. Additionally,
in order to better explore differences in recidivism timing, Huebner and Berg (2011) subdivided
study participants into three groups: early, mid, and late failure. Men who failed within the first
six months following release constituted as an early failure; those who failed between six months
and three years after release were considered mid failure; and men who failed either on a
reconviction or technical violation at any time greater than three years following the original
study period were identified as a late failure. Notably, it is particularly the recognition of late
failure that is unique to research on recidivism and desistance.

57

fairly large geographic units, and it is possible that certain towns or neighborhoods within
counties can be more rural and isolated than others (Bouffard & Muftić, 2006). It is important
that future studies look beyond the county level and more towards the neighborhood level,
especially if social disorganization is to remain the theoretical foundation in explaining variances
in rural recidivism outcomes.
Another limitation is that it is unknown if participants actually utilized public bus
services during the original time of study as this study was more concerned with the level of
access and frequency available to them in the research site. As some measures of accessibility to
public transportation proved to be statistically significant in predicting probabilities and
estimated time to failure in offenders, not all models were significant and the current study
demonstrated inconsistent and varied results. In order to better capture these measures, future
examinations should conduct in-depth qualitative interviews with rural offenders in order to gain
increased knowledge on their specific transportation patterns, habits, and needs. It is further
suggested that by combining qualitative and ethnographic methods along with empirical,
quantitative methods, future multilevel and multimethod examinations are needed and may help
provide more clarity on rural prisoner reentry.
Finally, and perhaps most important of all, is how the current study chose to define
3

rurality . Constituting exactly what is “rural” has proven a difficult challenge in scholarly
literature. To date, one of the most commonly used operational definitions of rural is by U.S.
Bureau of the Census standards. A place is deemed rural if it has a population of 2,500 or less,
and also includes all other populations, housing, and territory not located within a designated

3For a complete and detailed understanding what is meant by the term “rural”, see “What is

Rural” located in the Appendix of Weisheit, Falcone, and Wells’ (2006) Crime and Policing in
Rural and Small-Town America, 3rd Edition.
58

urban area (U.S. Census Bureau, 2010). The terms urban, urbanized area, urban cluster and
rural are all specifically defined by the Census Bureau; however, other agencies, whether
federal, state, local or private, may use these same terms to identify areas, but may be based on
entirely different criteria (Ricketts et al., 1998). It has even been suggested by some researchers
that this definition should only serve as a rough guide for distinguishing rural from urban as it
useful, but imprecise (Weisheit & Donnermeyer, 2000).
Additionally, rural communities can vary culturally, economically and socially (Weisheit
& Donnermeyer, 2000). Residents tend to be distrustful of outsiders, stress self-reliance, be less
supportive of public programs, and can often resist government involvement in their lives
(Beltrame, 1978; Coward, DeWeaver, Schmidt & Jackson, 1983; Keller & Murray, 1982;
Kenkel, 1986; Littrell&Littrell, 1991; Robertson & Donnermeyer, 1997; Seroka&Subramaniam,
1991; Swanson, Cohen & Swanson, 1979; Warner &Leukefeld, 2001; Weisheit & Donnermeyer,
2000). Due to such social diversity, making across-the-board generalizations about rural
communities should be avoided at all costs (Murray & Keller, 1991). It is important to strictly
define, operationalize, and measure rurality, but these measures need to look beyond simple
population counts and acknowledge how rural communities and its members are affected by their
surrounding landscape (Wodahl, 2006). Some scholars have even suggested that only when the
concept of rural is broadly defined that the true nature of prisoner reentry to rural communities
can be considered (Wodahl, 2006). In short, “rural” needs to be a flexible variable; one that is
precisely clear, but broad enough to include the unique elements and influences of rural
geography, culture and society.

59

Conclusion
While we do seem to have an overall idea of some of the specific challenges to reentry
and risks associated with recidivism, this knowledge tends to only aptly apply to the urban
offender. It is important that we also have similar amounts of knowledge from the rural
perspective in order to provide a broader sense of the reentry process. This study has not only
extended our knowledge of recidivism by focusing on parolees returning to rural counties, but
also by incorporating a new obstacle to reentry that is beginning to be recognized in the field:
transportation (Bucks County CCAP Commission, 2012; Due et al., 1990; Garland, Wodahl, &
Mayfield, 2011; Myers & Associates, Inc., 2012). Additionally, the current study also looked at
differences in probabilities of recidivism and estimated time to failure between reconvictions and
technical violations.
First, this study verifies that certain demographic, legal and environmental characteristics
influence the likelihood of recidivism, and that these influences react differently across rural and
urban environments. Confirming the first set of hypotheses, offenders returning to rural counties
showed higher probabilities for recidivating, as well as quicker times to failure. Confirming the
second set of hypotheses, differences in probabilities for recidivism were found between
offenders who failed on a reconviction and offenders who failed on a technical violation. As this
analysis is only a beginning into looking at these differences, it is recommended that further
reentry studies give more focus to the different methods, or types of recidivism. While the
current study was able to partially confirm the third set of hypotheses, results were too
inconsistent to make a fair assessment at this time. Therefore, an increased examination of how
transportation affects prisoner reentry and recidivism, especially among rural offender
populations, is warranted. This can be done through extensive multilevel methods that include

60

in-depth qualitative interview with rural offenders. Finally, this study also helps identify the
importance of parole office location. Because offenders are required to report to their parole
office on a regular basis, the location of these offices are critical, especially for offenders
residing in more rural communities. Future research should also look to incorporate the
geographic location of parole offices available to the particular study site in question and how
close offenders are located to the nearest office.
With an increase in the amount of adults released to some form of community
correctional supervision annually, exploration in areas of prisoner reentry and recidivism is still
needed. We have only barely begun to make a dent in the study of recidivism by solely
addressing the urban perspective, and much more knowledge is required from the rural
perspective. In addition, it is also important to continually recognize and address any new
potential challenges and obstacles to prisoner reentry as we advance during the 21st century. As
modes of public transportation have been made available for all to use, it is necessary to further
examine how offenders travel and navigate the reentry process and how this may pose a risk on
their chance for recidivism.

61

APPENDIX

62

Table 1: Comparing Variable Characteristics Between Missouri And The United States
Missouri
U.S.
General Characteristics
Population
5,988,927
Land area (square miles)
68,885.93
Population density (per square mile)
81.2
79.6
Median Age
37.9
37.2
Percent Male
49.0
49.2
Female-Headed Households
12.3
13.1
Racial/Ethnic Composition
Percent White
Percent Black
Percent Hispanic
Percent that speak second language
Percent native born
Economic Characteristics
Median family income
Percent below poverty
Percent unemployed
Percent on public assistance
Housing Characteristics
Percent vacant
Percent owner-occupied
Criminal Characteristics
Violent Crimes (per 100,000)
Homicide rate
Forcible rape rate
Robbery rate
Aggravated assault rate
Property Crimes (per 100,000)
Burglary rate
Larceny-theft rate
Motor vehicle theft rate

82.8
11.6
3.5
5.1
97.3

72.4
12.6
16.3
17.9
87.7

$59,020
14.3
8.1
14.6

$64,293
14.3
8.7
15.9

12.4
68.8

11.4
65.1

6.1
24.3
104.3
312.7

4.7
26.8
113.7
241.1

745.7
2308.3
254.8

702.2
1976.9
229.6

Data obtained from U.S. Decennial Census (2010) and FBI Uniform Crime Reports (2011)

63

Table 2: Description of Variables Measured in Study
Type of Factor
Dependent variable
Recidivism

Independent variables
Demographic characteristics

Legal characteristics

Environmental characteristics
Release setting
Living situation

Rurality

Population density

Variables

Source

Year

MDOC

1998 - 2006

(1)
(2)
(3)
(4)

Age (at release)
Minority
Dependent children
High school
graduate/GED
(5) Stable work history
(6) Stable mental health

MDOC

1998 - 2006

(1) Number of prior
convictions
(2) Current offense
(3) Property
(4) Personal
(5) Drug
(6) Other
(7) Salient factor score
(8) Number of
misconduct violations
(9) Time served in prison
(days)

MDOC

1998 - 2006

(1) Drug dependent

MDOC

1998 - 2006

(1) Sustained marriage
(2) Returned to family
(3) Returned to
transitional housing
(4) Enhanced supervision
(5) Concentrated
disadvantage

MDOC
U.S. Census
Bureau

1998 – 2006
2000

(1) Metro (RUCC 1-3)
(2) Nonmetro (RUCC 49)
Number of people per square
mile.

Economic
Research
Service
U.S. Census
Bureau

2013

Returned to prison through
reconviction or revocation
of parole due to technical
violation.

64

2000

Table 2 (cont’d)
Type of Factor

Variables

Parole office location

(1) Has contiguous county
office
(2) Has county office
(3) Has more than 1 county
office

Type of public
transportation available

(1)
(2)
(3)
(4)
(5)
(6)

Has OATS
Has SMTS
Has City
Has University
Has County
Has Other

Source

Year

MDOC, Division
of Probation and
Parole

2013

Missouri
Department of
Transportation

2013

Missouri Public
Transportation
Association
American Public
Transportation
Association

Level of access

(1) No coverage (Has OATS
or SMTS)
(2) Partial coverage (Has
University, County or
Other)
(3) Regular coverage (Has
City)

Missouri
Department of
Transportation
Missouri Public
Transportation
Association
American Public
Transportation
Association

65

2013

Table 3: Descriptive Statistics for Study Sample (n= 3,296)
All Returning
Parolees
(n= 3,296)
Mean
Dependent variable
Recidivated (1 = yes)
Reconviction
Technical violation
Independent variables
Demographic characteristics
Age at release
Minority (1 = Black)
Dependent children (1 = one or more)
High school graduate (1 = yes)
Stable work history (1 = yes)
Stable mental health (1 = yes)
Legal characteristics
Number of prior convictions
Current offense
Property
Personal
Drug
Other
Salient factor score
Number of misconduct violations
Time served in prison (days)
*p< .05; **p < .01; ***p < .001

66

Returning to
Rural
(n= 840 )

Returning to
Urban
(n= 2,456)

SD

66.7%
34.7%
32.1%

.47
.48

70.47%
49.18%
29.64%

65.47%**
46.87%***
32.90%

32.47
33.19%
33.28%
55.37%
25.79%
93.11%

9.53
.47

31.59
11.55%
34.4%
55.95%
26.55%
92.5%

32.76**
40.59%***
32.90%
55.17%
25.53%
93.32%

1.64

2.11

1.50

40.84%
23.82%
19.60%
15.75%
7.41
6.86
768.74

.49
.43
.39
.36

44.88%
19.52%
15.60%
20.00%
7.51
5.46
590.53

8.49
871.52

1.68*
39.35%**
25.29%***
20.97%***
14.29%***
7.37
7.34***
829.68***

Table 3 (cont’d)
All Returning
Parolees
(n = 3,296)
Mean

Returning to
Rural
(n = 840)

Returning to
Urban
(n = 2,456)

22.86%

21.5%

15.24%
15.48%
52.38%
8.21%

12.05%*
13.84%
47.80%*
19.91%***

1898.68

32.86%
-0.41
25.49%
62.71

20.24%***
0.14***
74.51%
1398.39***

SD

Independent Variables
Environmental characteristics
Release setting
Drug dependent (1 = yes)

21.84%

Sustained marriage (1 = yes)
Intimate partner (1 = yes)
Returned to family (1 = yes)
Returned to transitional
housing (1 = yes)
Enhanced supervision
Concentrated disadvantage

12.86%
14.26%
48.97%
16.93%

.33

23.45%
.000

.42
.99

Living situation

Rurality
Population density
Parole office location

1057.99

.50
.38

Has contiguous county office
Has county office
Has more than one county
office

13.77%
40.26%
45.96%

.34
.49
.50

35.36%
31.92%
0%

6.39%***
64.64%***
61.69%***

No regular coverage
Partial coverage
Regular coverage

26.24%
59.50%
14.26%

.45
.39
.45

72.62%
27.38%
0.00%

10.38%***
70.48%***
19.14%***

Level of access

*p< .05; **p < .01; ***p < .001

67

Table 4: Logistic Regression Models of Recidivism for Reconviction and Technical Violation
Model 1
Model 2
B
SE
OR
B
SE
Demographic characteristics
Age at release
-.037*** .005
.964
-.037*** .005
Minority (Black)
.488*** .107
1.628
.435*** .106
Dependent children
.071
.088
1.073
.075
.088
High school graduate
-.143
.084
.867
-.154
.084
Stable work history
.087
.096
1.091
.084
.096
Stable mental health
.100
.163
1.105
.110
.162
Legal characteristics
Number of prior convictions
.023
.021
1.023
.019
.021
Property offense (current)
.221
.120
1.247
.215
.120
Drug offense (current)
-.260*
.128
.771
-.263*
.128
Other offense (current)
-.177
.143
.838
-.188
.143
Salient factor score
-.117*** .020
.889
-.118*** .020
Misconduct violations
.034*** .007
1.035
.035*** .007
Time served in prison (days)
.000*** .000
1.000
.000*** .000
Environmental characteristics
Drug dependent
.784*** .109
2.190
.789*** .109
Sustained marriage
-.970*** .120
.379
-.959*** .120
Returned to family
.083
.082
1.087
.072
.082
Transitional housing
.333**
.121
1.395
.339** .122
Enhanced supervision
.036
.100
1.036
.042
.100
Concentrated disadvantage
-.057
.056
.945
-.030
.057
Population Density
.000**
.000
1.000
.000** .000
Measures of Rurality
Rurality (Model 1)
.348*** .100
1.416
Regular coverage (Model 2)
1.465*
.612
More than one parole office (Model 3)
Model Fit
2
.190
.188
Nagelkerke R
*p< .05; **p < .01; ***p < .001

68

Model 3
SE

OR

B

OR

.964
1.545
1.078
.858
1.087
1.116

-.037***
.501***
.080
-.144
.090
.122

.005
.109
.088
.084
.096
.162

.964
1.650
1.083
.866
1.094
1.130

1.019
1.240
.769
.828
.899
1.035
1.000

.022
.215
-.265*
-.188
-.114***
.035***
.000***

.021
.120
.128
.143
.020
.007
.000

1.023
1.240
.767
.829
.892
1.036
1.000

2.200
.383
1.075
1.404
1.043
.970
1.000

.785***
-.965***
.089
.342**
.046
-.021
.000

.109
.120
.082
.121
.100
.057
.000

2.192
.381
1.093
1.408
1.047
.980
1.000

-.314**

.104

.731

4.328

.189

Table 5: Logistic Regression Models of Recidivism for Reconviction Only
Model 1
B
SE
OR
Demographic characteristics
Age at release
-.023*** .005
.978
Minority (Black)
.361*** .101 1.435
Dependent children
-.510*** .089
.601
High school graduate
.004
.081 1.004
Stable work history
.238*
.096 1.268
Stable mental health
.252
.158 1.287
Legal characteristics
Number of prior convictions
.120*** .020 1.127
Property offense (current)
-.175
.113
.839
Drug offense (current)
-.474*** .129
.622
Other offense (current)
-.186
.143
.831
Salient factor score
-.089*** .019
.915
Misconduct violations
.030*** .006 1.030
Time served in prison (days)
.000*** .000 1.000
Environmental characteristics
Drug dependent
.278** .091 1.320
Sustained marriage
-.734*** .143
.480
Returned to family
-.075
.079
.927
Transitional housing
-.293** .116
.746
Enhanced supervision
-.098
.095
.907
Concentrated disadvantage
-.017
.055
.983
Population Density
.000
.000 1.000
Measures of Rurality
Rurality (Model 1)
.395*** .094 1.484
Regular coverage (Model 2)
More than one parole office (Model 3)
Model Fit
2
.135
Nagelkerke R
*p< .05; **p < .01; ***p < .001

69

B

Model 2
SE

OR

B

Model 3
SE

OR

-.023***
.295**
-.503***
-.007
.233*
.250

.005
.099
.089
.081
.096
.157

.977
1.343
.605
.993
1.263
1.284

-.022***
.382***
-.499***
.005
.239*
.268

.005
.102
.089
.081
.096
.157

.978
1.465
.607
1.005
1.270
1.307

.115***
-.178
-.479***
-.191
-.091***
.030***
.000***

.020
.113
.129
.143
.019
.006
.000

1.122
.837
.619
.826
.913
1.030
1.000

.120***
-.182
-.480***
-.197
-.086***
.030***
.000***

.020
.113
.129
.143
.019
.006
.000

1.127
.834
.619
.821
.918
1.031
1.000

.285**
-.721***
-.091
-.289**
-.091
.014
.000**

.091
.143
.079
.117
.095
.055
.000

1.330
.486
.913
.749
.913
1.014
1.000

.281**
-.728***
-.070
-.285**
-.095
.033
.000

.091
.143
.079
.117
.095
.056
.000

1.325
.483
.932
.752
.909
1.003
1.000

1.619**

.610

5.047
-.409***

.102

.664

.131

.134

Table 6: Logistic Regression Models of Recidivism for Technical Violation Only
Model 1
Model 2
B
SE
OR
B
SE
Demographic characteristics
Age at release
-.017** .005
.983
-.017** .005
Minority (Black)
.117
1.00 1.124
.131
.098
Dependent children
.587*** .084 1.799
.586*** .084
High school graduate
-.116
.080
.890
-.114
.080
Stable work history
-.161
.097
.851
-.160
.097
Stable mental health
-.120
.152
.887
-.120
.152
Legal characteristics
Number of prior convictions
-.106*** .023
.899
-.106*** .023
Property offense (current)
.379** .115 1.461
.379** .115
Drug offense (current)
.219
.127 1.245
.221
.127
Other offense (current)
.025
.146 1.026
.026
.147
Salient factor score
-.033
.019
.968
-.032
.019
Misconduct violations
.001
.006 1.001
.001
.006
Time served in prison (days)
.000
.000 1.000
.000
.000
Environmental characteristics
Drug dependent
.360*** .091 1.433
.358*** .091
Sustained marriage
-.447** .129
.639
-.451** .129
Returned to family
.142
.078 1.153
.143
.079
Transitional housing
.542*** .109 1.720
.545*** .110
Enhanced supervision
.135
.096 1.144
.132
.096
Concentrated disadvantage
-.026
.053
.974
-.030
.054
Population Density
.000
.000 1.000
.000
.000
Measures of Rurality
Rurality (Model 1)
-.093
.096
.912
Regular coverage (Model 2)
-.209
.588
More than one parole office (Model 3)
Model Fit
2
.075
.075
Nagelkerke R
*p< .05; **p < .01; ***p < .001

70

OR

B

Model 3
SE

OR

.983
1.140
1.796
.892
.852
.887

-.017**
.105
.584***
-.117
-.162
-.127

.005
.101
.084
.080
.097
.152

.983
1.110
1.794
.889
.850
.881

.900
1.461
1.247
1.026
.968
1.001
1.000

-.107***
.382**
.220
.031
-.033
.001
.000

.023
.115
.127
.147
.019
.006
.000

.899
1.465
1.246
1.032
.967
1.001
1.000

1.431
.637
1.154
1.724
1.141
.970
1.00

.358***
-.449***
.141
.536***
.135
-.041
.000

.091
.129
.078
.109
.096
.054
.000

1.431
.638
1.151
1.709
1.145
.960
1.000

.100

1.137

.811
.128
.075

Table 7: Survival Analysis of Time to Failure to Recidivism for Reconviction and Technical Violation
Model 1
Model 2
B
SE
OR
B
SE
OR
Demographic characteristics
Age at release
-.024*** .003
.976
-.024*** .003
.976
Minority (Black)
.213*** .056 1.238
.184**
.055 1.202
Dependent children
.144**
.048 1.155
.146**
.048 1.158
High school graduate
-.061
.045
.941
-.065
.045
.937
Stable work history
.032
.054 1.032
.028
.054 1.029
Stable mental health
-.046
.086
.955
-.049
.086
.952
Legal characteristics
Number of prior convictions
.014
.011 1.014
.012
.011 1.012
Property offense (current)
.044
.063 1.045
.045
.063 1.046
Drug offense (current)
-.233**
.073
.792
-.236**
.072
.790
Other offense (current)
-.204*
.082
.815
-.209*
.082
.811
Salient factor score
-.078*** .011
.925
-.079*** .011
.925
Misconduct violations
.020*** .003 1.020
.020*** .003 1.020
Time served in prison (days)
.000*** .000 1.000
.000*** .000 1.000
Environmental characteristics
Drug dependent
.369*** .049 1.446
.372*** .049 1.451
Sustained marriage
-.582*** .081
.559
-.574*** .081
.563
Returned to family
.054
.044 1.055
.049
.044 1.051
Transitional housing
.240*** .062 1.271
.236*** .062 1.266
Enhanced supervision
.038
.053 1.038
.040
.053 1.041
Concentrated disadvantage
-.035
.030
.965
-.025
.031
.975
Population Density
.000
.000 1.000
.000
.000 1.000
Measures of Rurality
Rurality (Model 1)
.152**
.053 1.165
Regular coverage (Model 2)
.472
.331 1.603
More than one parole office (Model 3)
Model Fit
-2 Log likelihood
33,071.178
33,077.404
*p< .05; **p < .01; ***p < .001

71

B

Model 3
SE

OR

-.024***
.225***
.147**
-.059
.029
-.044

.003
.056
.048
.045
.054
.086

.976
1.252
1.158
.942
1.030
.957

.013
.038
-.235**
-.215**
-.077***
.020***
.000***

.011
.063
.073
.082
.011
.003
.000

1.013
1.039
.790
.807
.926
1.020
1.000

.377***
-.577***
.055
.246***
.038
-.015
.000

.049
.081
.044
.062
.053
.031
.000

1.458
.562
1.057
1.278
1.039
.985
1.000

-.175**

.057

.839

33,069.787

Table 8: Survival Analysis of Time to Failure to Recidivism for Reconviction Only
Model 1
Model 2
B
SE
OR
B
SE
Demographic characteristics
Age at release
-.023*** .004
.977
-.024***
.004
Minority (Black)
.282*** .007 1.326
.223**
.076
Dependent children
-.229**
.072
.795
-.225**
.072
High school graduate
-.013
.062
.987
-.021
.062
Stable work history
.154*
.074 1.166
.145*
.074
Stable mental health
.077
.124 1.080
.068
.124
Legal characteristics
Number of prior convictions
.073*** .014 1.075
.069***
.014
Property offense (current)
-.137
.085
.872
-.134
.085
Drug offense (current)
-.448*** .102
.639
-.451***
.102
Other offense (current)
-.260*
.111
.771
-.267*
.111
Salient factor score
-.089*** .014
.915
-.090***
.014
Misconduct violations
.027*** .004 1.028
.027***
.004
Time served in prison (days)
.000*** .000 1.000
.000***
.000
Environmental characteristics
Drug dependent
.303*** .068 1.354
.311***
.068
Sustained marriage
-.691*** .124
.501
-.680***
.124
Returned to family
-.027
.061
.973
-.039
.061
Transitional housing
-.051
.093
.951
-.056
.093
Enhanced supervision
-.020
.073
.981
-.006
.043
Concentrated disadvantage
-.026
.043
.974
-.006
.043
Population Density
.000
.000 1.000
.000*
.000
Measures of Rurality
Rurality (Model 1)
.282*** .071 1.326
Regular coverage (Model 2)
.982*
.476
More than one parole office (Model 3)
Model Fit
-2 Log likelihood
17,178.827
17,189.929
*p< .05; **p < .01; ***p < .001

72

Model 3
SE

OR

B

OR

.977
1.249
.799
.979
1.156
1.070

-.023***
.299***
-.225**
-.013
.148*
.079

.004
.078
.072
.063
.074
.124

.977
1.349
.799
.987
1.159
1.082

1.072
.875
.637
.765
.914
1.028
1.000

.071***
-.148
-.451***
-.277*
-.087***
.028***
.000***

.014
.085
.102
.111
.014
.004
.000

1.074
.862
.637
.758
.917
1.028
1.000

1.365
.506
.962
.945
.994
.994
1.000

.315***
-.686***
-.024
-.044
-.015
.011
.000

.068
.124
.061
.093
.072
.043
.000

1.370
.503
.976
.957
.985
1.011
1.000

-.319***

.080

.727

2.670

17,177.792

Table 9: Survival Analysis of Time to Failure to Recidivism for Technical Violation Only
Model 1
Model 2
B
SE
OR
B
SE
Demographic characteristics
Age at release
-.025***
.004
.976
-.025*** .004
Minority (Black)
.147
.080 1.158
.147
.079
Dependent children
.489***
.066 1.631
.489*** .066
High school graduate
-.104
.064
.901
-.104
.064
Stable work history
-.094
.080
.911
-.094
.080
Stable mental health
-.158
.119
.854
-.158
.119
Legal characteristics
Number of prior convictions
-.066***
.019
.937
-.065*** .019
Property offense (current)
.241*
.94
1.272
.241*
.094
Drug offense (current)
-.013
.104
.987
-.014
.104
Other offense (current)
-.162
.122
.851
-.161
.122
Salient factor score
-.067***
.016
.935
-.067*** .016
Misconduct violations
.013**
.104
.987
.013**
.005
Time served (days)
.000***
.000 1.000
.000*** .000
Environmental characteristics
Drug dependent
.436***
.071 1.546
.435*** .071
Sustained marriage
-.478***
.107
.620
-.478*** .107
Returned to family
.142*
.063 1.152
.142*
.064
Transitional housing
.506***
.085 1.659
.506*** .085
Enhanced supervision
.104
.078 1.109
.104
.078
Concentrated disadvantage
-.040
.043
.960
-.041
.043
Population Density
.000
.000 1.000
.000
.000
Measures of Rurality
Rurality (Model 1)
-.002
.079
.998
Regular coverage (Model 2)
-.034
.463
More than one parole office (Model 3)
Model Fit
-2 Log likelihood
15,717.218
15,717.214
*p< .05; **p < .01; ***p < .001

73

Model 3
SE

OR

B

OR

.976
1.158
1.631
.901
.911
.854

-.025***
.152
-.489***
-.104
-.094
-.157

.004
.081
.066
.064
.080
.119

.976
1.164
1.631
.902
.910
.855

.937
1.272
.987
.851
.935
1.013
1.000

-.066***
.240*
-.013
-.163
-.067***
.013**
.000

.019
.094
.104
.122
.016
.005
.000

.937
1.271
.987
.849
.936
1.013
1.000

1.546
.620
1.153
1.658
1.110
.960
1.000

.437***
-.478***
.142*
.510***
.102
-.038
.000

.071
.107
.063
.085
.078
.044
.000

1.548
.620
1.152
1.665
1.108
.963
1.000

-.023

.081

.977

.967

15,717.140

REFERENCES

74

REFERENCES

Allen, J. (1988). Administering quality education in an adult correctional facility.
Community Services Catalyst, 18(4), 28-29.
Amato, P. R. (1993). Urban-rural differences in helping friends and family.Social Psychology
Quarterly, 56(4), 249–262.
Andrews, D. A., &Bonta, J. ([1994], 2003). The Psychology of Criminal Conduct, 3rd Edition.
Cincinnati: Anderson.
Austin, J. (2001). Prisoner reentry: Current trends, practices, and issues. Crime &
Delinquency, 47(3), 314-334.
Barnett, C., & Mencken, F. C. (2002).Social Disorganization Theory and the Contextual Nature
of Crime in Nonmetropolitan Counties*. Rural Sociology, 67(3), 372-393.
Batiuk, M. (1997). The state of post secondary correctional education in Ohio. Journal of
Correctional Education, 48(2), 70-72.
Beggs, J., Hurlbert, J., & Haines, V. (1996). Community Attachment in a Rural Setting: A
Refinement and Empirical Test of the Systemic Model. Rural Sociology, 61(3), 407-426.
Bellair, P. (1997). Social Interaction and Community Crime: Examining the Importance of
Neighbor Networks. Criminology 35(4), 677–703.
Beltrame, T. F. (1978). Meeting the special needs of Appalachian alcoholics. Hospital and
Community Psychiatry, 29(12), 792-794.
Berg, M. T., & Huebner, B. M. (2011). Reentry and the ties that bind: An examination of social
ties, employment, and recidivism. Justice Quarterly, 28(2), 382-410.
Blumstein, A., & Beck, A. J. (2005). Reentry as a Transient States between Liberty and
Recommitment.In J. Travis & C. Visher (Eds.), Prisoner reentry and crime inAmerica,
pp. 50-79. New York: Cambridge University Press.
Bottoms, A. (1994). Environmental criminology. In R. Morgan (Ed.), The Oxford Handbook of
Criminology, pp. unknown. London: Oxford University Press.
Bouffard, L. A., & Muftić, L. R. (2006). The “rural mystique”: Social disorganization and
violence beyond urban communities. Western Criminology Review, 7(3), 56-66.
Braithwaite, J. (1989). Crime, shame, and reintegration. London: Cambridge University Press.

75

Braman, D. (2005). Doing time on the outside: Incarceration and family life in urban America.
Ann Arbor: University of Michigan Press.
Brooks, L. E., Solomon, A. L., Keegan, S., Kohl, R., &Lahue, L. (2005).Prisoner Reentry in
Massachusetts (Research Report). Washington, DC: The Urban Institute, Urban Institute
Justice Policy Center. [Online]. Available:
http://www.urban.org/UploadedPDF/411167_Prisoner_Reentry_MA.pdf.
Brown, D. M. (1996). Rural America’s Transportation Network: Issues for the 1990s. Rural
Development Perspectives, 11(2), 10-17.
Brown, D. M. (2008).Public transportation on the move in rural America. Washington, DC:
Economic Research Service, U.S. Department of Agriculture. [Online]. Available:
http://www.nal.usda.gov/ric/ricpubs/publictrans.htm.
Bucks County CCAP Committee. (2012). CCAP Committee on County Criminal Justice System
Best Practices for the 21st Century. Doylestown: County of Bucks Department of
Corrections, Pennsylvania.
Burden, F. F. (2009). Examining Parolees in Their Communities: Poverty, Rurality, and
Criminal Justice Resources (Doctoral dissertation, Pennsylvania State University).
Bursik, R. J. (1988). Social Disorganization and Theories of Crime and Delinquency: Problems
and Prospects. Criminology, 26(4), 519–551.
Bursik, R. J., &Grasmick, H. G. (1993). Neighborhoods and crime: The dimensions of effective
community control. New York: Lexington Books.
Bushway, S. D., Piquero, A., Broidy, L. M., Cauffman, E., &Mazerolle, P. (2001). An empirical
framework for studying desistance as a process.Criminology, 39(2), 491–515.
Bushway, S., Thornberry, T. P., &Krohn, M. D. (2003). Desistance as a developmentalprocess:
A comparison of static and dynamic approaches. Journal of Quantitative Criminology, 19
(2), 129–153.
Carson, E. A., &Sabol, W. J. (2012). Prisoners in 2011 (NCJ 239808). Washington, DC: U.S.
Department of Justice, Office of Justice Programs, Bureau of Justice Statistics.
Chaiken J. M. &Chaiken, M. R. (1990). Drugs and predatory crime. In M. Tonry& J. Q. Wilson
(Eds.), Crime and Justice: A Review of Research, Vol. 13, pp. 203–240. Chicago:
University of Chicago Press.
City of Joplin MAPS Transit. (2010). General Guidelines for MAPS Transit & The Sunshine
Lamp Trolley.[PDF file].Available from
https://www.freemanhealth.com/upload/docs/Maps/2010%20maps%20added%20to%20s
ite/Trolley_Brochure2%5B1%5D.pdf.

76

Clear, T., Rose, D., Warring, E., & Scully, K. (2003). Coercive mobility and crime: A
preliminary examination of concentrated incarceration and social disorganization. Justice
Quarterly, 20(1), 33–64.
Clinard, M. B. (1942).The process of urbanization and criminal behavior.American Journal of
Sociology, 48(2), 202-213.
Conger, R. D. (1997).The special nature of rural America. In E. B. Robertson, Z. Sloboda, G. M.
Boyd, & L. Beatty (Eds.), Rural substance abuse: State of knowledge and issues, pp. 3752. Rockville: National Institute on Drug Abuse.
Coward, R. T., DeWeaver, K. L., Schmidt, F. E., & Jackson, R. W. (1983). Distinctive features
of rural environments: A frame of reference for mental health practice. International
Journal of Mental Health, 12(Spring-Summer 1 & 2), 3-24.
Cox, D. R. (1972).Regression models and life tables.Journal of the Royal Statistical Society,
34(2), 187–220.
Crutchfield, R. D., &Pitchford, S. R. (1997). Work and crime: The effects of labor stratification.
Social Forces, 76(1), 93-118.
Crutchfield, R. D., Geerken, M. R. & Gove, W. R. (1982). Crime Rate and Social Integration.
Criminology, 20(3-4), 467–478.
Cullen, F. T. (2002).Rehabilitation and treatment programs. In J. Q. Wilson & J. Petersilia
(Eds.), Crime: Public policies for crime control (pp. 253–290). Oakland: ICS Press.
Cumberland, A. K. & Boyle, G. J. (1997). Psychometric prediction of recidivism: Utility of the
Risk Needs Inventory. Australian and New Zealand Journal of Criminology, 30(1), 7286.
Daly, K. (1992). Women’s pathways to felony court: Feminist theories of lawbreaking and
problems ofrepresentation. Southern California Review of Law and Women’s Studies, 2,
11-52.
Daly, K. (1994). Gender, crime, and punishment. New Haven, CT: Yale University Press.
Daly, K., & Chesney-Lind, M. (1988). Feminism and criminology.Justice Quarterly, 5, 497-538.
Davidson, J. T., & Chesney-Lind, M. (2009). Discounting women: Context matters in risk and
need assessment. Critical Criminology, 17(4), 221-245.
DeJong, C. (1997). Survival analysis and specific deterrence: Integrating theoretical and
empirical models of recidivism. Criminology, 35(4), 561–575.
DeLisi, M. (2003). Criminal careers behind bars. Behavioral Science and the Law, 21,653–661.

77

Due, J. F., Allen, B. J., Kihl, M. R., & Crum, M. R. (1990).Transportation Service to Small
Rural Communities: Effects of Deregulation. Ames: Iowa State University Press.
Economic Resource Center. (2004). Rural poverty at a glance (Rural Development Research
Report Number 100). Washington, DC: U.S. Department of Agriculture. [Online].
Available: http://www.ers.usda.gov/publications/rdrr100/rdrr100_lowres.pdf.
Economic Resource Center. (2005). Rural transportation at a glance (Agriculture Information
Bulletin Number 795]. Washington, DC: U.S. Department of Agriculture. [Online].
Available: http://www.ers. usda.gov/publications/AIB795_lowres.pdf.
Economic Resource Center. (2013). Rural-Urban Continuum Codes.Washington, DC: U.S.
Department of Agriculture. [Online]. Available: http://www.ers.usda.gov/dataproducts/rural-urban-continuum-codes/.aspx#.UjbX42TF08o.
Fagan, J. (1989). Cessation of family violence: Deterrence and dissuasion. In L. Ohlin & M.
Tonry (Eds.), Crime and Justice: A Review of Research, Vol. 11,pp. 427-480. Chicago:
University of Chicago Press.
Fagan, J. (1990). Intoxication and aggression. In M. Tonry& J. Q. Wilson (Eds.), Drugs and
Crime. Crime and Justice: A Review of Research, Vol. 13, pp. 241-320. Chicago:
University of Chicago Press.
Farrington, D. P., & West, D. J. (1995). Effects of marriage separation, and children on
offending by adult males.In J. Hagan (Ed.), Current perspectives on aging and thelife
cycle, Vol. 4, pp. 249–281. Greenwich: JAI Press.
Federal Bureau of Investigation (FBI). (2011). Uniform crime reports: Crime in the United
States. Washington, DC: U.S. Government Printing Office.
Fischer, C. S. (1982). To dwell among friends: Personal networks in town and city. Chicago:
University of Chicago Press.
Fitchen, J. M. (1991). Homelessness in rural places: Perspectives from upstate New York. Urban
Anthropology, 20(2), 177-210.
Freudenberg, W. & Jones, R. (1991). Criminal Behavior and Rapid Community Growth:
Examining the Evidence. Rural Sociology, 56(4), 619–645.
Gaes, G. D., Flanagan, T. J., Motiuk, L., & Stewart, L. (1999). Adult correctional treatment. In
M. Tonry& J. Petersilia (Eds.), Prisons (Vol. 26, pp. 361–426). Chicago:University of
Chicago Press.

78

Garland, B., Wodahl, E. J., & Mayfield, J. (2011). Prisoner reentry in a small metropolitan
community: Obstacles and policy recommendations. Criminal Justice Policy
Review, 22(1), 90-110.
Garrett, M. & Taylor, B. (1999). Reconsidering social equity in public transit. Berkeley Planning
Journal, 13(1), 6-27.
Gatersleben, B. &Uzzell, D. (2007). Affective appraisals of the daily commute. Environment and
Behavior, 20(3), 416-431.
Gendreau, P., Goggin, C., & Smith, P. (2002). Is the PCL-R really the “unparalleled” measure of
offender risk? Criminal Justice and Behavior, 29, 397–426.
Gendreau, P., Little, T., &Goggin, C. (1996). A meta-analysis of the predictors of adult offender
recidivism: What works!.Criminology, 34(4), 575–607.
Ghelfi, L. M., &McGranahan, D. A. (2004). One in five rural counties depends on farming.
Amber Waves, 2(3), 11.
Gibbs, R., Kusmin, L., &Cromartie, J. (2004). Low-skill jobs: A shrinking wave of the rural
economy. Amber Waves, 2(5), 38-44.
Glaze, L. E., &Bonczar, T. P. (2011).Probation and parole in the United States, 2010 (NCJ
236019). Washington, DC: US Department of Justice, Office of Justice Programs, Bureau
of Justice Statistics.
Glaze, L. E., &Maruschak, L. M. (2008).Parents in prison and their minor children.Washington,
DC: Bureau of Justice Statistics.
Glenn, N. O., & Hill, L. (1977) Rural-urban differences in attitudes and behavior in the United
States.The ANNALS of the American Academy of Political and Social Science, 429(1),
36-50.
Gottfredson, M. R. (1990). A general theory of crime. Stanford: Stanford University Press.
Gottfredson, S. D., & Gottfredson, D. M. (1994). Behavioral prediction and the problems of
incapacitation.Criminology, 32(3), 441-474.
Gottschalk, M. (2010). Cell blocks & red ink: mass incarceration, the great recession & penal
reform. Daedalus, 139(3), 62-73.
Gruidl, J. J., Merrett, C. D., & Connelly, D. (2007). You can’t get there from here: Issues in rural
passenger transportation. In J. F. Plant, V. R. Johnston, & C. E. Ciocirlan (Eds.),
Handbook of transportation policy and administration, pp. 239-252. Boca Raton: CRC
Press.

79

Harlow, C. W. (2003). Education and correctional populations (NCJ 195670). Washington, DC:
U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics.
Harrison, L. (2001). The revolving door for drug-involved offenders: Challenges and
opportunities. Crime and Delinquency, 47(3), 462–485.
Hattery, A. J., & Smith, E. (2010). Prisoner reentry and social capital: The long road to
reintegration. Lanham, MD: Lexington Books.
Hiller, M., Knight, K., & Simpson, D. D. (1999). Prison-based substance abuse
treatment,residential aftercare and recidivism. Addiction, 94, 833–842.
Hofferth, S. L., & Iceland, J. (1998). Social capital in rural and urban communities.Rural
Sociology, 63(4), 574– 598.
Hogg, R. & Carrington, K. (1998). Crime, rurality and community.The Australian and New
Zealand Journal of Criminology, 31(2), 160–181.
Holzer, H. J., Raphael, S., & Stoll, M. A. (2004). Will employers hire former offenders?
Employer preferences, background checks, and their determinants. In M. Pattillo, D. F.
Weiman, & B. Western (Eds.), Imprisoning America: The social effects of mass
incarceration, pp. 205, 210-211. New York: Russell Sage Foundation.
Horney, J., Osgood, D. W., & Marshall, I. H. (1995). Criminal careers in the short-term: Intraindividual variability in crime and its relation to local life circumstances.
AmericanSociological Review, 60(5), 655–673.
Housing Assistance Council. (2003). Rural Research Note: Rental housing in rural America.
Washington: DC. [Online]. Available:
http://www.ruralhome.org/storage/research_notes/rrn_rural_rental_housing.pdf.
Huebner, B. M. & Berg, M. T. (2011). Examining the Sources of Variation in Risk for
Recidivism. Justice Quarterly, 28(1), 146-173.
Huebner, B. M., Varano, S. P., & Bynum, T. S. (2007). Gangs, guns, and drugs: Recidivism
among serious offenders. Criminology & Public Policy, 15(4), 437–461.
Huggins, C. M. (2009). Returning Home: Residential mobility, neighborhood context and
recidivism (Doctoral dissertation, Ohio State University).
Hughes, T. A., Wilson, D. J., & Beck, A. J. (2001). Trends in state parole, 1990–2000 (NCJ
184735). Washington, DC: U.S. Department of Justice, Office of Justice Programs,
Bureau of Justice Statistics.
Jobes, P. C. (1999). Residential Stability and Crime in Small Rural Agricultural and Recreational
Towns. Sociological Perspectives, 42(3), 498–524.

80

Keller, P. A. & Murray, J. D. (1982).Handbook of rural community mental health. New York:
Human Sciences Press.
Kenkel, M. B. (1986).Stress-coping-support in rural communities: A model for primary
prevention. American Journal of Community Psychology, 14(5), 457-478.
Kerbs, J. J., Jones, M., &Jolley, J. M. (2009).Discretionary Decision Making by Probation and
Parole Officers The Role of Extralegal Variables as Predictors of Responses to Technical
Violations.Journal of Contemporary Criminal Justice, 25(4), 424-441.
Klinger, D. A. (1997).Negotiating order in patrol work.Criminology, 35, 277-306.
Klinger, D. A., & Bridges, G. S. (1997). Measurement error in calls-for-service as an indicator of
crime.Criminology, 35, 705-726.
Knight, K., Simpson, D. D., &Moyey, J. T. (2002).An evaluation of the TCU drug screen.Fort
Worth: Institution for Behavioral Research, Texas Christian University.
Kornhauser, R. 1978.Social Sources of Delinquency: An Appraisal of Analytic Models. Chicago:
University of Chicago Press.
Kowalski, G. & Duffield, D. (1990). The Impact of the Rural Population Component on
Homicide Rates in the United States: A County-Level Analysis. Rural Sociology, 55(1),
76–90.
Kposowa, A. &Breault, K. (1993). Reassessing the Structural Covariates of U.S. Homicide
Rates: A County Level Study. Sociological Focus, 26(1), 27–46.
Kposowa, A., Breault, K., & Harrison, B. (1995). Reassessing the Structural Covariates of
Violent and Property Crimes in the USA: A County Level Analysis. The British Journal
of Sociology, 46(1), 79–105.
Kroner, D. G. & Mills, J. F. (2001).The accuracy of five risk appraisal instruments in predicting
institutional misconduct and new convictions.Criminal Justice and Behavior, 28(4), 471489.
Kubrin, C., & Stewart, E. A. (2006). Predicting who reoffends: The neglected role of
neighborhood context in recidivism studies. Criminology, 44(1), 165–197.
La Vigne, N. G., Visher, C., & Castro, J. (2004). Chicago prisoners' experiences returning
home. Washington, DC: The Urban Institute Press. [Online]. Available:
http://www.urban.org/UploadedPDF/311115_ChicagoPrisoners.pdf.
Land, K. C., McCall, P., & Cohen, L. E. (1990). Structural covariates of homicide rates: Are
there any invariances across time and space? The American Journal of Sociology, 95(4),
922-963.

81

Langan, P. A., & Levin, D. J. (2002). Recidivism of prisoners released in 1994 (NCJ 193427).
Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of
Justice Statistics.
Laub, J. H., Nagin, D. S., & Sampson, R. (1998). Trajectories of change in criminal offending:
Good marriages and the desistance process. American Sociological Review, 63(2), 225–
238.
Laub, J. H., & Sampson, R. (2001). Understanding desistence from crime. In M. Tonry& N.
Morris (Eds.), Crime and Justice: A Review of Research, Vol. 28, pp. 1–69. Chicago:
University of Chicago Press.
Laub, J. H., & Sampson, R. (2003). Shared beginnings, divergent lives: Delinquent boys to age
70. Boston: Harvard University Press.
Lee, M. R., Maume, M. O., &Ousey, G. C. (2003). Social Isolation and Lethal Violence Across
the Metro/Nonmetro Divide: The Effects of Socioeconomic Disadvantage and Poverty
Concentration on Homicide. Rural Sociology, 68(1), 107-131.
Lerner, M. J. (1977).The Effectiveness of a Definite Sentence Parole Program.Criminology,
15(2), 211-224.
Leukefeld, C. G., Narevic, E., Hiller, M. L., Staton, M., Logan, T. K., Gillespie, W., Webster, J.
M., Garrity, T. F. & Purvis, R. (2002). Alcohol and Drug Use Among Rural and Urban
Incarcerated Substance Abusers. International Journal of Offender Therapy and
Comparative Criminology, 46(6), 715-728.
Littrell, D. W. &Littrell, D. P. (1991). Civic education, rural development, and the Land Grant
institutions. In K. E. Pigg (Ed.), The future of rural America: Anticipating policies for
constructive change, pp. 195-212. Boulder: Westview Press.
Loeber, R., & Le Blanc, M. (1990). Toward a developmental criminology. In M. Tonry& N.
Morris (Eds.), Crime and Justice: A Review of Research, Vol. 12, pp. 375–473. Chicago:
University of Chicago Press.
Long, J. S. (1997). Regression models for categorical and limited dependent variables.Thousand
Oaks: Sage.
Lowenkamp, C. T., Holsinger, A. M., &Latessa, E. J. (2001). Risk/need assessment, offender
classification and the role of child abuse.Criminal Justice and Behavior, 10(5), 269-284.
Loza, W. &Loza-Fanous, A. (2001). Predictive validity of the self-appraisal questionnaire
(SAQ): A tool for assessing violent and nonviolent release failures. Journal of
Interpersonal Violence, 15(11), 1183-1191.

82

Lynch, J. P., &Sabol, W. J. (2001). Prisoner reentry in perspective (Crime Policy Report Vol. 3).
Washington, DC: The Urban Institute, Urban Institute Justice Policy Center. [Online].
Available: http://www.urban.org/UploadedPDF/410213_reentry.pdf.
Mallik-Kane, K., &Visher, C. A. (2008).Health and Prisoner Reentry: How Physical, Mental,
and Substance Abuse Conditions Shape the Process of Reintegration (Research Report).
Washington, DC: The Urban Institute, Urban Institute Justice Policy Center. [Online].
Available: http://www.urban.org/UploadedPDF/411617_health_prisoner_reentry.pdf.
Maltz, M. D. (1984). Recidivism. Orlando: Academic Press.
Maruna, S. (2001).Making good: How ex-convicts reform and rebuild their lives. Washington,
DC: American Psychological Association.
Matsueda, R. L., Gartner, R., Piliavin I., and Polakowski, M. (1992). The prestige of criminal
and conventional occupations: A subcultural model of criminal activity.
AmericanSociological Review, 57, 752–770.
Miethe T. D., & McDowell, D. (1993). Contextual effects in models of criminal
victimization.Social Forces, 71(3), 741-759.
Missouri Public Transportation Association. (2013). Missouri Public Transportation
Association. Retrieved May 5, 2013, from
http://mptaonline.typepad.com/missouri_public_transit_a/missouris-public-transpor.html.
Missouri Rural Public Transportation Providers. (2013). Rural Public Transportation.Missouri
Department of Transportation. Retrieved May 5, 2013 from
http://www.modot.org/othertransportation/transit/rptransportation.htm.
Motiuk, L. L., Bonta, J., & Andrews, D. A. (1986). Classification in correctional halfway houses:
The relative and incremental predictive criterion validities of the Megargee-MMPI and
LSI systems. Criminal Justice and Behavior, 13(1), 33-46.
Motor Carrier Ratemaking Study Commission, Collective Ratemaking in the Trucking Industry.
(1983). A Report to the President and the Congress of the United States. Washington,
DC: 98th Congress.
Mumola, C. J., &Karberg, J. C. (2006). Drug Use And Dependence, State And Federal
Prisoners, 2004 (NCJ 213530).Washington, DC: U.S. Department of Justice, Office of Justice
Programs, Bureau of Justice Statistics.
Murray, A. T., Davis, R., Stimson, R. J., & Ferreira, L. (1998). Public transportation
access. Transportation Research Part D: Transport and Environment, 3(5), 319-328.
Murray, J. D. & Keller, P. A. (1991). Psychology and Rural America: Current Status and Future
Directions. American Psychologist, 46(3), 220-231.

83

Myers, D. T. and Associates, Inc. (2012). Housing Needs Study for the York County Criminal
Justice Advisory Board. York County, Pennsylvania.[Online]. Available:
http://www.dma-housing.com/wp-content/uploads/2012/01/York-Co.-Crim.-JusticeReport.pdf.
National Research Council. (2008). Parole, Desistance from Crime, and Community
Integration.Committee on Community Supervision and Desistance from Crime.
Washington, DC: National Academies Press.
Needles, K. E. (1996). Go directly to jail and do not collect? A long-term study of recidivism,
employment, and earning patterns among prison releasees.Journal ofResearch in Crime
and Delinquency, 33(4), 471-496.
Nelson, M., Deess, P., & Allen, C. (1999). The first month out: Post-incarceration experiences in
New York City. Federal Sentencing Reporter, 24(1), 72-75.
Nutley, S. (2005). Monitoring rural travel behaviour: A longitudinal study in Northern Ireland
1979-2001. Journal of Urban Transport Geography, 13(3), 247-263.
O’Brien, P. (2001). “Just like baking a cake”: Women describe the necessary ingredients for
successful reentry after incarceration. Families in Society: Journal of Contemporary
Human Services, 82(3), 287-295.
O'Connor, M. E., & Gray, D. E. (1989). Crime in a rural community. Sydney: Federation Press.
O’Keefe, M., Klebe, K. J., &Hromas, C. S. (1998). Validation of the Level of Suppression
Inventory (LSI) for community-based offenders in Colorado: Phase II. Colorado Springs:
Colorado Department of Corrections.
O’Sullivan, A. (2007). Urban economics. New York: McGraw-Hill.
OATS, Inc. (2013). Bus Schedules. [Online]. Available: http://www.oatstransit.org/busschedules.
Oser, C. B., Biebel, E. P., Havens, J. R., Staton-Tindall, M., Knudsen, H. K., Mooney, J. L.,
&Leukefeld, C. G. (2009). Inmate Prerelease Assessment (IPASS) Aftercare Placement
Recommendation as a Predictor of Rural Inmate's 12-Step Attendance and Treatment
Entry Postrelease. Journal of Offender Rehabilitation, 48(8), 725-743.
Osgood, D. W., & Chambers, J. M. (2000). Social disorganization outside the metropolis:An
analysis of rural youth violence. Criminology, 38(1), 81-115.
Ouimet, M., & Le Blanc, M. (1996).The role of life experiences in the continuation of the adult
criminal career.Criminal Behaviour and Mental Health, 6(1), 73–97.
Pager, D. (2003).The mark of a criminal record.American Journal of Sociology, 108(5),937-975.

84

Palen, J. J. (1977). The Urban Nexus: Toward The Year 2000. In A. Hawley (Ed.), Societal
Growth, pp. 141-156. New York: The Free Press.
Petee, T. & Kowalski, G. (1993). Modeling Rural Violent Crime Rates: A Test of
SocialDisorganization Theory. Sociological Focus, 26(1), 87–89.
Petersilia, J. (2001). Prisoner Reentry: Public Safety and Reintegration Challenges. The Prison
Journal, 81(3): 360-375.
Petersilia, J. (2003). When prisoners come home: Parole and prisoner reentry. New York:
Oxford University Press.
Petersilia, J. & Turner, S. (1991). An evaluation of intensive probation in California.Journal of
Criminal Law and Criminology, 82(3), 610–658.
Pew Center on the States. (2008). One in 100: Behind bars in America 2008. Washington, DC:
Pew Charitable Trusts.
Pucher, J. &Renne, J. L. (2005). Rural mobility and mode choice: Evidence from the 2001
National Household Travel Survey. Transportation, 32(2), 165-186.
Raynor, P., Kynch, J., Roberts, C., &Merrington, S. (2000). Risk and need assessment in
probation services: an evaluation. Home Office Research Study No. 211. London: Home
Office.
Rees, G. &Wragg, R. (1975). A Study of the Passenger Transport Needs of Rural Wales. Cardiff:
Welsh Council.
Reisig, M. D. &Cancino, J. M. (2004). Incivilities in Nonmetropolitan Communities: The Effects
of Structural Constraints, Social Conditions, and Crime. Journal of Criminal Justice,
32(1), 15–29.
Rephann, T. J. (1999).Links between rural development and crime. Papers in Regional
Science, 78(4), 365-386.
Richie, B. E. (2001). Challenges incarcerated women face as they return to their communities:
Findings from life history interviews. Crime and Delinquency, 47(3), 368-389.
Ricketts, T. C., Johnson-Webb, K. D., & Taylor, P. (1998). Definitions of rural: A handbook for
health policy makers and researchers. Bethesda: Federal Office of Rural Health Policy.
Ritter, S. E. (1997).Exploring the relationship between prison misconduct andRecidivism
(Doctoral dissertation, Sam Houston State University).
Robertson, E. B. (1997). Introduction: Intervention and Services. In E. B. Robertson, Z. Sloboda,
G. M. Boyd, and L. Beatty (Eds.), Rural Substance Abuse: State of Knowledge and
Issues, pp. 246-249. Rockville: National Institute on Drug Abuse.

85

Robertson, E. B. & Donnermeyer, J. F. (1997). Illegal Drug Use Among Rural Adults: Mental
Health Consequences and Treatment Utilization. American Journal of Drug and Alcohol
Abuse, 23(3), 467-484.
Rost, K., Smith, R., & Taylor, L. (1993). Rural-urban differences in stigma and the use of care
for depressive disorders.Journal of Rural Health, 9(1), 57-62.
Rubel, T. (1990). Future of Rural Transportation.In Proceedings of the 76th Annual Road
School (Held at Purdue University, March 6-8), 161, pp. 55-60.
Sagarin, E., Donnermeyer, J.F., & Carter, T.J. (1982). Crime in the countryside - A prologue. In
T.J. Carter, G.H. Phillips, J.F. Donnermeyer, & T.N. Wurschmidt (Eds.), Rural crime, pp.
10-19. Totowa: Allanheld, Osmun, & Co. Publishers, Inc.
Sampson, R. J. (1985). Neighborhood and crime: The structural determinants of
personalvictimization. Journal of Research in Crime and Delinquency, 22(1), 7-40.
Sampson, R. J. (1992). Family management and child development: Insights from social
disorganization theory. In J. McCord (Ed.), Advances in criminological theory, Vol.3, 6393. Piscataway: Transaction Publishers.
Sampson, R. J., & Groves, W. B. (1989). Community structure and crime: Testing
socialdisorganization theory. American Journal of Sociology, 94(4), 774-802.
Sampson, R. J., &Laub, J. H. (1993). Crime in the making: Pathways and turning pointsthrough
life. Cambridge: Harvard University Press.
Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: A
multi-level study of collective efficacy. Science, 277(5328), 918-924.
Sarkisian, N., &Gerstel, N. (2004). Kin support among Blacks and Whites: Race and family
organization. American Sociological Review, 69(6), 812-837.
Scheer, S. D., Borden, L. M., & Donnermeyer, J. F. (2000). The relationship between family
factors and adolescent substance use in rural, suburban, and urban settings. Journal of
Child and Family Studies, 9(1), 105-115.
Schmidt, P., & Witte, A. D. (1988). Predicting recidivism using survival models. New
York: Springer.
Schoeneberger, M. L., Leukefeld, C. G., Hiller, M. L., &Godlaski, T. (2006). Substance Abuse
Among Rural and Very Rural Drug Users at Treatment Entry*. The American Journal of
Dug and Alcohol Abuse, 32(1), 87–110.
Scroggins, J. R., &Malley, S. (2010). Reentry and the (unmet) needs of women. Journal of
Offender Rehabilitation, 49(2), 146-163.

86

Seroka, J. &Subramaniam, S. (1991). Governing the countryside: Linking politics and
administrative resources. In K. E. Pigg (Ed.), The future of rural America: Anticipating
policies for constructive change, 213-231. Boulder: Westview Press.
Shaw, C. R., & McKay, H. D. (1942). Juvenile delinquency and urban areas. Chicago:
University of Chicago Press.
Shover, N. (1996).Great pretenders: Pursuits and careers of persistent thieves. Boulder:
Westview.
Shover, N., & Thompson, C. (1992).Age, differential expectations, and crime
desistance.Criminology, 30(1), 89–104.
Singer, J. D., & Willett, J. B. (2003). Applied longitudinal data analysis: Modeling change and
event occurrence. New York: Oxford University Press.
Slocum, L. A., Simpson, S., & Smith, D. A. (2005). Strained lives and crime: Examining intraindividual variation in strain and offending in a sample of incarcerated women.
Criminology, 43, 1067–1110.
Southeast Missouri Transportation Service, Inc. (SMTS). (2013). Services & Area.[Online].
Available: http://ridesmts.org/services.html.
Solomon, A. L., Gouvis, C., &Waul, M. (2001). Summary of Focus Group with Ex-Prisoners in
the District: Ingredients for Successful Reintegration. Washington, DC: The Urban
Institute, Urban Institute Justice Policy Center. [Online]. Available:
http://www.urban.org/UploadedPDF/410492_ExPrisoners.pdf.
Spano, R. & Nagy, S. (2005). Social Guardianship and Social Isolation: An Application and
Extension of Lifestyle/Routine Activities to Rural Adolescents. Rural Sociology, 70(3),
414–437.
Spohn, C., &Holleran, D. (2002). The effect of imprisonment on recidivism rates offelony
offenders: A focus on drug offenders. Criminology, 40(2), 329–357.
Stark, R. (1987). Deviant places: A theory of the ecology of crime. Criminology, 25(4),893-909.
Stark, R., Bainbridge, W., Crutchfield, R., Doyle, D., & Finke, R. (1983).Crime andDelinquency
in the Roaring Twenties.Journal of Research in Crime and Delinquency, 20(1), 4–23.
Stern, E. (1979).Inter-nodal association of bus services. GeoJournal, 3(1), 89-96.
Stern, E. (1982). Bus Services in Rural Areas A Consumers' Evaluation. Environment and
Behavior, 14(1), 94-112.
Sullivan, M. (1989). Getting paid: Youth crime and work in the inner city. Ithaca:Cornell

87

University Press.
Swanson, B. E., Cohen, R. A., & Swanson, E. P. (1979). Small towns and small towners: A
framework for survival and growth. Beverly Hills: Sage Publications.
Taylor, R. B. (1997). Social order and disorder of street blocks and neighborhoods:Ecology,
microecology, and the systemic model of social disorganization. Journal of Research in
Crime and Delinquency, 34(1), 113-155.
Travis, J. & Lawrence, S. (2002). Beyond the prison gates: The state of parole in America
(Research Report). Washington, DC: The Urban Institute, Urban Institute Justice Policy
Center. [Online]. Available:
http://www.urban.org/UploadedPDF/310583_Beyond_prison_gates.pdf.
Travis, J. & Petersilia, J. (2001). Reentry Reconsidered: A New Look at an Old Question. Crime
and Delinquency, 47(3): 291-313.
Travis, J., Solomon, A. L., &Waul, M. (2001). From Prison to Home: The Dimensions and
Consequences of Prisoner Reentry. Washington, DC: The Urban Institute, Urban Institute
Justice Policy Center. [Online]. Available:
http://www.urban.org/UploadedPDF/from_prison_to_home.pdf.
Uggen, C., Manza, J., & Behrens, A. (2003). Stigma, role transition, and the civic reinte- gration
of convicted felons. In S. Maruna& R. Immarigeon (Eds.), After crime and punishment:
Ex-offender reintegration and desistance from crime (pp. 258–290). New York: State
University of New York Press.
Uggen, C., & Thompson, M. (2003). The socioeconomic determinants of ill-gotten gains:
Within-person changes in drug use and illegal earnings. American Journal of Sociology,
109(1), 146–185.
U.S. Census Bureau. (2000). Urban/Rural and Metropolitan/Nonmetropolitan Population.U.S.
Department of Commerce.[Online]. Available: http://factfinder2.census.gov.
U.S. Census Bureau. (2010a). Missouri: Profile of General Population and Housing
Characteristics: 2010 Demographic Profile Data.[Online]. Available:
http://factfinder2.census.gov.
U.S. Census Bureau. (2010b). United States: Profile of General Population and Housing
Characteristics: 2010 Demographic Profile Data.[Online]. Available:
http://factfinder2.census.gov.
Vacca, J. (2004). Educated prisoners are less likely to return to prison. Journal of Correctional
Education, 55(4), 297-305.

88

Vesterby, M., Bucholtz, S., Baez, A., & Roberts, M. J. (2006). Major uses of land in the United
States, 2002. Wash Washington, DC: Economic Research Service, U.S. Department of
Agriculture. [Online]. Available:
http://www.farmlandinfo.org/documents/37124/eib14.pdf.
Visher, C. A., Kachnowski, V., La Vigne, N. G., & Travis, J. (2004).Returning Home:
Understanding the Challenges of Prisoner Reentry Maryland Pilot Study: Findings from
Baltimore (Research Report). Washington, DC: The Urban Institute, Urban Institute
Justice Policy Center. [Online]. Available:
http://www.urban.org/UploadedPDF/410974_ReturningHome_MD.pdf.
Visher, C. A., & Travis, J. (2003). Transitions from prison to community: Understanding
individual pathways. Annual Review of Sociology, 29, 89–113.
Vose, B., Cullen, F. T., & Smith, P. (2008). The empirical status of the level of service
inventory.Federal Probation, 72, 22–29.
Waite, L. J., &Spitze, G. (1981). Young women’s transition to marriage.Demography,18(4),
681–694.
Warner, B. D. &Leukefeld, C. G. (2001).Rural-urban differences in substance use and treatment
utilization among prisoners.American Journal of Drug and Alcohol Abuse, 27(2), 265280.
Warr, M. (1998). Life-course transitions and desistance from crime.Criminology, 36(2), 183–
216.
Weisheit, R. A. & Donnermeyer, J. F. (2000).Change and Continuity in Crime in Rural America.
In G. LaFree (Ed.), The Nature of Crime: Continuity and Change, pp. 309-358.
Washington, DC: U.S. Department of Justice.
Weisheit, R. A., Falcone, D. N., & Wells, L. E. (1995). Crime and policing in rural and smalltown America: An overview of the issues (NIJ Research Report). Washington, DC: U.S.
Department of Justice, Office of Justice Programs, National Institute of Justice.
Weisheit, R. A., Falcone, D. N. and Wells, L. E. (2006) Crime andPolicing in Rural and SmallTown America. 3rd Edition. Prospect Heights: Waveland Press.
Weisheit, R. A., & Wells, L. E. (1996). Rural crime and justice: Implications for theory and
research. Crime and Delinquency, 42(3), 379-397.
Wells, L. E. & Weisheit, R. A. (2004). Patterns of Rural and Urban Crime: A County Level
Comparison. Criminal Justice Review, 29(1), 1–22.
Western, B., & Beckett, K. (1999). How unregulated is the U.S. labor market: The penal system
as a labor market institution. American Journal of Sociology, 104(4), 1030-1060.

89

Western, B., &McLanahan, S. (2000). Fathers in fragile families: The impact of incarceration on
family formation. In G. L. Fox (Ed.), Families and Crime, pp. 307-322. London: Oxford
University Press.
Western, B., Kling, J. R., &Weiman, D. F. (2001). The labor market consequences of
incarceration.Crime & Delinquency, 47(3), 410–427.
White, P. R. (1976). Planning for public transport. London: Hutchinson.
Wilkinson, K. P. (1984). A research note on homicide and rurality.Social Forces, 63(2), 445452.
Wilkinson, K. P. (1991). The community in rural America. New York: Greenwood Press.
Wilkinson, K., Thompson, J., Reynolds, R. &Ostresh, L. (1982). Local Social Disruption and
Western Energy Development: A Critical Review. The Pacific Sociological Review,
25(3), 275-296.
Wilkinson, K., Reynolds, R., Thompson, J., &Ostresh, L. M. (1984).Violent Crime in the
Western Energy-Development Region. Sociological Perspectives, 27(2), 241–256.
Wilson, J. Q. (1968). Varieties of police behavior. Cambridge: Harvard University Press.
Wilson, W. J. (1997).When Work Disappears: The World of the New Urban Poor. New York:
Vintage.
Wirth, L. (1938). Urbanism as a way of life.American Journal of Sociology, 44(1), 1-24.
Wodahl, E. J. (2006). The Challenges of Prisoner Reentry from a Rural Perspective. Western
Criminology Review, 7(2), 32-47.
Wright, R. T., & Decker, S. H. (1997). Armed robbers in action: Stickups and streetculture.
Boston: Northeastern University Press.
Yahner, J., &Visher, C. (2008). Illinois prisoners’ reentry success three years
afterrelease.Washington, DC: Urban Institute.

90