Revisiting the Effect of Visitation on Recidivism

Abstract

Scholarship suggests that prison inmates who are visited may be less likely to recidivate. Questions exist, however, about whether the observed relationship is causal and, if so, whether it is consistent for different groups of inmates. To address these questions, this study employs two methodological approaches – first, conventional regression analyses and, second, instrumental variable (IV) analyses – to examine the effects of visitation in general and specifically for females, young inmates, and individuals incarcerated for the first time. Effect size estimates are similar across the two analytic approaches, but conventional regression analyses identify a statistically significant effect of visitation, whereas IV analyses do not. Subgroup analyses suggest differences between males and females and by age. Combined, the results raise questions about whether visitation exerts a causal effect on offending. Implications for theory, research, and policy of the divergent results and the potential for a generalized visitation effect are discussed.

Keywords:

• Incarceration
• social ties
• visitation
• recidivism
• instrumental variable

Acknowledgments

We thank the Florida Department of Corrections for providing the data, Avi Bhati and Daniel Nagin for discussions about distance as a potential instrumental variable, Claudia Anderson for editorial assistance, and Shawn Bushway and Sarah Tahamont for guidance on related work. We also thank the Editor and three anonymous reviewers for the helpful feedback and advice for strengthening the manuscript. A version of this manuscript was presented at the 2017 annual conference for the European Society of Criminology in Cardiff, Wales. The views expressed in this paper are those of the authors alone, not the Florida Department of Corrections.

Notes

1 Two hundred and seventy cases were dropped from the original sample due to discrepancies in the inmates’ movement records that made it impossible to identify precisely when they were moved out of an intake facility and into a prison facility.

2 However, it appears that more than half of inmates in Florida stay in the same facility over the course of their stay. In our sample, 9723 out of 16,289 (59.7%) inmates did not experience any distance change over the course of their stay. Thus, for nearly 60% of our sample, the visitation measure is capturing all of their visits.

3 We conducted ancillary analyses that included county-level economic disadvantage and index crime rates to account for any specific effects of social disadvantage on the relationships between distance and visitation and recidivism. The economic disadvantage was an index measure using four 2,000 Census measures: percent female headed households, percent below the poverty line, unemployment rate, and percent of households on public assistance. The index crime rate used county-level data from the Uniform Crime Reports. Results from the full sample analysis indicated significant associations between economic disadvantage and recidivism and index crime rates and recidivism. However, the estimated effects of visitation on recidivism were substantively the same as those shown in the paper. These results are available upon request.

4 As a sensitivity check, we re-estimated all models using a conventional probit and IV probit. All results were substantively the same to those reported in the text. Because the probit model imposes stricter assumptions than the linear model, we proceed with linear models for both the conventional regression and the IV estimation. Linear models, in this context, can be interpreted as the linear probability model (Long, 1997).

5 We will refer to visitation in the first facility simply as “visitation” going forward.

6 An IV analysis will almost always produce larger standard errors than conventional regression because the standard error in the IV estimation is adjusted (almost always upward) by the inverse of the covariance between the endogenous predictor (visitation) and the instrumental variable(s) (distance; see Wooldridge, 2010:109).

7 This played out as expected when we compared the transfers short-term inmates experienced relative to long-term inmates. Only 30% of short-term inmates transferred at least once, but 47% of long-term inmates transferred at least once.

Additional information

Notes on contributors

Joshua C. Cochran

Joshua C. Cochran, PhD, is an Assistant Professor at the University of Cincinnati, School of Criminal Justice, Cincinnati, OH, USA. His research interests include criminological theory, imprisonment, and sentencing. His work has appeared in Criminology, the Journal of Quantitative Criminology, the Journal of Research in Crime and Delinquency, Justice Quarterly, and in the book Prisoner Reentry in the Era of Mass Incarceration (Sage).

J. C. Barnes

J.C. Barnes, PhD, is an Associate Professor in the School of Criminal Justice at the University of Cincinnati, Cincinnati, OH, USA. His research seeks to understand how genetic and environmental factors combine to impact criminological phenomena. He also has interest in applied statistical analysis.

Daniel P. Mears

Daniel P. Mears, PhD, is the Mark C. Stafford Professor of Criminology at Florida State University’s College of Criminology and Criminal Justice, Tallahassee, FL, USA. He conducts research on crime and policy. His work has appeared in such journals as Criminology and Justice Quarterly, and in the books Out-of-Control Criminal Justice (Cambridge University Press), American Criminal Justice Policy (Cambridge University Press), and Prisoner Reentry in the Era of Mass Incarceration (Sage).

William D. Bales

William D. Bales, PhD, is a Professor at Florida State University’s College of Criminology and Criminal Justice, Tallahassee, FL, USA. Dr. Bales focuses on a range of crime and policy topics, including factors that contribute to recidivism, the effectiveness of electronic monitoring, and tests of labeling theory. He has published in Criminology, Criminology and Public Policy, Justice Quarterly, and other crime and policy journals.