Research That Matters (January 17 - 20, 2008)

Approximately 630,000 offenders were released from prison in 2006, affecting families and communities throughout the United States, particularly in lower-income areas. A recent national recidivism study found that 68 percent of offenders were re-arrested within three years of release and 25 percent were re-incarcerated for committing a new crime. Over 50 percent were re-incarcerated when including re-incarceration for parole revocation.

Considering these high re-incarceration rates, it is important to understand how life trajectories of recidivists and trajectories of offenders who abstain from continuing a criminal lifestyle differ. The life-course theory provides a conceptual framework for understanding how and why offenders desist from committing additional crimes by theorizing that social activities such as obtaining employment, seeking educational pursuits, and getting married decrease the likelihood of further criminal activity.

The specific aim is to assess the influence of social bonding variables on recidivism and recidivists' time crime-free in the community based on the attachments to social bonds offenders develop upon release from prison.

The study is a secondary data analysis on a random selection of 250 Texas parolees released from prison since 2001. The researcher conducted Cox proportional hazard models to assess the influence of social bonding variables on offenders' likelihood of re-incarceration and recidivists' time crime-free before re-incarceration while controlling for race, age, number of prior offenses, and sentence length.

The researcher conducted three Cox regression models. The first model included all recidivists and censored cases. The p-value for the chi-square change (2.205) from Block 1 (control variables) and Block 2 (inclusion of independent variables) was p=.524, indicating the life-course variables did not contribute to the model. The second model excluded censored cases to assess if life-course variables extended recidivists' time crime-free in the community. The chi-square change from the block with only the control variables and the block including the elements of life-course theory (15.512) was statistically significant (p=.004). One of the three predictor variables significantly reduced the monthly hazard ratio. The p-value of the Wald Statistic for employment (13.487) was p<.001. The value of the odds ratio for employment (.315) indicates that the recidivism hazard was reduced by 68.5 percent for employed offenders. The third model conducted the same analysis for offenders younger than 37 years old and also found that employment extends time crime-free in the community before re-incarceration.

Results indicate that employment appears to provide a temporary desistance from crime, but its influence may diminish over time. The primary elements of life-course theory did not decrease the likelihood of recidivating, but employment did delay re-incarceration. Considering most recidivism studies find the offender to be most at-risk during the first year upon release, employment could potentially enable the offender to remain crime-free during this hazardous period.

After social workers and case managers help their clients released from prison obtain employment, they should continuously measure their clients' motivation levels regarding the desire to remain crime free. Employment does not seem to engender a permanent desistance from criminal behavior.