Michigan Study of Life After Prison, Administrative Data on 2003 Cohort of Michigan Parolees (ICPSR 32681)

Version Date: Mar 28, 2019

Cite this study |

Principal Investigator(s):
Jeffrey D. Morenoff, University of Michigan; David J. Harding, University of Michigan

https://doi.org/10.3886/ICPSR32681.v1

Version V1

Slide tabs to view more

Summary

These data are part of NACJD's Fast Track Release and are distributed as they were received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except for the removal of direct identifiers. Users should refer to the accompanying readme file for a brief description of files available with this collection and consult the investigator(s) if further information is needed.

The Michigan Study of Life After Prison examined the association between neighborhood context and outcomes related to employment and recidivism among the cohort of former prisoners released on parole from Michigan state prisons in one calendar year (2003), controlling for pre-incarceration neighborhood context, local labor market conditions, and a large set of individual characteristics. The primary goals of this study were to answer two questions: (1) "Are ex-offenders who are released to more disadvantaged neighborhoods (those with greater poverty, unemployment, residential turnover, etc.) more likely to recidivate?" (2) "Are ex-offenders who are released to more disadvantaged neighborhoods less likely to gain stable employment?" This research sought to supplement available literature on prisoner reentry and criminal desistance, which the researchers posit existing literature has largely ignored the role that neighborhoods play in shaping the recidivism and employment of returning prisoners.

The 31 data files included as part of this collection are as follows:

Cleaned Data Files:

  • casenotearrestsreps1-4_ICPSR-EDITED.dta: 4,932 Cases, 12 Variables
  • casenotearrestsreps5-8_ICPSR-EDITED.dta: 5,302 Cases, 13 Variables
  • casenotearrestsrep9_ICPSR-EDITED.dta: 2,321 Cases, 13 Variables
  • casenoteemploymentreps1-4_ICPSR-EDITED.dta: 4,871 Cases, 28 Variables
  • casenoteemploymentreps5-8_ICPSR-EDITED.dta: 4,754 Cases, 23 Variables
  • casenoteemploymentrep9_ICPSR-EDITED.dta: 2,610 Cases, 23 Variables
  • cleanedcasenoteaddressesreps1-8_ICPSR-EDITED.dta: 50,207 Cases, 72 Variables
  • cleanedcasenoteaddressesrep9_ICPSR-EDITED.dta: 10,309 Cases, 69 Variables
  • preprisonaddress_all_ICPSR-EDITED.dta: 5,183 Cases, 30 Variables
  • preprisonaddress_all_rep9_ICPSR-EDITED.dta: 1,017 Cases, 63 Variables
  • postprisads_ICPSR-EDITED.dta: 11,064 Cases, 41 Variables
  • cleaned-demographics-population_ICPSR-EDITED.dta: 11,064 Cases, 57 Variables
  • simplecrimhistory.dta: 11,064 Cases, 4 Variables
  • popSAhistory.dta: 11,064 Cases, 8 Variables
  • deathdates_ICPSR-EDITED.dta: 308 Cases, 3 Variables
  • popprisonenterdates.dta: 11,064 Cases, 7 Variables
  • discharge dates.dta: 7,369 Cases, 5 Variables
  • parole and release dates for pop.dta: 11,064 cases, 3 Variables
  • mdoc_recidivism_measures.dta: 11,064 Cases, 6 Variables
  • recidivism dates from transits.dta: 11,064 Cases, 8 Variables
  • recidivism from bir.dta: 11,064 Cases, 3 Variables
  • sample marker.dta: 3,689 Cases, 2 Variables
  • samplereps.dta: 3,689 Cases, 2 Variables
  • tta_rsid_rep.dta: 1,363 Cases, 2 Variables
    •

    Contextual Data Files:

  • Complete.data.file.dta: 2,757 Cases, 1,055 Variables
  • countyemployment.dta: 10,956 Cases, 6 Variables
  • places.dta: 5,004 Cases, 5 Variables
  • TractDataInterpolated-long.dta: 57,036 Cases, 50 Variables
  • TractDataInterpolated-wide.dta: 2,716 Cases, 1,009 Variables
  • tractscales2000.dta: 2,716 Cases, 49 Variables
  • urbanicity + density.dta: 2,716 Cases, 9 Variables
    •

    Demographic variables included: gender, race, educational attainment, age, employment, and marital status.

    Citation

    Morenoff, Jeffrey D., and Harding, David J. Michigan Study of Life After Prison, Administrative Data on 2003 Cohort of Michigan Parolees. Inter-university Consortium for Political and Social Research [distributor], 2019-03-28. https://doi.org/10.3886/ICPSR32681.v1

    Export Citation:

    • RIS (generic format for RefWorks, EndNote, etc.)
    • EndNote

    Funding

    United States Department of Justice. Office of Justice Programs. National Institute of Justice (2008-IJ-CX-0018), United States Department of Health and Human Services. National Institutes of Health. Eunice Kennedy Shriver National Institute of Child Health and Human Development (1R21HD06016001A1), Russell Sage Foundation, University of Michigan. National Poverty Center, University of Michigan. Center for Local, State, and Urban Policy, National Science Foundation (F027757)

    Subject Terms

    Geographic Coverage

    Smallest Geographic Unit

    County

    Restrictions

    Access to these data is restricted. Users interested in obtaining these data must complete a Restricted Data Use Agreement, specify the reason for the request, and obtain IRB approval or notice of exemption for their research.

    Distributor(s)

    Inter-university Consortium for Political and Social Research
    Hide

    Time Period(s)

    2003 (Data pertains to the 2003 population cohort of Michigan parolees released to Michigan communities.)

    Date of Collection

    2007 -- 2012 (IRB Approval Period)

    Data Collection Notes

    1. These data are part of NACJD's Fast Track Release and are distributed as they were received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except for the removal of direct identifiers. Users should refer to the accompanying readme file for a brief description of the files available with this collection and consult the investigator(s) if further information is needed.

    2. The variable rsid may be used to link the data files in this collection.
    Hide

    Study Purpose

    This project involved collecting data through a unique arrangement with the Michigan Department of Corrections (MDOC) and prospectively analyzing the role of neighborhood context in structuring recidivism and labor market outcomes of returning parolees.

    Study Design

    This study included the population of parolees released from Michigan prisons in 2003 and paroled to Michigan communities (n= 11,064). Some results were based on data from the entire population of those paroled in 2003. Other analyses were based on a random 1/6 sample from this population (n= 1,848) for which detailed residential information was collected. Data came from four sources: (1) Michigan Department of Corrections (MDOC) administrative databases that provided criminal history and demographic data as well as recidivism outcomes; (2) narrative electronic case notes and paper files written by MDOC parole and probation agents and coded by the research team; (3) arrests reported to the Michigan State Police; and (4) unemployment insurance records provided by the Michigan Unemployment Insurance Agency.

    Post-release residential histories were assembled from the case notes that included move dates, residence types, and addresses, which were then geocoded and linked to census tracts and their characteristics from the 2000 Census. Data were also collected on the pre-prison addresses from hard copies of pre-sentence investigation reports, in which addresses were usually verified by the MDOC agent preparing the report, as well as from parole violation reports and parole agent case notes.

    The study included five measures of recidivism for each parolee: (1) arrests for a new offense (as recorded in case notes by parole agents and as recorded by the Michigan State Police), (2) re-commitment to prison due to parole violation, (3) re-commitment to prison for a new conviction, (4) absconding, and (5) new felony convictions, whether or not they resulted in return to prison.

    Through a data sharing agreement between MDOC and the Michigan Unemployment Insurance Agency, this study also obtained unemployment insurance earnings records for the entire population, both during the post-incarceration period and prior to incarceration. UI records were based on employer reports, included employer information, and provided pre-incarceration data, but they were reported in three-month increments, excluding temporary or "under the table" employment, and required matching by social security number.

    Sample

    This study included three nested sets of subjects:

  • Population of parolees released from Michigan prisons in 2003 and paroled to Michigan communities (n= 11,064).
  • A 1/3 sample of the population, selected via a two-stage cluster sampling design (see below), with first post-prison neighborhoods as clusters (n= 3,689). Pre-prison residential information, residential histories, employment, and arrests coded from parole agent casenotes are available for the 1/3 sample.
  • A 1/6 sample of the population, selected at random from the 1/3 sample (n= 1,848). Residential histories, employment, and arrests coded from the parole agent casenotes are available for the 1/6 sample.
    •

    Two-Stage Cluster Sample Design for 1/3 Sample

    A two-stage clustered sampling design (in which parolees were clustered within census tracts) with probabilities proportional to size (PPS) was employed. In the first stage, census tracts were sampled with probability proportionate to their size (i.e., the number of parolees who returned to each tract). In the second stage, individuals within each selected tract were sampled with probability inversely proportionate to the tract selection rate. When the first- and second-stage selection rates were multiplied together, the sampling probability was equal for every individual. This approach ensured the final sample size of parolees remained the same no matter which tracts were sampled in the first stage.

    Time Method

    Cross-sectional

    Universe

    Population of all individuals paroled from Michigan prisons to Michigan communities in 2003 (n= 11,064).

    Unit(s) of Observation

    Counties and Census Tracts, Individual

    Data Source

    Michigan Department of Corrections. Offender Management Network Information (OMNI) system.

    Michigan State Police. Arrest Records.

    Michigan Department of Corrections. Corrections Management Information System (CMIS).

    Michigan Unemployment Insurance Agency.

    Data Type(s)

    Mode of Data Collection

    Description of Variables

    Cleaned Data Files:

  • case note arrests reps 1-4 10-21-10final.dta: Variables include casenote recorded date, arrested date, what individual was arrested for, by whom the individual was arrested, charges, and information on being held in custody.
  • case note arrests reps 5-8.dta: Variables include casenote recorded date, arrested date, what individual was arrested for, by whom the individual was arrested, charges, and information on being held in custody.
  • case note arrests rep 9.dta: Variables include casenote recorded date, arrested date, what individual was arrested for, by whom the individual was arrested, charges, and information on being held in custody.
  • case note employment reps 1-4 9-14-10final.dta: Variables include information on employment.
  • case note employment reps 5-8.dta: Variables include information on employment.
  • case note employment rep 9.dta: Variables include information on employment.
  • cleaned casenote addresses reps 1-8.dta: Variables include information on parole, release dates, casenote recorded date, discharge dates, and geocoded location and residency variables.
  • cleaned casenote addresses rep 9.dta: Variables include information on parole, release dates, casenote recorded date, discharge dates, and geocoded location and residency variables.
  • preprisonaddress_all.dta: Includes variables on geocoded location and residency pre-prison.
  • preprisonaddress_all_rep9.dta: Includes variables on geocoded location and residency pre-prison.
  • postprisads.dta: Includes variables on geocoded location and residency post-prison.
  • cleaned demographics for population.dta: Includes variables related to gender, race, education attainment, juvenile and adult crimes committed, Assault Risk and Property Risk variables, age, marital status, and other demographic variables.
  • simplecrimhistory.dta: Includes variables on offense group and types.
  • popSAhistory.dta: Includes variables on individuals' substance abuse histories.
  • death dates.dta: Includes date of death variable.
  • popprisonenterdates.dta: Includes variables related to prison entry dates.
  • discharge dates.dta: Includes first and second discharge dates following 2003 parole.
  • parole and release dates for pop.dta: Includes date of release and parole variables.
  • mdoc_recidivism_measures.dta: Includes Michigan Department of Corrections (MDOC) recidivism measure variables.
  • recidivism dates from transits.dta: Includes dates and reasons for individuals' return to prison (re-commitment).
  • recidivism from bir.dta: Includes variables noting felony convictions from the Bureau of Internal Revenue (BIR).
  • sample marker.dta: Includes sample identification variable.
  • samplereps.dta: Includes sample replicate number variable.
  • tta_rsid_rep.dta: Includes sample replicate number variable.
    •

    Contextual Data Files:

  • Complete.data.file.dta: Contains variables pertaining to census tract data.
  • countyemployment.dta: Contains variables pertaining to employment associated with FIPS codes.
  • places.dta: Contains variables that describe the place associated with FIPS codes.
  • TractDataInterpolated-long.dta: Contains variables by census tract that detail proportion of population that have varying characteristics, such as receiving public assistance, education attainment, race, and disadvantage/affluence factor and alpha scales.
  • TractDataInterpolated-wide.dta: Contains imputed variables by census tract and year that detail proportion of population that have varying characteristics, such as receiving public assistance, education attainment, race, and disadvantage/affluence factor and alpha scales.
  • tractscales2000.dta: Contains variables by census tract, from the 2000 Census, that detail proportion of population that have varying characteristics, such as receiving public assistance, education attainment, race, and disadvantage/affluence factor and alpha scales.
  • urbanicity + density.dta: Contains variables that determine the population density and whether it is categorized as rural, urban, central, or suburban.

    • Response Rates

    N/A

    Presence of Common Scales

    N/A

    Hide

    Original Release Date

    2019-03-28

    Hide

    Weight

    Data do not require weights when combined with cases from original sample who are ages 18-25.

    Hide

    Notes

    • These data are part of NACJD's Fast Track Release and are distributed as they were received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except for the removal of direct identifiers. Users should refer to the accompanying readme file for a brief description of the files available with this collection and consult the investigator(s) if further information is needed.

    • The public-use data files in this collection are available for access by the general public. Access does not require affiliation with an ICPSR member institution.

    • One or more files in this data collection have special restrictions. Restricted data files are not available for direct download from the website; click on the Restricted Data button to learn more.