Skip to main content

Investigating the Assignment of Probation Conditions: Heterogeneity and the Role of Race and Ethnicity



Explore the typical “packages” of probation conditions and assess general and interactive racial and ethnic disparities in assignment to these “packages”.


Latent class analysis (LCA) is applied as an exploratory modeling technique to uncover the nature of heterogeneity in probation conditions using novel and detailed data for over 2100 adults convicted of a felony in two urban counties. Class enumeration procedures use the Bayesian Information Criterion and correct for potential violations of conditional independence assumption. Model verification is assessed using mean posterior probabilities, confidence intervals for class differentiability, split sample checks and multiple iterations with random starting points. The influence of race and ethnicity, along with other legal and extralegal factors on assignment of probation condition “packages” identified is estimated using multinomial regression.


Meaningful and discrete classes that ranged in their likely combinations, type, and number of conditions are identified by LCA with high assignment accuracy. Results from regression models suggest race and ethnicity play a role in assignment of probation conditions. Young black offenders and black drug offenders are particularly more likely to receive a wider range of and more restrictive conditions.


LCA has the potential of classifying the multiple components of probation sentences without masking the heterogeneous nature of the conditions and imposing classification systems. Heterogeneity in the assignment of probation conditions represents a source of racial and ethnic sentencing disparities.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2


  1. 1.

    The Minnesota Sentencing Commission states that incarceration sentences should be neutral with respect to the race, gender, and social or economic status (e.g. employment, educational attainment, living situation, and marital status) of convicted felons. All common offenses are given a ranking between 1 and 11 of severity level. The criminal history calculation includes all felony offenses from the last 15 years, ranging from 0.5 to 2 points depending on their severity, all gross misdemeanors from the last 10 years (typically 0.25 points), and felony juvenile adjudications (maximum of 1 point unless the presumptive sentence was prison). Offenders receive an additional point if they are under custody at the time of the current offense. Each cell in the grid provides the presumptive sentence length in months, and the range within which a judge may sentence without the sentence being considered a departure. The cells in the bottom left corner of the grid, including severity levels I and II for criminal history scores 0–5, severity levels III and IV for criminal history scores 0–3, and severity levels V- VII for criminal history scores 0–2, are given a presumptive stayed sentence.

  2. 2.

    While a jail sentence may be imposed, it is not always executed and may be replaced with a community sanction.

  3. 3.

    The description of Minn. Stat. Ann. §609.135 applies both to current practices and to 2009 when the sample in the present study was collected. In 2009 Subd. 8 was repealed which stated that a defendant’s obligation to pay court-ordered fines and fees will survive for six years from the date of expiration of the stayed sentence. Since then, the only update to the statute has been an introduction of a pilot project in 2014 which changed the standards for ordering offenders charged with domestic abuse to use an electronic monitoring device and indicated that violations of location restrictions in situations where the victim and the defendant are both mobile does not automatically constitute a violation of conditions. For 2009 felony offenders, electronic monitoring was explicitly specified as a condition of probation in only a handful of cases, and was more commonly given as an alternative for those eligible to avoid part or all of their jail terms.

  4. 4.

    When the statutory maximum is less than four years, judges may exceed the statutory maximum and give probation sentences of up to four years.

  5. 5.

    Only 16 offenders in the sample (0.007%) received unsupervised probation.

  6. 6.

    Offenses without a severity level ranking, usually because they are rare or new, are not included in analysis. 107 total criminal sexual conduct cases (55 sentenced to probation) were coded but not included in the sample. Criminal sexual conduct is treated differently by the guidelines, as these cases have their own separate sentencing grid, severity classification system, and different eligibility for terms of probation. Felony DWI cases were also excluded since they have certain sentencing requirements and severity level 7 is reserved solely for this offense.

  7. 7.

    Minnesota’s sentencing guidelines are intimately tied to its correctional population (Robina Institute of Criminal Law and Criminal Justice 2017). Minnesota has one of the lowest prison population rates, but ranks one of the highest in probation (Kaeble and Bonczar 2017).

  8. 8.

    There may have been clerical errors that cause omissions of conditions of probation in the court documents or changes that were not updated in the system. However, it seems unlikely that these types of errors would result in systematic differences that would bias the results in a significant way.

  9. 9.

    Information is not available on Minnesota Trial Court Public Access website or otherwise on the specific court clerk writing each court document.

  10. 10.

    Hispanic offenders were not evaluated for the interaction effect of being young and Hispanic or a Hispanic drug offender given power issues. Only 39 offenders in the sample were Hispanic and between the ages of 18–21, and 31 were Hispanic drug offenders.

  11. 11.

    Levels 10 and 11 are for murder offenses which are not included in the sample.

  12. 12.

    The number of years sentenced to probation is missing for 17 observations (less than 0.008% of the sample) and were omitted from related analyses. No other variables are missing in the data.

  13. 13.

    The two percent without explicit basic probation conditions are likely a result of clerical error in the court documents. It seems unlikely that these types of errors would result in systematic differences that would bias the results in a significant way.

  14. 14.

    Drug-involved offenders account for a much larger percentage of the criminal justice population than those convicted of a drug crime (Belenko et al. 2013; Office of National Drug Control Policy 2014). The presentence report is likely to provide information to judges on the offender’s involvement with drugs and alcohol, regardless of their convicted crime.

  15. 15.

    The entropy scores are a summary measure of how good classification is, ranging from 0 to 1. The closer the values are to 1, the better the classification certainty. More indicators are generally associated with higher classification certainty. Therefore a lower entropy score associated with a higher number of indicators suggests problems with the model, as is the case when the drug and alcohol conditions are used as two separate indicators.

  16. 16.

    Although they had a relatively low correlation (0.166), the model was also estimated by combining the employment and school related indicators, as these conditions may potentially tap into the same construct and not necessarily provide additional information as two separate indicators. However, the BIC still clearly identified a five group model as the best fit and indicating that, unlike the drug and alcohol related conditions, maintaining employment and participating in employment counseling tap into separate types of conditions and they do not generate extraneous classes. Given this and their relatively low correlation they were left in the model as separate indicators, unlikely to violate the conditional independence assumption.

  17. 17.

    Jail confinement is separate from the probation “package” itself since it constitutes a separate element of the overall sentence. The classes are representative of the combinations of types of probation conditions assigned, but the overall sentence may include other elements which may change its harshness. Jail sentences may be imposed, but not necessarily executed and the lengths of confinement may not reflect actual time served.

  18. 18.

    The classes did not substantially differ in their likelihood of having a prison sentence length pronounced at sentencing, with similar distributions as the full sample (0.588). Limited-Low Financial: 0.588, Limited-Restorative: 0.560, Multiple-Treatment: 0.667, Multiple-Restrictive: 0.586, Multiple-Mixed: 0.582.

  19. 19.

    Unlike the overall number of conditions which may not operate in an additive manner to imply harshness, additional types of conditions are more likely to intrude on multiple domains of an offender’s life.

  20. 20.

    Appendix 2 provides descriptive statistics for all the classes.

  21. 21.

    The table is not completed for visual simplicity. All coefficients that are not shown in the text are available upon request.

  22. 22.

    Drug offenders were most likely to fall into the Limited-Low Financial “package” and the Multiple- Mixed “package”. Supplemental analyses (available upon request) suggest that the severity of the offense and the county sentenced explain assignment into one of these “packages” over the other.

  23. 23.

    In Minnesota, when probationers violate any condition of probation the probation revocation process may be initiated up to six months after the end of the probation period (Mitchell and Reitz 2014). Probation terms may also be extended up to two years for failing to pay restitution at least 60 days before the term of probation expires, or for those deemed likely not to pay by that time (Minn. Stat. Ann. §609.135). It can be extended up to three years if the probationer has not completed any of the court-ordered treatments at least 60 days before the term of probation expires or is likely not to complete treatment by that time (Minn. Stat. Ann. §609.135). Of offenders sentenced to probation in 2009 in Minnesota, 16% had their probation revoked within five years (Minnesota Sentencing Guidelines Commission, 2016). Ramsey County has a relatively high revocation rate especially for technical violations of probation, demonstrating the potential impact that the assigned packages of probation conditions from which violations may happen can have (Ruhland and Alper 2016).


  1. Akaike H (1973) Information theory and an extension of the maximum likelihood principle. International symposium on information theory.

    Google Scholar 

  2. Albonetti CA (1991) An integration of theories to explain judicial discretion. Soc Probl.

    Article  Google Scholar 

  3. Applegate BK (2014) Of race, prison, and perception: seeking to account for racially divergent views on the relative severity of sanctions. Am J Crim Justice.

    Article  Google Scholar 

  4. Bales WD, Piquero AR (2012) Racial/ethnic differentials in sentencing to incarceration. Justice Q.

    Article  Google Scholar 

  5. Baumer EP (2013) Reassessing and redirecting research on race and sentencing. Justice Q.

    Article  Google Scholar 

  6. Belenko S, Hiller M, Hamilton L (2013) Treating substance use disorders in the criminal justice system. Curr Psychiatry Rep.

    Article  Google Scholar 

  7. Beto DR, Corbett RPJ, DiIulio JJJ (2000) Getting serious about probation and the crime problem. Correct Manag Q 4(2):1–8

    Google Scholar 

  8. Bridges GS, Steen S (1998) Racial disparities in official assessments of juvenile offenders: attributional stereotypes mediating mechanisms. Am Sociol Rev.

    Article  Google Scholar 

  9. Bushway SD, Forst B (2013) Studying discretion in the processes that generate criminal justice sanctions. Justice Q.

    Article  Google Scholar 

  10. Bushway SD, Piehl AM (2007) Social science research adn the legal threat to presumptive sentencing guidelines. Criminol Public Policy.

    Article  Google Scholar 

  11. Chiricos T (1996) Moral panic as idealogy: drugs, violence, race and punishment in America. In: Lynch MJ, Patterson EB (eds) Justice with prejudice: race and criminal justice in America. Harrow and Heston, Albany, pp 19–49

    Google Scholar 

  12. Corbett RP (2015) The Burdens of leniency: the changing face of probation. Minn Law Rev 99:1697–1732

    Google Scholar 

  13. Crouch BM (1993) Is incarceration really worse? Analysis of offenders’ preferences for prison over probation. Justice Q.

    Article  Google Scholar 

  14. D’Alessio SJ, Stolzenberg L (1993) Socioeconomic status and the sentencing of the traditional offender. J Crim Justice.

    Article  Google Scholar 

  15. D’Alessio SJ, Stolzenberg L (1995) The impact of sentencing guidelines on jail incarceration in Minnesota. Criminology.

    Article  Google Scholar 

  16. Daly K, Tonry M (1997) Gender, race, and sentencing. Crime Justice.

    Article  Google Scholar 

  17. DeMichele M (2014) Studying the community corrections field: applying neo-institutional theories to a hidden element of mass social control. Theor Criminol.

    Article  Google Scholar 

  18. Demuth S (2003) Racial and ethnic differences in pretrial release decisions and outcomes: a comparison of Hispanic, black, and white felony arrestees. Criminology.

    Article  Google Scholar 

  19. Doerner JK, Demuth S (2010) The independent and joint effects of race/ethnicity, gender, and age on sentencing outcomes in U.S. federal courts. Justice Q 27(1):1–27.

    Article  Google Scholar 

  20. Eisenstein J, Jacob H (1977) Felony justice: an organizational analysis of criminal courts. Little, Brown and Co, Boston

    Google Scholar 

  21. Engen RL, Gainey RR, Crutchfield RD, Weis JG (2003) Discretion and disparity under sentencing guidelines: the role of departures and structured sentencing alternatives. Criminology.

    Article  Google Scholar 

  22. Frase RS (2005a) Sentencing guidelines in Minnesota, 1978–2003. Tonry, Michael.

    Article  Google Scholar 

  23. Frase RS (2005b) State sentencing guidelines: diversity, consensus, and unresolved policy issues. In: Columbia law review

  24. Frase RS (2009) What explains persistent racial disproportionality in Minnesota’s prison and jail populations? Crime Justice.

    Article  Google Scholar 

  25. Gainey RR, Steen S, Engen RL (2005) Exercising options: an assessment of the use of alternative sanctions for drug offenders. Justice Q 22(4):488–520.

    Article  Google Scholar 

  26. Gray MK, Fields M, Maxwell SR (2001) Examining probation violations: who, what, and when. Crime Delinq.

    Article  Google Scholar 

  27. Hagenaars JA, Mccutcheon AL (2002) Applied latent class analysis. Cambridge University Press, Cambridge

    Book  Google Scholar 

  28. Jannetta J, Breaux J, Ho H, Porter J (2014) Examining racial and ethnic disparities in probation revocation: summary findings and implications from a multisite study

  29. Johnson BD (2015) Examining the “life course” of criminal cases. Criminol Public Policy 14(2):183–186.

    Article  Google Scholar 

  30. Johnson BD, Dipietro SM (2012) The power of diversion: intermediate sanctions and sentencing disparity under presumptive guidelines. Criminology.

    Article  Google Scholar 

  31. Kaeble D, Bonczar TP (2017) Probation and parole in the United States, 2015. Washington, DC

  32. Klingele C (2013) Rethinking the use of community supervision. J Crim Law Criminol 103(4)

  33. Kutateladze BL, Andiloro NR, Johnson BD, Spohn CC (2014) Cumulative disadvantage: examining racial and ethnic disparity in prosecution and sentencing. Criminology.

    Article  Google Scholar 

  34. Lanza ST, Dziak JJ, Huang L, Wagner AT, Collins LM (2015) LCA Stata plugin users’ guide (Version 1.2). University Park, PA

  35. Lazarsfeld PF, Henry NW (1968) Latent structure analysis. Houghton, Mifflin, New York

    Google Scholar 

  36. May DC, Wood PB, Mooney JL, Minor KI (2005) Predicting offender-generated exchange rates: implications for a theory of sentence severity. Crime Delinq.

    Article  Google Scholar 

  37. McCutcheon AL (1987) Latent class analysis. Sage, Thousand Oaks

    Book  Google Scholar 

  38. Miethe TD, Moore CA (1989) Sentencing guidelines: their effect in Minnesota. NIJ Research in Brief

  39. Minnesota Sentencing Guidelines Commission (2009) Minnesota Sentencing Guidelines and Commentary. St. Paul, Minnesota. Retrieved from

  40. Minnesota Sentencing Guidelines Commission (2016) Probation revocations: Offenders sentenced from 2001-2013 and revoked to prison. St. Paul, MN. Retrieved from Revocations Report.pdf

  41. Minnesota Statutes (2009) Stay of Imposition or Execution of Sentence, Pub. L. No. 609.135 (2009). The Revisor of Statutes. Retrieved from

  42. Mitchell O (2005) A meta-analysis of race and sentencing research: explaining the inconsistencies. J Quant Criminol.

    Article  Google Scholar 

  43. Mitchell KL, Reitz KR (2014) Profiles in probation revocation: examining the legal framework in 21 states. Robina Institute of Criminal Law & Criminal Justice

  44. Nagin DS (2005) Group-based modeling of development. Harvard University Press, Harvard.

    Book  Google Scholar 

  45. National Research Council (1983) Research on sentencing: the search for reform. National Academies Press, Washington, DC.

    Book  Google Scholar 

  46. Nylund KL, Asparouhov T, Muthén BO (2007) Deciding on the number of classes in latent class analysis and growth mixture modeling: a Monte Carlo simulation study. Struct Equ Model.

    Article  Google Scholar 

  47. Office of National Drug Control Policy (2014) 2013 Annual Report, Arrestee Drug Abuse Monitoring Program II. Washington, DC. Retrieved from

  48. Ostrom CW, Ostrom BJ, Kleiman M (2004) Judges and discrimination: assessing the theory and practice of criminal sentencing. Final Grant Report to the National Institute of Justice (Grant 98-CE-VX-0008)

  49. Petersilia J (1990) When probation becomes more dreaded than prison. Federal Probat.

    Article  Google Scholar 

  50. Petersilia J (1997) Probation in the United States. Crime Justice.

    Article  Google Scholar 

  51. Petersilia J (2011) Community corrections: probation, parole, and prisoner reentry—Book, Section—Stanford Law School. In: Wilson JQ, Petersilia J (eds) Crime and public policy. Oxford University Press, New York

    Google Scholar 

  52. Petersilia J, Deschenes EP (1994) Perceptions of punishment: inmates and staff rank the severity of prison versus intermediate sanctions. Prison J.

    Article  Google Scholar 

  53. Raftery AE (1995) Bayesian model selection in social research. Sociol Methodol.

    Article  Google Scholar 

  54. Raykov T (2017) Latent class analysis in social science research. In: Inter-University Consortium for Political and Social Research (ICPSR). University of California at Berkeley, Berkeley, CA. Retrieved from

  55. Rios VM (2011) Punished: policing the lives of Black and Latino boys. New York University Press. Retrieved from

  56. Robina Institute of Criminal Law and Criminal Justice (2014) Profiles in probation revocation: Examining the legal framework in 21 states. Retrieved from

  57. Robina Institute of Criminal Law and Criminal Justice (2017) Jurisdiction profile: Minnesota. Retrieved from

  58. Ruhland E, Alper M (2016) Probation revocation and its causes: Profiles of state and local jurisdictions. Ramsey County, Minnesota

    Google Scholar 

  59. Ruhland EL, Robey JP (2016) Probation revocation and its causes: Profiles of state and local jurisdictions, Hennepin County, Minnesota. Retrieved from

  60. Schwarz G (1978) Estimating the dimension of a model. Ann Stat.

    Article  Google Scholar 

  61. Spohn CC (2000) Thirty years of sentencing reform: the quest for a racially neutral sentencing. Crime Justice 3:427–501

    Google Scholar 

  62. Spohn CC (2009) How do judges decide? The search for fairness and justice in punishment. Sage, Thousand Oaks.

    Book  Google Scholar 

  63. Starr SB (2015) Estimating gender disparities in federal criminal cases. Am Law Econ Rev.

    Article  Google Scholar 

  64. Steffensmeier D, Kramer J, Ulmer J (1995) Age differences in sentencing. Justice Q.

    Article  Google Scholar 

  65. Steffensmeier D, Ulmer J, Kramer J (1998) The interaction of race, gender, and age in criminal sentencing: the punishment cost of being young, black, and male. Criminology.

    Article  Google Scholar 

  66. Steffensmeier D, Painter-Davis N, Ulmer J (2017) Intersectionality of race, ethnicity, gender, and age on criminal punishment. Sociol Perspect.

    Article  Google Scholar 

  67. Steinmetz KF, Henderson H (2016) Inequality on probation: an examination of differential probation outcomes. J Ethn Crim Justice.

    Article  Google Scholar 

  68. Tapia M, Harris PM (2006) Race and revocation: is there a penalty for young, minority males? J Ethn Crim Justice.

    Article  Google Scholar 

  69. Tonry MH (1995) Malign neglect: race, crime, and punishment in America. Oxford University Press, Oxford

    Google Scholar 

  70. Ulmer J (2012) Recent developments and new directions in sentencing research. Justice Q.

    Article  Google Scholar 

  71. Ulmer JT, Bradley MS (2006) Variation in trial penalties among serious violent offenses. Criminology.

    Article  Google Scholar 

  72. Ulmer J, Painter-Davis N, Tinik L (2016) Disproportional imprisonment of black and Hispanic males: sentencing discretion, processing outcomes, and policy structures. Justice Q.

    Article  Google Scholar 

  73. Vera Institute of Justice (2013) The potential of community corrections to improve safety and reduce incarceration. New York, NY

  74. Vermunt J, Magidson J (2002) Latent class cluster analysis. Appl Latent Class Anal.

    Article  Google Scholar 

  75. Walker S, Spohn C, DeLone M (2012) The color of justice: race, ethnicity, and crime in America. Contemp Issues Crime Justice Ser.

    Article  Google Scholar 

  76. Watts AL (2016) Probation in-depth: the length of probation sentences. Minneapolis

  77. Wicklund C (2004) Assessment, treatment, and compliance. Recommendations to the community corrections field: From the national community corrections forum. Alexandia, VA

  78. Wood PB, Grasmick HG (1999) Toward the development of punishment equivalencies: male and female inmates rate the severity of alternative sanctions compared to prison. Justice Q 16(1):19–50

    Article  Google Scholar 

  79. Wood PB, May DC (2003) Racial differences in perceptions of the severity of sanctions: a comparison of prison with alternatives. Justice Q.

    Article  Google Scholar 

  80. Yan S (2017) Search for the hidden punishments: an alternative approach to studying alternative sanctions. J Quant Criminol.

    Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Anat Kimchi.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


Appendix 1

See Table 8.

Table 8 Frequency of conviction offenses

Appendix 2

See Table 9.

Table 9 Descriptive statistics by class

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kimchi, A. Investigating the Assignment of Probation Conditions: Heterogeneity and the Role of Race and Ethnicity. J Quant Criminol 35, 715–745 (2019).

Download citation


  • Probation conditions
  • Latent class analysis
  • Sentencing packages
  • Disparities
  • Race/ethnicity