Maryland Data Analysis Part 2: Community Corrections Drivers - - PowerPoint PPT Presentation

Maryland Data Analysis Part 2: Community Corrections Drivers

Justice Reinvestment Coordinating Council August 18, 2015

1

Justice Reinvestment Coordinating Council Focus

SB 602

The Council shall “develop a statewide framework of sentencing and corrections policies to further reduce the state’s incarcerated population, reduce spending

• n corrections, and reinvest in strategies to increase

public safety and reduce recidivism … ”

June 22, 2015

2

Outline

• Prison drivers review
• Follow-up questions
• Community corrections drivers
• Research on what works to reduce recidivism

3

PRISON DRIVERS

4

Pretrial Population

Prison Population Down 5% in Last Decade

22,466 21,326

• 5,000

10,000 15,000 20,000 25,000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Prison Population, Annual Snapshot

Prison Population

Note: 2005-2013 stock population snapshot count in August, 2014 snapshot count in July

5

Pretrial Population

If Not for Baltimore City, State Prison Population Would Have Grown in Last Decade

Prison Population

9,564 6,706 12,853 14,559

• 2,000

4,000 6,000 8,000 10,000 12,000 14,000 16,000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Prisoners by Jurisdiction of Origin, Annual Snapshots

Baltimore City All other jurisdictions

6

Pretrial Population

Baltimore City and County Still Largest Contributors to Prison Population

1,000 2,000 3,000 4,000 5,000 6,000 7,000 BALT CITY BALT COUNTY PR GEORGE'S MONTGOMERY ANNE ARUNDL WICOMICO HARFORD WASHINGTON CHARLES FREDERICK HOWARD CECIL CARROLL DORCHESTER WORCESTER ST MARY'S CALVERT CAROLINE SOMERSET ALLEGANY TALBOT QUEEN ANNES KENT GARRETT

Prisoners by Jurisdiction, July 2014 Snapshot

Prison Population

7

Pretrial Population

2/3 of Prisoners in for Person Crimes

Person, 65% Property, 13% Drugs, 19% Public order, 3%

Prisoners by Offense Type, July 2014

Prison Population

8

Pretrial Population

Almost 2/3 of Prisoners from New Sentences, 28% from Probation Revocations

Prison Population

Sentenced to prison, 63% Mandatory supervision return, 5% Parole return, 4% Probation revocation, 28%

Prisoners by Admit Type, August 2014

9

Pretrial Population

Prison Admissions Down 19% in Last Decade

11,078 8,928

• 2,000

4,000 6,000 8,000 10,000 12,000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Prison Admissions, by FY

Prison Admissions

10

Pretrial Population

58% of Admissions Are for Nonviolent Crimes

Person, 42% Property, 20% Drugs, 32% Public

• rder,

7%

Prison Admissions by Offense Type, FY14

Prison Admissions

11

Pretrial Population

58% of Admissions Were Previously on Supervision

Prison Admissions

Sentenced to prison, 42% Mandatory supervision return, 20% Parole return, 17% Probation revocation, 21%

Prison Admissions by Type, FY14

12

Pretrial Population

Decline in Newly Sentenced Prisoners Due Almost Entirely to a Drop in Drug Admissions

1,632 885 1,572 284 1,665 749 991 293

• 200

400 600 800 1,000 1,200 1,400 1,600 1,800 Person Property Drugs Public order

Newly Sentenced Prisoners by Offense Type, FY05 vs FY14

2005 2014

Prison Admissions

37% decline in

• ffenders sentenced to

prison for drug crimes

13

Pretrial Population

PWID Still #1 Crime at Admission, Distribution and Possession Also in Top 10

Top 10 Offenses at Admission in FY14, Newly Sentenced Prisoners Admitted to Prison Offense 2005 2014 % Change, 2005-2014 Possession w/ Intent to Distribute Narcotics 964 462

• 52%

Assault-2nd Degree 342 340

• 1%

Robbery with a Deadly Weapon 248 281 13% Narcotics Distribution 285 240

• 16%

Robbery 172 229 33% Theft Felony 204 221 8% Assault-1st Degree 245 214

• 13%

Burglary-1st Degree* 210 Possession of a CDS (Excluding Marijuana) 178 144

• 19%

Murder-1st Degree 66 132 100%

Prison Admissions

*Burglary-1st Degree did not exist in its current form in 2005

14

Pretrial Population

Admissions from Baltimore City Down 43%, All Others Up 4%

Prison Admissions

3,206 5,704

• 1,000

2,000 3,000 4,000 5,000 6,000 7,000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Prison Admissions by Jurisdiction, by FY

Baltimore City All other jurisdictions

15

Pretrial Population

25% Increase in Average Sentence Length for Newly Sentenced Prisoners

77.4 96.7 20 40 60 80 100 120 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Months

Average Sentence for Newly Sentenced Prisoners, by FY

Prison Admissions

16

Pretrial Population

Time Served Up 23% in Last Decade

Time Served in Prison

29 35.7 5 10 15 20 25 30 35 40 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Average Time Served, by FY (Months)

17

Pretrial Population

Time Served Up for All Offense Types

Time Served in Prison

61.6 27.9 30.1 18.5 75.3 31.4 33.3 24.7 10 20 30 40 50 60 70 80 Person Property Drugs Public order Months

Average Time Served for New Court Commitments by Offense Type, FY05 vs FY14

2005 2014 22% increase for person offenders 13% increase for property offenders 34% increase for public order offenders

18

Pretrial Population

Proportion of Parole Releases Increased but Still Less Than 40% of All Releases

Parole 30% Mandatory release 68% Other 2%

Prison Release Type, FY05

Parole 37% Mandatory release 59% Other 4%

Prison Release Type, FY14

Time Served in Prison

19

Pretrial Population

Violent Offenders Released Closer to Parole Eligibility Date Than Nonviolent Offenders

Offense

% of sentence served by new court commitments released to parole, FY14 Must serve 50% Robbery with a Deadly Weapon 56% Assault-1st Degree 55% Robbery 54% Burglary-1st Degree 51% Must serve 25% Possession w/ Intent to Distribute Narcotics 40% Assault-2nd Degree 38% Narcotics Distribution 43% Theft Felony 38% Possession of a CDS (Excluding Marijuana) 36% Possession of Regulated Gun 37% Time Served in Prison

20

FOLLOW-UP QUESTIONS

21

Pretrial Population

Baltimore City Leads State in Admissions per 100,000 Residents

Follow-Up Questions

33 48 66 72 79 81 101 125 135 140 142 149 149 153 160 200 207 242 267 274 313 348 365 515 100 200 300 400 500 600 MONTGOMERY HOWARD FREDERICK ANNE ARUNDL CARROLL PR GEORGE'S ST MARY'S QUEEN ANNES GARRETT CALVERT BALT COUNTY TALBOT ALLEGANY CECIL CHARLES HARFORD KENT WORCESTER WASHINGTON CAROLINE SOMERSET WICOMICO DORCHESTER BALT CITY

Admissions per 100,000 Residents by Jurisdiction, FY14

22

Pretrial Population

Time Served by Offense

FY14 New Court Commitments Released from Prison Offense Time Served (months) % Paroled % Sentence Served, Those Paroled Possession w/ Intent to Distribute Narcotics 36.6

57% 40%

Assault-2nd Degree 20.4

35% 38%

Narcotics Distribution 37.8

61% 43%

Robbery with a Deadly Weapon 71.5

36% 56%

Theft Felony 25.4

51% 38%

Assault-1st Degree 79.5

33% 55%

Robbery 45.5

31% 54%

Burglary-1st Degree 44.9

31% 51%

Possession of a CDS (Excluding Marijuana) 12.4

47% 36%

Possession of Regulated Gun 29.1

24% 37%

Murder-2nd Degree 158.2

35% 59%

Burglary-2nd Degree 51.2

29% 48%

Theft Misd $100 - <$1K 12.4

31% 36%

Rape-2nd Degree 110.4

12% 47%

Burglary-4th Degree 17.1

34% 36%

Conspiracy Possession CDS (Excluding Marijuana) 17.9

53% 36%

DWI/Alcohol 9.6

38% 33%

Possession of Handgun 16

24% 38%

Unauth Use Of Goods 13.9

29% 37%

Other CDS Charge (Including Marijuana) 22.9

55% 34%

Follow-Up Questions

23

Pretrial Population

Black Share of Prison Population Has Declined in Past Two Decades, Still Disproportionate

78% 71% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Black Share of Prison Population by FY

Follow-Up Questions Source: DPSCS, June Snapshots

24

Pretrial Population

Black Offenders More Likely to Be Sentenced to Prison for Drug

• r Person Crimes; White Offenders for Property Crimes

1,174 417 361 376 752 219 209 75 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% BLACK WHITE

Newly Sentenced Prisoners by Offense Type, by Race, FY14

Person Property Drugs Public order

Follow-Up Questions

25

Pretrial Population

Black Offenders Serve Longer in Prison than White Offenders

82.3 30.8 35.5 27.1 54.5 60.7 32.7 23.4 18.9 41.5 10 20 30 40 50 60 70 80 90 Person Property Drugs Public order Total Months

Average Time Served by Released New Court Commitments by Offense Type and Race, FY14

Black White

Follow-Up Questions

26

COMMUNITY CORRECTIONS DRIVERS

27

Outline

• Active population
• Discharges from supervision
• Time served on supervision

28

Data

• Maryland Department of Public Safety and Correctional

Services – Division of Parole and Probation data:

• OBSCIS Snapshots, August 2005-2012
• OCMS Snapshots, August 2013-2014

29

Supervision Types

Probation Probation Before Judgment (PBJ)

Supervision before the court imposes a verdict

Probation After Judgment (PAJ)

Supervision under which the court suspends a prison sentence and allows the offender to serve a term in the community

Post-Release Supervision Parole

Supervision while on a period of discretionary, conditional release from prison granted by the Maryland Parole Commission

Mandatory Release Supervision

Supervision while serving the remainder of an

• ffender’s sentence less diminution of confinement

credits after mandatory release from prison; only applies to offenders with sentences of 18 months or more

30

Supervision Levels

Violence Prevention Initiative (VPI)

Supervision level for cases assigned to VPI. Individuals under this supervision level will be assigned to one of two supervision levels: VPI 1 or VPI 2

High

Supervision level for offenders with a risk score of fifteen or above

Moderate

Supervision level for offenders with a risk score above 6 or below 15

Low-Moderate

Supervision level for offenders with a risk score of 6 or lower

Low

Least intensive supervision level for offenders. This type of supervision level has no contact reporting requirements

Sex Offender

Specialized caseload for offenders who have been convicted of a sex offense. Offenders under this supervision level are assigned to one of four supervision levels (LV1 through LV4)

31

Discharge Types

Unsatisfactory Discharge

Revocation: New Offense

The offender is guilty of a new offense committed while under supervision and the court or parole commission finds the offender guilty of a Violation of Probation or Parole (VOP) that includes the new charge as a basis of the VOP (regardless of whether or not the VOP results in incarceration)

Revocation: Technical Violation

Violations other than new convictions that result in the offender being found guilty of a VOP (regardless of whether or not the VOP results in incarceration)

Unsatisfactory: New offense

The offender is guilty of an offense that was committed during the supervision or monitoring period, and the case is closed (with or without a hearing) by the court or parole commission without finding the offender guilty of a VOP

Unsatisfactory: No New Offense

Violations other than new convictions have been documented in a report to the court or parole commission and the case is closed (with or without a hearing) without the offender being found guilty of a VOP

Satisfactory Discharge

Expiration of sentence

The case reaches the legal expiration date

Early termination

The court agrees to close the case in a satisfactory status prior to the legal expiration date

Commutation

The case that resulted in the offender being placed under supervision is commuted

Other Discharge

Transferred out of state

The offender is transferred to supervision in another jurisdiction

32

COMMUNITY CORRECTIONS ACTIVE POPULATION

33

Pretrial Population

Outline

• Population over time
• Demographics
• By supervision type
• By supervision level
• By geographic region or jurisdiction

DPP Active Population

34

Pretrial Population

5% Decrease in Community Corrections Population in Last Decade

49,734 47,467 10,000 20,000 30,000 40,000 50,000 60,000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Community Corrections Population by FY

DPP Active Population

35

Pretrial Population

Active Cases Per 100,000 Residents Dropped 29% in Baltimore City Over Last Decade, Steady in the Rest of the State

DPP Active Population

2,778 589 1,976 595 500 1,000 1,500 2,000 2,500 3,000 Baltimore City All Other Jurisdictions

Community Corrections Population Per 100,000 Residents, FY05 vs FY14

2005 2014

36

Pretrial Population

80% of Offenders on Community Supervision on Probation

Probation 80% Parole 11% Mandatory Supervision 8% Other 1%

Community Corrections Population by Supervision Type, FY14

DPP Active Population

37

Pretrial Population

Probation Population Has Larger Proportion of Females Than Post-Release Supervision Population

MALE, 92% FEMALE, 8%

Post-Release Supervision Population by Gender, FY14

DPP Active Population

MALE, 77% FEMALE, 23%

Probation Population by Gender, FY14

38

Pretrial Population

Blacks Are Overrepresented in Probation and Post- Release Supervision Populations

WHITE, 43.3% BLACK, 54.0% INDIAN, 0.1% ASIAN, 0.8% UNKNOWN, 1.8%

Probation Population by Race, FY14

WHITE, 27% BLACK, 72% INDIAN, 0% ASIAN, 0% UNKNOWN, 1%

Post-Release Supervision Population by Race, FY14

DPP Active Population

39

Pretrial Population

4% Decline in Probation Population Over Last Decade

DPP Active Population

39,844 38,206 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Probation Population by FY

40

Pretrial Population

Declines in Both Supervision Types

DPP Active Population

6,427 33,417 5,617 32,588 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 Probation Before Judgment Probation After Judgment

Probation Population by Supervision Type, FY05 vs FY14

2005 2014

41

Pretrial Population

Central, South Regions Supervise Over Three Quarters of Probation Population

South, 37% Central, 40% North, 24%

Probation Population by Region, FY14

DPP Active Population

42

Pretrial Population

71% of Probation Population on Low or Moderate Supervision

DPP Active Population

VPI, 5% High, 19% Moderate, 31% Low-Moderate, 26% Low, 14% Sex Offender, 6%

Probation Population by Supervision Level, FY14

43

Pretrial Population

PAJ Has Higher Percent of VPI, Sex Offender, High Risk Cases

DPP Active Population

2% 6% 21% 13% 33% 25% 29% 31% 11% 20% 3% 5% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Probation Before Judgment Probation After Judgment

Probation Population by Supervision Level by Supervision Type, FY14

VPI High Moderate Low-Moderate Low Sex Offender

44

Pretrial Population

8% Decline in Post-Release Supervision Population Over Last Decade

DPP Active Population

9,717 8,981 2,000 4,000 6,000 8,000 10,000 12,000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

Post-Release Supervision Population by FY

45

Pretrial Population

Decline Driven by 29% Drop in Mandatory Supervision Population; Parole Population Up 17%

DPP Active Population

4,528 5,189 5,306 3,675 1,000 2,000 3,000 4,000 5,000 6,000 Parole Mandatory Supervision

Post-Release Supervision Population by Supervision Type, FY05 vs FY14

2005 2014

46

Pretrial Population

Central Region Supervises Half of Post-Release Supervision Population

South, 33% Central, 50% North, 16%

Post-Release Supervision Population by Region, FY14

DPP Active Population

47

Pretrial Population

62% of Post-Release Supervision on Moderate or Low Supervision

DPP Active Population

VPI, 8% High, 21% Moderate, 28% Low-Moderate, 23% Low, 11% Sex Offender , 9%

Post-Release Supervision Population by Supervision Level, FY14

48

Pretrial Population

Key Takeaways

• 5% decline in community supervision population in last

decade

• Probationers make up 80% of community supervision

population

• 71% of probation population is moderate or low risk
• 62% of post-release supervision population is moderate or

low risk

DPP Active Population

49

COMMUNITY CORRECTIONS DISCHARGES

50

Pretrial Population

Outline

• Satisfactory vs unsatisfactory discharges

– Changes over time

• New criminal convictions
• By supervision type
• By supervision level
• By geographic region or jurisdiction

DPP Discharges

51

Pretrial Population

38% of Probationers Fail, Down in Last Decade

Satisfactory closing, 51% Unsatisfactory closing, 48% Other closing, 1%

Probation Discharges by Type, FY05

Satisfactory closing, 58% Unsatisfactory closing, 38% Other closing, 4%

Probation Discharges by Type, FY14

DPP Discharges

52

Pretrial Population

Large Increase in Successful Discharge for Baltimore City Probationers

DPP Discharges

43% 61% 55% 57% 57% 36% 45% 40% 0% 3% 0% 2% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2005 2014 2005 2014 Baltimore City All other jurisdictions

Probation Discharges by Type and Jurisdiction, FY05 vs FY14

Other closing Unsatisfactory closing Satisfactory closing

53

Pretrial Population

PBJ Has More Successful Discharges Than PAJ

DPP Discharges

71% 56% 27% 40% 2% 4% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Probation Before Judgment Probation After Judgment

Probation Discharges by Discharge Type and Supervision Type, FY14

Other closing Unsatisfactory closing Satisfactory closing

54

Pretrial Population

84% of Probationers Discharged Without a New Criminal Conviction While on Supervision

No new offense, 84% New offense, 16%

Probation Discharges by New Criminal Conviction Status, FY14

DPP Discharges

55

Pretrial Population

VPI Offenders More Likely to Fail Supervision

26% 33% 56% 75% 84% 58% 69% 63% 40% 23% 13% 35% 5% 4% 4% 3% 3% 7% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% VPI High Moderate Low-Moderate Low Sex Offender

Probation Discharges by Supervision Level and Discharge Type, FY14

Other closing Unsatisfactory closing Satisfactory closing

DPP Discharges

56

Pretrial Population

Low Risk Offenders Fail Supervision for Reasons Other Than a New Criminal Conviction

51% 46% 40% 34% 30% 45% 0% 10% 20% 30% 40% 50% 60% VPI High Moderate Low-Moderate Low Sex Offender

% of Unsuccessful Probation Discharges Convicted of a New Offense, by Supervision Level, FY14

DPP Discharges

57

Pretrial Population

Large Variation in Probation Failure Rate Across Jurisdictions

DPP Discharges

52% 52% 51% 49% 47% 46% 45% 44% 44% 44% 39% 39% 38% 37% 37% 36% 36% 36% 36% 35% 35% 33% 32% 31% 0% 10% 20% 30% 40% 50% 60% Allegany Cecil Caroline Anne Arundel Calvert

• St. Mary's

Wicomico Talbot Dorchester Frederick Harford Prince George's Carroll Garrett Queen Anne's Charles Baltimore County Montgomery Baltimore City Worcester Washington Kent Somerset Howard

Unsatisfactory Discharge Rate by Jurisdiction, FY14

58

Pretrial Population

Most Jurisdictions Have New Conviction Rate Around 15%, But Some Variation

DPP Discharges

23.7% 20.9% 18.9% 18.2% 15.8% 15.4% 15.4% 14.8% 14.5% 14.3% 14.2% 13.4% 13.1% 12.9% 12.9% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% Wicomico Allegany Anne Arundel Harford Baltimore City Baltimore County Carroll Frederick Cecil Montgomery Worcester Howard Washington Prince George's Charles

Percent of Probation Discharges with New Criminal Conviction by Jurisdiction, FY14

59

Pretrial Population

39% of Offenders on Parole or Mandatory Supervision Fail Supervision

Satisfactory closing, 57% Unsatisfactory closing, 42% Other closing, 1%

Post-Release Supervision Discharges by Type, FY05

Satisfactory closing, 57% Unsatisfactory closing, 39% Other closing, 4%

Post-Release Supervision Discharges by Type, FY14

DPP Discharges

60

Pretrial Population

83% of Parole or Mandatory Supervision Offenders Discharged Without a New Criminal Conviction While on Supervision

No new offense, 83% New offense, 17%

Post-Release Supervision Discharges by New Criminal Conviction Status, FY14

DPP Discharges

61

Pretrial Population

VPI Offenders Most Likely to Fail Supervision

30% 45% 69% 80% 89% 48% 67% 52% 28% 17% 9% 46% 4% 4% 4% 3% 2% 6% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% VPI High Moderate Low-Moderate Low Sex Offender

Post-Release Supervision Discharges by Supervision Level and Discharge Type, FY14

Other closing Unsatisfactory closing Satisfactory closing

DPP Discharges

62

Pretrial Population

VPI Offenders More Likely to Fail Post-Release Supervision Without a New Criminal Conviction

33% 50% 49% 55% 56% 25% 0% 10% 20% 30% 40% 50% 60% VPI High Moderate Low-Moderate Low Sex Offender

% of Unsuccessful Post-Release Discharges Convicted of a New Offense, by Supervision Level, FY14

DPP Discharges

63

Pretrial Population

Key Takeaways

• Just under 40% of community supervision cases fail supervision
• Probation success rates are up over the last decade, driven by

improvement in Baltimore City

• Less than 20% of probationers, parolees, and offenders on mandatory

release supervision are convicted of a new crime committed while on supervision

• Almost 60% of unsuccessful cases do not involve a new criminal

conviction – For probationers, low risk offenders more likely to fail without a new criminal conviction – For parolees and offenders on mandatory release supervision, VPI

• ffenders more likely to fail without a new criminal conviction

DPP Discharges

64

TIME SERVED ON COMMUNITY SUPERVISION

65

Pretrial Population

Outline

• By supervision type
• By supervision level
• By geographic region or jurisdiction

DPP Time Served

66

Pretrial Population

Offenders Serve 18 Months on PBJ, Two Years on PAJ

DPP Time Served

17.4 23.9 18.2 24.1 5 10 15 20 25 30 Probation Before Judgment Probation After Judgment Months

Average Time Served on Probation by Supervision Type, FY05 vs FY12

2005 2012

67

Pretrial Population

Very Little Difference in Time Served by Outcome

DPP Time Served

22 23.7 15.1 22 24.7 17 5 10 15 20 25 30 Satisfactory closing Unsatisfactory closing Other closing Months

Average Time Served on Probation, by Discharge Type, FY05 vs FY12

2005 2012

68

Pretrial Population

Low Risk Offenders Serve Almost as Long on Probation as High Risk Offenders

DPP Time Served

22.3 24.8 22.7 21.2 20.8 30.2 5 10 15 20 25 30 35 VPI High Moderate Low-Moderate Low Sex Offender Months

Average Time Served on Probation by Supervision Level, Before Satisfactory Close, FY12

69

Pretrial Population

Large Variation in Probation Time Served by Jurisdiction

DPP Time Served

39.5 29.6 29.9 29.8 27.6 26.1 25.8 24.3 23.4 23.2 22.7 22.4 22.5 21.5 21.8 20.7 20.8 19.7 19.1 19.5 18.7 18 16.7 16.2 5 10 15 20 25 30 35 40 45 Charles Calvert Kent Prince George's Carroll Baltimore City Worcester Cecil Wicomico

• St. Mary's

Talbot Dorchester Howard Frederick Anne Arundel Baltimore County Queen Anne's Harford Caroline Garrett Allegany Montgomery Somerset Washington Months

Average Time Served on Probation by Jurisdiction, FY12

70

Pretrial Population

PBJ Offenders Serve Less Time Across the Board

DPP Time Served

16.7 23.5 21.5 25.2 5 10 15 20 25 30 Probation Before Judgment Probation After Judgment Months

Average Time Served on Probation by Supervision Type and Discharge Type, FY12

Satisfactory closing Unsatisfactory closing

71

Pretrial Population

Parolees Serve Longer on Supervision; Both Types Up Since 2005

DPP Time Served

26.4 17.5 29.6 21.3 5 10 15 20 25 30 35 Parole Mandatory Supervision Months

Average Time Served on Post-Release Supervision by Supervision Type, FY05 vs FY12

2005 2012

72

Pretrial Population

Time Served on Supervision Up Across Both Discharge Types

DPP Time Served

21.8 19.2 26.9 21.5 5 10 15 20 25 30 Satisfactory closing Unsatisfactory closing Months

Average Time Served on Post-Release Supervision by Discharge Type, FY05 vs FY12

2005 2012

73

Pretrial Population

Low Risk Offenders Serve Longest

DPP Time Served

30.9 49.9 33.5 25.3 19.5 15 10 20 30 40 50 60 Sex Offender Low Low-Moderate Moderate High VPI Months

Average Time Served on Post-Release Supervision by Supervision Level, Before Satisfactory Closing, FY12

74

Pretrial Population

Key Takeaways

• Low risk probationers serve the same amount of time on supervision as

high risk probationers

• Time spent on probation varies widely by jurisdiction
• Offenders on parole and mandatory release supervision are serving

longer than they did a decade ago – Last month we saw prison sentences were growing

• Low risk offenders on parole and mandatory release supervision serve

an average of 49 months compared to 19 months for high risk offenders

DPP Time Served

75

What Works to Reduce Recidivism?

Justice Reinvestment Coordinating Council August 18, 2015

76

Outline

• Research on incarceration
• Research on reducing recidivism

77

RESEARCH ON INCARCERATION

78

Does more incarceration result in less crime?

Research on Incarceration

• Researchers have examined the question of whether increased

incarceration caused the crime decline in the 1990’s, and have found that it was responsible for 10-30% of the crime decline

• Difficult to isolate the impact, because of other simultaneous

variables

• Improved police strategies, technology, and personal security habits
• Demographic shifts
• Changes in drug markets

Source: National Research Council (2014), The Growth of Incarceration in the United States

79

Does more incarceration result in less crime?

Research on Incarceration

• Diminishing returns: The marginal impact of incarceration (the value

to society of sending one more person to prison) has declined since the 1990’s

• Agreement among researchers: Increasing incarceration today will

have little if any effect on crime

Source: National Research Council (2014), The Growth of Incarceration in the United States

80

Does more incarceration result in less crime?

Research on Incarceration

Steve Levitt (2004)

“Expenditures on prisons appear to have benefits that outweigh the direct costs of housing prisoners.”

Steve Levitt (2012)

“Today, my guess is that the costs [of incarceration]

• utweigh the benefits at the margins. I think we

should be shrinking the prison population by at least

• ne-third.”

Sources: Levitt (2004), Understanding Why Crime Fell in the 1990s; New York Times (Dec. 11, 2012), For Lesser Crimes, Rethinking Life Behind Bars, quoting Steve Levitt.

81

Does incarceration reduce recidivism?

Research on Incarceration

Researchers have examined whether incarceration reduces recidivism more than non-custodial sanctions

• Research models:
• Matched samples or experimental design: incarceration vs.

non-custodial sanctions

• Comparing recidivism outcomes
• Findings:
• No significant difference in recidivism rates or a criminogenic

effect of incarceration

Sources: Campbell Collaborative (2015); Nagin & Snodgrass (2013); Nagin, Cullen, and Lero Jonson (2009)

82

Does incarceration reduce recidivism?

Research on Incarceration

• Campbell Collaboration (2015) (meta-analysis):
• Found incarceration has a null or criminogenic effect on re-arrest

and re-conviction rates

• Nagin & Snodgrass (2013):
• Found no significant difference in 1, 2, 5, and 10-year re-arrest

rates

• Nagin, Cullen & Jonson (2009) (systematic review):
• Found incarceration has a null or criminogenic effect compared to

non-custodial sanctions

83

Does incarceration reduce recidivism?

Research on Incarceration

• Spohn and Holleran (2002)
• Found that drug offenders sentenced to prison were 5-6 times more likely

than probationers to be rearrested and charged, controlling for offender characteristics

• Drake and Aos (2012)
• Found that technical violators of probation serving a period of confinement

(jail or prison) had significantly higher recidivism than offenders sanctioned in the community

• Nieuwbeerta, Nagin, and Blokland (2009)
• Found first-time imprisoned offenders who served less than 1 year were

1.9 times as likely to be reconvicted within 3 years, compared to offenders sentenced in the community

• Property crimes: 2 times as likely
• Other nonviolent crimes: 1.8 times as likely

84

Does incarceration reduce recidivism?

Research on Incarceration

Researchers have also examined whether longer periods of incarceration reduce recidivism more than shorter periods

• Research models:
• Matched samples: shorter periods vs. longer periods
• Compared: recidivism outcomes
• Findings:
• No increased benefit of longer periods of incarceration

Sources: Nagin (2009); Anwar & Stephens (2011); Meade, et al. (2012)

85

Does incarceration reduce recidivism?

Research on Incarceration

• Nagin, Cullen & Jonson (2009) (systematic review):
• Found no relationship between time served and recidivism
• The United States Sentencing Commission (2014):
• No difference in recidivism for drug offenders with reduced

sentences after retroactive application of a new sentencing law

• Meade, et al. (2012):
• For prison terms of 5 years or less: no effect on recidivism
• For prison terms of 10 years or longer: some reduction in re-

arrest due to aging out

86

Does incarceration reduce recidivism?

Research on Incarceration

“[L]engthy prison sentences are ineffective as a crime control measure… [and] an inefficient approach to preventing crime by incapacitation unless they are specifically targeted at very high-rate or extremely dangerous offenders.”

National Research Council

The Growth of Incarceration in the United States (2014)

Source: National Research Council (2014), The Growth of Incarceration in the United States

87

Summary

Research on Incarceration

• Prison expansion historically:
• Played a small but significant part in the U.S. crime decline
• Prison expansion today:
• Has little, if any, additional crime reduction effect (diminishing

returns)

• Reducing recidivism:
• Incarceration is not more effective than non-custodial sanctions
• Longer prison terms do not guarantee better outcomes

88

RECIDIVISM REDUCTION PRINCIPLES

89

Recidivism Reduction

• Focus on high risk offenders, target criminogenic

needs, address programming barriers (Risk, Need, Responsivity)

• Use sanctions and incentives to respond to behavior
• Frontload resources for offenders coming out of

prison

• Incorporate treatment into supervision
• Monitor quality, fidelity, and outcomes

Key Principles

90

The risk principle tells us who to target

The Risk Principle

• High risk offenders are more likely to recidivate
• Require the most intensive intervention (supervision and

treatment)

• Low risk offenders are not as likely to recidivate
• Too much intervention can increase likelihood of recidivism

Source: Bonta & Andrews (2007), Risk-Need-Responsivity Model for Offender Assessment and Rehabilitation

91

Risk of future offending ≠ seriousness of the current

• ffense

The Risk Principle

• Someone who committed a serious crime could be more likely to

reoffend (high-risk) or less likely to reoffend (low-risk)

• Same for someone who committed a low-level crime

Source: Andrews (1999), Recidivism Is Predictable and Can Be Influenced: Using Risk Assessments to Reduce Recidivism

92

The Risk Principle

Source: Andrews, Bonta & Wormith (2004), Level of Service / Case Management Inventory (LS/CMI): An Offender Assessment System (user’s manual)

93

4%

• 19%
• 25%
• 20%
• 15%
• 10%
• 5%

0% 5% 10% Low Risk High Risk Percent Change in Recidivism Rate

The Risk Principle

Source: Dowden & Andrews (1999) (meta-analysis)

Correctional Interventions Targeting Low-Risk and High-Risk Offenders (Meta-Analysis)

94

2%

• 6%
• 10%
• 12%
• 10%
• 8%
• 6%
• 4%
• 2%

0% 2% 4% Low Risk Moderate Risk High Risk Percent Change in Recidivism Rate Risk Level

Recidivism Outcomes for Ohio Halfway House and Community Residential Placements

The Risk Principle

Source: Latessa et al. (2010), Follow-up Evaluation of Ohio’s Community Based Correctional Facilities and Halfway House Programs

95

15% 32% 51% 32% 0% 10% 20% 30% 40% 50% 60% Minimal Intensive Recidivism Rate Dosage of Programming Low Risk High Risk

The Risk Principle

Source: Bonta et al. (2000), A Quasi-Experimental Evaluation of an Intensive Rehabilitation Supervision Program

Programming Intensity and Dosage

96

The needs principle tells us WHAT we should be paying attention to

The Needs Principle

• Certain factors (criminogenic needs) are tied to criminal behavior
• Criminogenic = crime-producing
• Criminogenic needs = risk factors which predict recidivism

AND are dynamic (can be targeted for change)

• Static = cannot be changed (e.g., age and criminal history)
• Targeting criminogenic needs has been shown to reduce

recidivism

Source: Bonta & Andrews (2007), Risk-Need-Responsivity Model for Offender Assessment and Rehabilitation

97

Criminogenic Needs

The Needs Principle

• “Big Four” Criminogenic Risk Factors:
• Antisocial attitudes (dynamic)
• Antisocial peers (dynamic)
• Antisocial personality (dynamic)
• History of antisocial behavior (static)
• Other Criminogenic Risk Factors:
• Substance abuse
• Employment/education
• Low family affection/poor supervision/poor communication
• Leisure/recreation

Source: Bonta & Andrews (2007), Risk-Need-Responsivity Model for Offender Assessment and Rehabilitation

98

The Needs Principle

Risk Factors of a Heart Attack: 1) Increased LDL/HDL ratios (i.e., elevated LDL and low HDL levels) 2) Smoking 3) Diabetes 4) Hypertension 5) Abdominal obesity 6) Psychosocial (i.e., stress or depression) 7) Failure to eat fruits and vegetables daily 8) Failure to exercise 9) Failure to drink any alcohol

99

The Needs Principle

Source: Gendreau, French & Taylor (2002), What Works (What Doesn’t Work)

+1%

• 32%
• 35%
• 30%
• 25%
• 20%
• 15%
• 10%
• 5%

0% 5% Non-Criminogenic Criminogenic Change in Recidivism Rate Needs Targeted

Effect of Criminogenic vs. Non-Criminogenic Programming on Recidivism

100

The responsivity principle tells us HOW to target

• ffender issues

The Responsivity Principle

• Responsivity factors impact the likelihood of an individual being

successful in a program, intervention, or service

• Targeting responsivity factors will increase the offender’s

likelihood of success

• Examples:
• Acute mental illness
• Child care needs
• Transportation needs

Source: Bonta & Andrews (2007), Risk-Need-Responsivity Model for Offender Assessment and Rehabilitation

101

SANCTIONS AND INCENTIVES

102

Sanctions

Swift, Certain, and Proportional Sanctions

• Respond to negative behavior in a manner that will change that

behavior

• Deterrence:
• Swift, certain, and proportional sanctions have a stronger

deterrent effect than delayed, random, and severe sanctions

Source: Nagin & Pogarsky (2001), Integrating Celerity, Impulsivity, and Extralegal Sanction Threats into a Model of General Deterrence: Theory and Evidence

103

Swift, Certain, and Proportional Sanctions

Sanctions

Source: Hawken and Kleiman (2009), Managing Drug Involved Probationers with Swift and Certain Sanctions: Evaluating Hawaii’s HOPE

47% 46% 23% 15% 21% 13% 9% 7% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Arrested Used Drugs Skipped Appointments Probation Revoked Percent of Offenders

Hawaii’s HOPE Program Outcomes

Control Hope

104

Sanctions

Swift, Certain, and Proportional Sanctions

• Less effective deterrent
• Letting multiple violations build up before a response
• Imposing sanctions after a delay
• Imposing sanctions that are out of proportion to the problem

behavior

• Strong deterrent
• Making consequences clear upfront
• Responding swiftly to problem behavior
• Responding with sanctions that are proportionate to the

problem behavior

Source: Nagin & Pogarsky (2001), Integrating Celerity, Impulsivity, and Extralegal Sanction Threats into a Model of General Deterrence: Theory and Evidence

105

Sanctions

Swift, Certain, and Proportional Sanctions

• Harrell & Roman (2001) examined whether using swift, certain, and

proportional sanctions as part of a drug court program reduced recidivism

• Research model
• Matched samples: Participants in drug court program with

swift, certain, and proportional sanctions vs. participants in drug court programs without

• Compared: Re-arrest rates after 2 years
• Finding
• Substantially lower re-arrest rates (19% vs. 27% for the

control group)

Source: Harrell & Roman (2001), Reducing Drug Use and Crime Among Offenders: The Impact of Graduated Sanctions

106

Rewards and Incentives

Incorporate Rewards and Incentives

• Identify opportunities for rewarding prosocial behavior and attitudes

(e.g., case plan progress, practicing a new skill, taking initiative, being honest, etc.)

• Develop a continuum of rewards to round out the continuum of

sanctions

• Offender change is most effective when rewards are utilized at a

higher rate than sanctions

Source: Wodahl, Garland, Culhane & McCarty (2011), Utilizing Behavioral Interventions to Improve Supervision Outcomes in Community-based Corrections

107

Rewards and Incentives

Incorporate Rewards and Incentives

• Allowing probationers and parolees to step-down their supervision

(e.g., reduced reporting, less frequent drug testing, etc.) or earn their way off supervision for compliance with conditions

• Encourages offenders to change their behavior and attitudes,

thereby reducing violations

• Allocates resources based on which offenders are exhibiting

antisocial behaviors

Source: Petersilia (2007), Employ Behavioral Contracting for ‘Earned Discharge’ Parole

108

Rewards and Incentives

Incorporate Rewards and Incentives

Source: Wodahl, et al. (2007), Utilizing behavioral intervention to improve supervision.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 1:10 1:08 1:06 1:04 1:02 2:01 4:01 6:01 8:01 10:01 Probability of ISP Success Ratio of Rewards to Punishments

Ratio of Rewards to Sanctions and the Probability of ISP Success

109

FRONTLOAD RESOURCES

110

Frontload Resources

Frontload Resources

• Focus community supervision resources in the first few months

when offenders are most likely to violate conditions or commit a new crime

• Identify offenders who need enhanced supervision and those who

do not

• Adjust reporting requirements / conditions for successful
• ffenders to offset costs of frontloading
• Deter future crime and technical violations by changing offender

behavior early in the supervision process

Source: National Research Council (2007), Parole, Desistance from Crime, and Community Integration

111

Frontload Resources

Frontload Resources

50 100 150 200 250 300 0-10 0-90 90-180 180-270 270-360 360-450 Number of Violations Days Since Release

Failure within Selected Time Periods (per 1,000 parolees)

Any Violation Technical Violation Criminal Violation

Source: Grattet, Petersilia & Lin (2008), Parole Violations and Revocations in California

112

INCORPORATE TREATMENT INTO SUPERVISION

113

Treatment and Supervision

Incorporate Treatment into Supervision

Incorporate treatment into supervision case plans rather than using surveillance alone Cost-Benefit Outcomes for Adult Criminal Justice Programs

Source: Washington State Institute for Public Policy (2012), available at: http://www.wsipp.wa.gov/BenefitCost?topicId=2

114

MONITOR QUALITY, FIDELITY, AND OUTCOMES

115

Quality Assurance and Fidelity

Monitor Quality, Fidelity, and Outcomes

• Higher quality evidence-based practices have bigger impacts on

recidivism

• Validate risk and needs assessment tools
• Train, supervise, and coach staff
• Manage caseloads
• Monitor programs for compliance and fidelity
• Collect data, set performance benchmarks, and monitor
• utcomes

Source: Andrews & Bonta (2006), The Psychology of Criminal Conduct

116

Quality Assurance and Fidelity

Monitor Quality, Fidelity, and Outcomes

Programs designed to meet offenders’ criminogenic needs must be delivered with fidelity to the program model

• Functional Family Therapy
• Followed model: 38% decrease in recidivism
• Didn’t follow model: 17% increase in recidivism
• Aggression Replacement Therapy
• Followed model: 24% decrease in recidivism
• Didn’t follow model: 7% increase in recidivism

Source: Washington State Institute of Public Policy (2010)

117

Quality Assurance and Fidelity

Monitor Quality, Fidelity, and Outcomes

Source: Latessa et al. (2010), Follow-up Evaluation of Ohio’s Community Based Correctional Facilities and Halfway House Programs

• 7%
• 6%
• 5%
• 4%
• 3%
• 2%
• 1%

0% Internal QA No Internal QA

Effect of Internal Quality Assurance on Recidivism Outcomes

118

Reducing Recidivism

• Focus on high risk offenders, target criminogenic

needs, and address programming barriers (Risk, Need, Responsivity)

• Use sanctions and incentives to respond to behavior
• Frontload resources for offenders coming out of

prison

• Incorporate treatment into supervision
• Monitor quality, fidelity, and outcomes

Key Principles Summary

119

Next Meeting: September 11, 2:30 pm

• System review
• Introduction to policy development

Summary

120

Contact Information

• Connie Utada

– Office: 202.540.6423 – Email: cutada@pewtrusts.org

• Felicity Rose

– Office: 971.344.5556 – Email: frose@crj.org

• Len Engel

– Office: 617.482.2520 x129 – Email: lengel@crj.org

• Public Safety Performance Project

– www.pewtrusts.org/publicsafety