Governor’s Commission on Innovation, Efficiency, and Transparency

October 25, 2013

The Pew-MacArthurThe Pew-MacArthurResults First Initiative:Results First Initiative:

Investing in Programs that WorkInvesting in Programs that Work

2

The Policy Challenge

• While we talk about making strategic choices, the budget process relies on inertia and anecdote

• Very limited data on:– What programs are funded

– What each costs

– What programs accomplish

– How they compare

3

The Solution: Bring Evidence Into the Process

• Target funds using rigorous evidence

• Stop funding ineffective programs

• Ensure programs are implemented effectively

• Achieve dramatic improvements without increased spending

4

The cost-benefit analysis approach

•Long-standing approach to policy analysis

•Widely used in the private sector

•Increasingly used by states

Goal: Assess whether a program generates enough benefits to justify funding

5

States are increasing use of CBA

Report found number of studies done by states is growing significantly each year

States are increasingly mandating that studies be done

States are using results in policy and budget processes

6

Predict program impacts in your state

Calculate and compare long-term costs and benefits

Use the best national research to identify what works

7

Step 1: Conduct a Program Inventory

• Identify the programs currently provided in New Hampshire and the population that is served by those programs

• Identify the current funding for programs

• Assess whether the programs are evidence-based

• Determine if programs are being implemented according to design

8

Step 1: Inventory Programs

*Washington State 2012 dollars

POLICY/PROGRAMPOLICY/PROGRAM COSTCOST LONG-TERM LONG-TERM BENEFITSBENEFITS

COST/BENEFIT COST/BENEFIT RATIORATIO

Intensive supervision (only) $4,140 -$578 -$0.14

Mental health court $2,935 $20,424 $6.96

Community drug treatment $1,602 $17,711 $11.05

Correctional education in prison $1,128 $21,426 $19.00

Work release $661 $7,117 $10.77

Cognitive behavioral therapy $412 $9,695 $23.55

Community job training & aid $135 $5,501 $40.76

JUVENILE PROGRAMSJUVENILE PROGRAMS

Functional Family Therapy $3,262 $70,370 $21.57

Aggression replacement training $1,508 $62,947 $41.75

Coordination of services $395 $5,501 $13.94

Scared Straight $65 -$4,949 -$76.35

9

Step 2: Identify Program Costs

• Identify the costs of serving persons in each program

• Include direct and indirect costs

• Calculate marginal costs for each program

10

Step 2: Identify Costs

*Washington State 2012 dollars

POLICY/PROGRAMPOLICY/PROGRAM COSTCOST LONG-TERM LONG-TERM BENEFITSBENEFITS

COST/BENEFIT COST/BENEFIT RATIORATIO

Intensive supervision (only) $4,140 -$578 -$0.14

Mental health court $2,935 $20,424 $6.96

Community drug treatment $1,602 $17,711 $11.05

Correctional education in prison $1,128 $21,426 $19.00

Work release $661 $7,117 $10.77

Cognitive behavioral therapy $412 $9,695 $23.55

Community job training & aid $135 $5,501 $40.76

JUVENILE PROGRAMSJUVENILE PROGRAMS

Functional Family Therapy $3,262 $70,370 $21.57

Aggression replacement training $1,508 $62,947 $41.75

Coordination of services $395 $5,501 $13.94

Scared Straight $65 -$4,949 -$76.35

11

Step 3: Predict and Monetize Outcomes

– Costs per felony conviction

– Convictions avoided per participant

– Other benefits throughout system

– Victimizations avoided per participant

Taxpayer benefits Taxpayer benefits (avoided costs)(avoided costs)

Victimization benefits Victimization benefits (avoided costs)(avoided costs)

12

• Taxpayer outcomes

– Avoided cost of delivery of services and programs

• Societal outcomes

– Avoided costs incurred of crime victims

– Tangible costs (e.g., lost wages, health care)

– Intangible costs (e.g., pain and suffering)

– Estimates based on medical records, insurance claims, and court judgments

Assess full program outcomes

13

Meta-analysis of Functional Family Therapy

Follow-up Years Source: Washington State Institute for Public Policy

Recidivism Rate

Without FFT (actual baseline)

RECIDIVISM RATES REDUCED BY RECIDIVISM RATES REDUCED BY 22%22%

With FFT

14

Cost-Benefit of Functional Family Therapy

BENEFITS PER FAMILY BENEFITS PER FAMILY WA STATE WA STATE 2010 DOLLARS2010 DOLLARS MAIN SOURCE OF BENEFITSMAIN SOURCE OF BENEFITS

Reduced crime $31,745 Lower state & victim costs

Increased high school graduation $5,686 Increased earnings

Reduced health care costs $307 Lower public costs

Total Benefits Per Family $37,739

Cost Per Family $3,190

Net Present Value $34,549

Benefits Per Dollar of Cost $11.86

15

Cost-benefit of Nurse Family Partnership

BENEFITS PER FAMILY BENEFITS PER FAMILY WA STATE WA STATE 2010 DOLLARS2010 DOLLARS MAIN SOURCE OF BENEFITSMAIN SOURCE OF BENEFITS

Reduced crime & abuse $6,611 Lower state & victim costs

Educational gains $17,770 Increased earnings

Reduced public assistance costs $5,944 Lower public costs

Total Benefits Per Family $30,325

Cost Per Family $9,421

Net Present Value $20,904

Benefits Per Dollar of Cost $3.23

16

Step 4: Compare Costs and BenefitsAcross Program Portfolio

*Washington State 2012 dollars

POLICY/PROGRAMPOLICY/PROGRAM COSTCOST LONG-TERM LONG-TERM BENEFITSBENEFITS

COST/BENEFIT COST/BENEFIT RATIORATIO

Intensive supervision (only) $4,140 -$578 -$0.14

Mental health court $2,935 $20,424 $6.96

Community drug treatment $1,602 $17,711 $11.05

Correctional education in prison $1,128 $21,426 $19.00

Work release $661 $7,117 $10.77

Cognitive behavioral therapy $412 $9,695 $23.55

Community job training & aid $135 $5,501 $40.76

JUVENILE PROGRAMSJUVENILE PROGRAMS

Functional Family Therapy $3,262 $70,370 $21.57

Aggression replacement training $1,508 $62,947 $41.75

Coordination of services $395 $5,501 $13.94

Scared Straight $65 -$4,949 -$76.35

17

Results First Work in States

18

Washington State’s Long-term Success

• 15+ years of using approach to help steer budget decisions

• Have achieved better outcomes at lower costs

LOWERED CRIME RATE LOWERED CRIME RATE and achieved and achieved

$2.7 BILLION $2.7 BILLION in higher long-term benefitsin higher long-term benefits

19

Juvenile Crime Reduction Benefits

Change since 1990 in the U.S. and Washington State

1991 1993 1995 1971 19991 2001 2003 2005 2007 2009

Washington StateWashington State67% lower67% lower

Washington StateWashington State67% lower67% lower

United StatesUnited States49% lower49% lower

United StatesUnited States49% lower49% lower

In 2003, Washington begins “full fidelity” implementation

In 2000, Washington begins evidence-based Juvenile Justice

program

Source: Washington State Institute for Public Policy

20

Participation in Results First

AK

HIHI

WAWA

TNTN

MTMT

OROR

IDID

WYWY

COCOUTUT

NVNV

CACA

AZAZ NMNM

NENE

KSKS

OKOK

TXTX

MEME

NDND

SDSD

MNMN

LALA

ARAR

MOMO

IAIA

WIWI

MIMI

ILIL ININOHOH

PAPA

NYNY

WVWV

KYKY

MSMS ALAL GAGA

SCSC

NCNC

FLFL

VAVA

AKAK

21

What are Results First states doing?

enacted legislation incorporating Results First into their policy making process

used models to analyze legislation, avoiding millions in potential costs

22

New Mexico

• Legislative Finance Committee leads initiative, expanding into agencies

• Implemented Results First in all available policy areas

• Innovative “Cost of Doing Nothing” report found $360M in recidivism-related corrections costs over next 15 years

• Used Results First model to target $17M for evidence-based programming in early education and criminal justice

23

Iowa

• Housed in the Departments of Corrections and Human Rights

• Found state’s domestic violence treatment program was ineffective

– Replacing with new program to achieve higher ROI

• Used model to analyze sentencing reform proposals

24

New York

• Used model to develop Governor’s public safety budget

– Referenced in 2013 State of the State Address

• Restructuring $11.4M in Alternatives to Incarceration funds to prioritize cost-effective programs

– $5M allocated through competitive grant process incorporating cost-benefit analyses

25

Mississippi

• Legislative PEER Committee implementing model

• Very strong legislative leadership support

• Using approach to re-energize performance budgeting system

• Currently assessing criminal justice and education programs

26

Vermont

• Legislative Joint Fiscal Office began initiative

– Legislatively established the Criminal Justice Consensus Cost-Benefit Working Group to expand the Vermont Results First model

• Used analysis to cut funding to inefficient correctional education program

27

Results First can be used to analyze many policy areas

28

What does it take to become a Results First state?

29

State

SelectionCriteria

1

2

3

30

The Role of Partner States

• Secure leadership support

• Appoint a policy work group

• Establish a staff work group with project manager

• Collaborate with Results First to strengthen the model and build a learning community of states

31

Services provided by Results First

• Provide software

• Train staff in the approach

• Provide ongoing technical assistance

• Help interpret results for policymakers

• Compile and share lessons learned with other participating states

• Expand and update model

• No charge for Results First services

32

Goal – Dramatically improve outcomes by:

–Fund programs that are proven to work (and cut those that don’t)

–Programs must be properly implemented

–Must target the right people

–Compare outcomes to predictions

–Require new programs to prove success

Using EvidenceUsing Evidence

Ensuring Program QualityEnsuring Program Quality

Tracking ResultsTracking Results

33

This approach should drive the system

Appropriations(investment

advice)

Research(test new programs)

Implementation (ensure fidelity)

Oversight (monitor

outcomes)

www.pewstates.org/ResultsFirstwww.pewstates.org/ResultsFirst

Contact:

Gary VanLandinghamDirectorgvanlandingham@pewtrusts.org