HANDLE WITH CARE: costbenefit studies and crime prevention

1
HANDLE WITH CAREcost-benefit studies and crime
prevention

• Daniel SANSFAÇON, Ph.D
• Deputy Director General
• International Centre for the Prevention of Crime

2
Summary

• What we (may know) reviewing some benefit-cost
  studies of crime prevention
• And may not know
• And what to do with it

3
About ICPC

• Created in 1994, it is an international NGO
• It is a unique international platform involving
  governments, cities, civil society and
  researchers
• To foster debate and exchange on crime prevention
  and community safety

4
About ICPC

• ICPCs 2 Crime Prevention Digests (1997, 1999)
  used existing cost-effectiveness and cost-benefit
  studies
• ICPCs on-going work on evaluating crime
  prevention, especially comprehensive, integrated,
  city-based approaches
• www.crime-prevention-intl.org

5
I - WHAT WE MAY KNOW (1)

• Cohens 5 point scale for cost-benefit studies
  from level 1 where only relevant program costs
  are fully assessed in monetary terms to level 5
  where all relevant program effectiveness measures
  are stated in monetary terms
• Shermans 5 point scale of scientific rigor for
  evaluation studies of program impacts from
  studies with correlational evidence only to
  randomized experiments where units are assigned
  at random to program and control groups

6
I - WHAT WE MAY KNOW (2)

• Situational crime prevention (Welsh and
  Farrington, 1999a)
• 13 studies reviewed
• 9 focus on residences and 4 on commercial
  premises or public facilities (transport
  systems)
• most interventions (9) commonly involved some
  form of surveillance, natural, formal (private
  security) or by citizens (block watch), others
  focused on target hardening or removal
• five studies have a quasi-experimental design,
  most use only before-after comparisons
• overall, interventions produce positive
  benefit-cost ratios in 8 cases, varying between
  1.31 and 5.04 five had turned undesirable ratios
  (0.22 to 0.68).

7
I - WHAT WE MAY KNOW (3)

• Developmental crime prevention (Welsh and
  Farrington, 1999b)
• 6 studies reviewed
• subjects ranging in age from prebirth to 18
  years
• variety of risk factors (parenting, education,
  cognitive development, behaviour problems)
• interventions from ten weeks to four years, and
  some had long follow-up periods
• five of the six studies yielded a desirable
  benefit-cost ratio, ranging from a low of 1.06 to
  a high of 7.16
• other benefits include improved educational
  achievement, reduced reliance on social welfare,
  increased wages from employment, etc.

8
I - WHAT WE MAY KNOW (4)

• Correctional crime prevention (Welsh and
  Farrington, 2000)
• 7 studies, all carried in the United States
• wide range of subjects and offending behaviours
  and used many different intervention approaches
• five used an experimental evaluation design, but
  none score more than 3 on the evaluation scale
• six of the seven studies reported positive
  effects on recidivism rates (measured by first
  rearrest or reconviction)
• four studies also examined and monetized other
  benefits such as education, employment, social
  service use and substance abuse, and found
  positive benefits larger than reduced recidivism
• benefit-cost ratios of these programs varied
  between 1.13 and 7.14. The measurement of costs
  included only direct out-of-pocket expenses and
  did not include intangible costs to victims.

9
II- AND MAY NOT KNOW

• (1) estimating the costs of crime is no easy
  task
• requires a sound basis re, extent of crime
  problem
• include costing crime control policies and their
  effects?
• (2) measuring the effectiveness in other words
  the benefits of preventive intervention remains
  challenging. For example, identifying and
  agreeing on the programs objective
• crime prevention is to maximise private security
  and social conformity
• crime reduction is designed to strengthen the
  institutions of the state with a view to uphold
  the rule of law
• community safety is the overriding collective
  good from which citizens derive their own private
  security and conformity. (Hope, in press)
• (3) good quality evaluations suffer from at least
  three major drawbacks
• their apparent lack of interest for the processes
  underlying the measured effects
• inability to disentangle the effects
• focus on crime reduction as a measure of impact

10
III- AND WHAT TO DO WITH IT

• Impacts on policy making ?
• Public Policy Institute of the State of
  Washington (Aos et al. 1999 and 2001). In 2003,
  the legislature directed the Institute to
  undertake a cost-benefit analysis of prevention
  and early intervention for at-risk youths.
• RANDs policy analysis scorecard (Karoly, 2001
  xv) impact on three-strikes legislation?
• Drug policy area as a prime example
• Re-integrating some degree of complexity to crime
  and delinquency.
• how crime is defined why focus only on street
  crime?
• what is excluded are incivilities part of
  crime?
• who is excluded are the lost lives of young
  delinquents discounted?
• whose costs are taken into consideration?

11
Conclusion

• The economy may be everywhere but so are the
  political and the social.
• In thinking about economic costs of crime and
  crime control studies, one needs to think
  simultaneously about
• What is to be counted and how ?
• What is to be done with the information and how
  to translate it for policy makers.

12
Conclusion (2)

• What sould be counted and how
• Benefit-cost analysis is an art that is built
  on many important assumptions. It is important
  to understand some of these assumptions before
  attempting either to conduct such a study or to
  interpret a study that has been done by others.
  (Cohen, 2000 266)
• It is not always clear that we make this effort
  to make our assumptions clear (e.g., drugs)

13
Conclusion (3)

• What to do with this information
• The findings suggest that tangible costs
  incurred by the justice system are not matched by
  tangible cost-saving benefits (of getting
  tougher). () This does not necessarily imply
  that getting tough on juveniles is wrong. For
  one thing, cost-benefit analysis does not have a
  special capacity to make such a judgement. Also,
  certain intangible benefits that we ignore,
  retribution for its own sake and vote enhancement
  for example, may compensate for the shortfall.
  (Fass and Pi , 2002 366).