Figure 1. Robbery rate vs. Homicide rate

advertisement
Mobilizing Private Inputs for
Crime Prevention
Philip J. Cook
Duke University
John MacDonald
University of Pennsylvania
March 31, 2010
Royal Economic Society Annual Conference
University of Surrey
Crime control without prison
The love affair with imprisonment is over (?)
Alternatives for crime control:
*Smarter use of the criminal justice system
OR
*Social policy –
mental health, early childhood, schooling,
employment, housing, alcohol control
Can private action influence crime
rates?
Conceptually:
1. situational crime prevention
Private actions to make crime targets less
alluring to criminals
Displacement vs. “taking profit out of crime”
2. Private inputs into police investigation
Co-production with criminal justice system of
deterrence and incapacitation
Police officers and security guards per capita
450
400
350
y = 3.3199x + 273.06
300
250
200
y = 2.6211x + 211.16
150
100
50
0
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08
Police Officers per 100,000 persons all age
Security Guards per 100,000 persons all age
Linear (Police Officers per 100,000 persons all age)
Linear (Security Guards per 100,000 persons all age)
Case study of private action:
Business Improvement Districts
Self-taxing non-profits that focus on
neighborhood safety, sanitation, marketing
“Safe and clean”
Local public goods supplement city services
800 BIDs in US
Crime-related efforts:
Private security, improved contact with police
Evaluation results for BIDs in LA
Question: Do BIDs reduce crime?
History: California changes law, LA adopts a
process for BID creation
Evaluation of 30 new BIDs in LA, 1997-2003
Annual panel data, 1995-2005, for 1,072
“reporting districts” (neighborhoods)
179 become part of a BID
Crime Count per neighborhood
Annual Effect of BID on Crimes/RD
BID
BID
(YES/NO) /$10,000
Crime #
-28.0
-3.37
Robbery
-4.7
-.59
Assault
-2.4
-.43
Burglary
-3.2
-.53
Value of crime reduction
Contingent valuation study from national survey
vote for continuation of a program that reduces
[robbery] by 10% if it will cost you $X ?
X ranges from $25 to $225, at random.
Responses form a demand curve.
Area under the demand = total value
Compute implied value per crime
Social value of crime reduction per
$10,000 expended
Crimes
Value
@soc cost /$10K exp
Robbery
.59
@$263K
Assault
.43
@$79K
Burglary
.53
@$21K
$155K
$34K
$11K
Are we ignoring some costs?
Displacement to other neighborhoods? NO
panel regression results show nil effect
Do BIDs increase the arrest rate?
NO, just the opposite:
-1.7 arrests per $10,000 @ $2,947 = $4,863
savings to the city
Other concerns?
Implications
BIDs in Los Angeles reduce crime, & pass the
cost-benefit test
Are there equity concerns? Or (possibly) equity
benefits?
How can the collective action problem be
overcome in other states?
Is BID security equivalent to more police?
What can we conclude about private action?
Auto Theft as a case study of
private action
2008 vs. 1989: 2X vehicles but fewer thefts!
Why?
Self-Protection
1. Normal precaution
2. “Target hardening”
steering column locks; immobilizers
The Club
Too much or too little effort? (Moral hazard?)
Private co-Production with CJS
CJS effectiveness depends on willingness of
victims and others to provide a public service,
largely uncompensated
Cooperation begins with reporting to the police
Reporting rates, residential burglary
0.8
0.7
0.6
y = 0.0018x + 0.4762
0.5
0.4
0.3
y = 0.0076x + 0.2658
y =0.0092x + 0.1345
0.2
0.1
0
73
75
77
79
81
83
85
87
89
91
93
95
97
NCVS Reporting Rates
UCR Offences/NCVS Victimizations
99
01
03
05
07
Reporting rates, motor vehicle theft
1.4
1.2
y =0.0267x + 0.3412
1
y = 0.0076x + 0.7158
0.8
y = 0.0053x + 0.6543
0.6
0.4
0.2
0
73
75
77
79
81
83
85
87
89
91
93
95
97
99
01
03
05
07
NCVS Reporting Rates
UCR Offences/NCVS Victimizations
Linear (NCVS Reporting Rates)
Linear (UCR Offences/NCVS Victimizations)
Private incentives
1. Insurance requirement
Crime Stoppers
Victim compensation
Restitution/ tort settlement
Regulatory requirement
Satisfaction from doing good/ revenge
2. Too much or too little? Too little!
best example of victim input
Example: LoJack: Private cost $700. Externality
= $1300 per year (Ayres & Levitt).
Some states give insurance discount
VIN is earlier example
Mandated
total property crime rate - NCVS
600
500
400
300
200
100
0
73
75
77
79
81
83
85
87
89
91
93
95
97
99
01
Total property crime rate per 1,000 households
03
05
07
Conclusions
Why the huge drop in crime since 1980?
Private action deserves some credit
Should private action be encouraged by public
policy?
Yes for cooperation; yes for some precautions
How?
BIDs; insurance regulation; incentives for
cooperation; trust in system
Download