Measuring Criminal Spillovers: Evidence from Three Strikes

by on May 27, 2009 at 7:05 am in Economics, Law | Permalink

California’s Attorney General was pleased to announce that “An unintended but positive consequence of ‘Three-Strikes’ has been the impact on parolees leaving the state….The growth in the number of parolees leaving California is staggering.” Law enforcement officers in other states were presumably less pleased. A displaced criminal is a benefit to California but a cost to other states. If such criminal spillovers are important, law enforcement will over-invest in policies that encourage displacement. We test whether California’s three-strikes law led to significant criminal spillovers.

That's the abstract from my latest paper, Measuring Criminal Spillovers: Evidence from Three Strikes (with Eric Helland). In an earlier paper, Helland and found–by comparing statistical doppelgangers some subject to the law some not–that three strikes does deter (whether it deters eneough to be good public policy in less clear).  In this paper, however, we found that three strikes does not cause appreciable exit of criminals from California (or appreciable reduction in entry), i.e deterrence does not occur on the relocation margin.  During the time that California's AG spoke, everyone was leaving CA not just criminals.  One conclusion of our research is that a federalist approach to crime remains viable. 

1 Compressor May 27, 2009 at 7:35 am

Need for Quality Compressor Parts can not be overemphasized. Since its inception in 1980, SPARECO have specialized in Standard Quality Spares for all types of Compressors and have achieved customer satisfaction consistently all along.

2 Eric Melin May 27, 2009 at 8:22 am

hi,

Very nice to visit your blog and found very interesting. I also have an blog http://www.jazd.net, which is a free online B2B business directory for Sales Solution with Guaranteed Results. I Want that You will give me a link on your blog and I will do the same for you.

Thank you

Eric Melin

3 A.M. May 27, 2009 at 9:35 am

I read too many police procedurals, but… all the ones set in California claim that P(commit a 3rd crime | three strikes law) < P(commit a 3rd crime | no three strikes law) BUT that P(commit a murder | two strikes) > P(commit another theft or drug offense | two strikes)

Maybe testable with your dataset?

4 IWantCookieNow May 27, 2009 at 9:37 am

Isn’t “three strikes” one of these policies that inflates the US prison population so much?

5 Alex Golubev May 27, 2009 at 11:05 am

any way to measure the “nothing/much less to lose mindset” when the criminal is on the third strike? ie when i was in Oakland a few months ago, one of these guys killed 3 cops and left one brain dead. not sure if he died. not sure if there’s a way to measure resistance to arrest by type of warrant between two populations (3rd strike and not).

6 LemmusLemmus May 27, 2009 at 11:27 am

The second link does not work.

7 Alex Golubev May 27, 2009 at 1:00 pm

Found the answer to my own question. This is an old study and I expect that this has gotten worse. Even violent criminals have a learning curve. In LA, “there is a notable increase in arrest rates, resisting and assaulting officers, and a significant increase (113% between 1996 and 2001) in two and three-strikes crimes with a police officer victim.”

Officer Down: Implications of Three Strikes for Public Safety
This article examines the impact of California’s 1994 Three Strikes law on front line law enforcement personnel—specifically whether there is a greater risk of injury by criminal suspects resisting arrest and contact with law enforcement due to Three Strikes. The authors evaluated data collected from six major police agencies and district attorney offices in California between 1990 and 2001, including total arrests, resisting arrest charges, assault on a peace officer, officer injuries or deaths, use of force incidents, officer-involved shootings, vehicle pursuits, and three-strike case filings. Although the resulting analysis did not evidence a statewide increase or trend, in the Los Angeles area (where there is a higher concentration of repeat offenders and three-strikes prosecution has been more actively pursued), there is a notable increase in arrest rates, resisting and assaulting officers, and a significant increase (113% between 1996 and 2001) in two and three-strikes crimes with a police officer victim.
http://cjp.sagepub.com/cgi/content/abstract/16/4/443

ALTHOUGH, this study is more comprehensive (even older) and the ends seem to justify the means :/ :
Moody, Carlisle E., Thomas B. Marvell, Robert J. Kaminski. “Unintended Consequences: Three-Strikes Laws and the Murders of Police Officers”, National Institute of Justice, 01/11/2002.
“The dataset was a pooled time series and cross section of 50 States for the period of 1973 to 1998. The dependent variable was the number of law enforcement officers feloniously killed in the line of duty. The target variable was a dummy variable that took the unit value in years following the passage of a three-strikes law. The results show an estimated impact of 44 percent more murders in years following the laws. In the average state there were 1.2 police murders per year in the 1990’s; so the typical three-strikes law led to an additional police murder roughly every other year. This means that approximately 0.0006 percent of arrests for major violent crimes in three-strike States involve police murders that would not have occurred without the laws. Several other criminal justice policies that might affect police murders were evaluated by the pooled regression model. Laws requiring sentencing enhancements for crimes committed with firearms appear to reduce police killings by roughly 18 percent. The size of the prison population, the number of executions, and the presence of right-to-carry concealed weapons have no discernible impact.”

8 Albert May 27, 2009 at 7:14 pm

there were 1.2 police murders per year in the 1990’s; so the typical three-strikes law led to an additional police murder roughly every other year. This means that approximately 0.0006 percent of arrests for major violent crimes in three-strike States involve police murders that would not have occurred without the laws.

Cool site check it out folks Interesting Finance & Economic Articles free link submissions for economics, finance, and technical analysis articles from bloggers/websites

9 lez sohbet November 7, 2009 at 2:11 pm

eywl

10 mardinkusu January 26, 2010 at 5:04 pm

Comments on this entry are closed.

Previous post:

Next post: