It would appear fiscal restraint has finally caught up with police services across the country. The recent release of Police Resources in Canada, 2013 by Statistics Canada documents a decline in police strength after nearly a decade of increases as well as a slowdown in per capita spending. The crime rate, however continues to fall.
Figure 1 combines the data on staffing and criminal code incidents (exluding traffic) from Police Resources in Canada, 2013 with Statistics Canada’s population data and presents for Canada, police officers per 100,000 and crimes per 100,000 from 1986 to the present. Figure 2 presents for Canada, real per capita policing costs in 2002 dollars.
Police per capita fell from a peak in 1991 of 202.5 officers per 100,000 of population to 181.6 by 1998 and then began to increase again. During the era of fiscal restraint of the 1990s, officers per 100,000 fell by 10 percent. Interesting enough, during this same period, the crime rate also fell by about 22 percent. After 1998, police officers per 100,000 of population begin to rise and they are now at a level approximately where they were in the late 1980s before the recent decline set in. Indeed, since 2010, police officers per 100,000 have declined by 3 percent. Police costs per capita after rising steadily since about 2000 have also flattened out since 2010.
Yet, crime rates have continued an almost uninterrupted decline. Between 1991 and 2012, the crime rate as measured by criminal code incidents per 100,000 dropped by 46 percent. One might argue that they have decreased as a result of increased police expenditures and resources and increased spending since 1998 may indeed be a factor. However, the fact remains that they decline even during periods when police resources are flat or declining.
The United States also experienced a pretty steep drop in the crime rate starting in the 1990s. The dramatic decline in Canadian crime rates since 1991 is paralleled in the United States where homicide rates fell 43 percent between 1991 and 2001 and violent and property crime fell 34 and 29 percent respectively (Levitt 2004: 163). In his 2004 JEP paper, Steve Levitt noted that the media explanations for this drop included the implementation of new innovative policing strategies, changes in the market for crack cocaine, population aging, a stronger economy, tougher gun control laws and increases in the number of police (and I suppose by extension expenditure increases also).
However, according to Levitt (2004) the key factors were fourfold: the rising prison population which takes criminals off the streets, the waning crack epidemic, the legalization of abortion in the 1970s and increases in the number of police. Regarding abortions, Levitt explains that legalizing abortion resulted in a reduction in unwanted births and unwanted children are at a greater risk of crime. As for the aging of the population and the reduction of a youthful population generally more disposed to commit crimes, it was not an important factor. Innovative police strategies were not a factor either as the drop in crime appears to predate the onset of such strategies in the US. As for the impact of the economy, on crime, the impact is likely more indirect via the impact on government budgets and by extension on police and prison spending.
I don’t know enough to really conclude whether or not all of these factors also apply to Canada (or the United States for that matter) but I will comment on a few things. In the United States, Roe versus Wade set guidelines for the availability of abortions starting in 1973 while in Canada abortion was illegal until 1969. So the availability of abortions starting in the 1970s for both countries may support the Levitt explanation for declining crime rates in Canada staring in the 1990s. As for a rising prison population, the total number of Canadians in prison custody has grown but the number of prisoners per 100,000 appears to have stayed flat according to adult correctional statistics. The American incarceration rate has gone up. Don’t know enough about the evolution Canadian crack cocaine industry as an explanatory factor in crime though though the industry does appear to have caused some issues for Toronto’s current mayor. Anyway, all very interesting stuff.
References:
Levitt, S.D. (2004) “Understanding Why Crime Fell in the 1990s: Four Factors that Explain the Decline and Six that Do Not,” Journal of Economic Perspectives, Winter, 18, 1, 163-190.
Juristat article, Component of Statistics Canada catalogue no. 85-002-X, Juristat ISSN 1209-6393, Police resources in Canada, 2013, by Hope Hutchins, Canadian Centre for Justice Statistics, Released on March 27th, 2014.
Juristat article, Component of Statistics Canada catalogue no. 85-002-X, Juristat, ISSN 1209-6393, Adult correctional statistics in Canada, 2010/2011, by Mia Dauvergne, Canadian Centre for Justice Statistics, Released on October 11, 2012
Probably worth mentioning Foote & Goetz finding of errors in the original Levitt & Donohue paper (though those two have since replied). The impact of lead has also been much discussed as a possible factor.
Posted by: Wonks Anonymous | March 28, 2014 at 12:38 PM
Kevin Drum has been publicizing research that suggests that reducing the use of lead has caused the reduction in violent crime. His post http://www.motherjones.com/environment/2013/01/lead-crime-link-gasoline has a link to the original paper that claimed the elimination of lead in gasoline was the main factor.
Posted by: Vladimir | March 28, 2014 at 01:24 PM
Vladimir & Wonks Anonymous: Thank you for the lit references.
Posted by: Livio Di Matteo | March 29, 2014 at 09:25 AM
Life expectancy
Hi Livio,
remember your December 2013 thread (now closed) of OECD health data ?
And that a lot of the observed variation could be explained by “eastern Europe” and health care coverage in US, Mexico,. …
The problem is, the factors are brutally high.
After some discussion over this weekend, originally related to other things, like what goes nowadays as a bachelor thesis here and there, ….
The intuition is, that at least half of these five year difference for eastern europe and 15 for health care coverage, should be an accounting fake/misinterpretation of the mortality tables used, AND, that
Life expectancy dependence on income, SES, whatever you name it, should be significantly different between Canada and the US,
And a wonderful real world important topic for your group, to explore in more detail, with proper access to, and understanding of Canadian and US data
Posted by: genauer | March 30, 2014 at 03:37 PM
Didn't Canada prohibit leaded gasoline in the early 1990s? Lead has been used to explain crime rates in the US, so this might be worth considering.
Posted by: Matthew | March 30, 2014 at 06:23 PM
You can't equate Canada's position on abortion with Roe v. Wade. The 1969 reforms instituted a "Therapeutic Abortion Committee" at hospitals which provided abotions (which were few and far between). Abortion was technically legal but still very hard to get in Canada. It was the various iterations of R v. Mortentaler in Quebec and Ontario in the 1970's and 1980's which really made abortion available in practice.
The "Therapeutic Abortion Committee" system, which was a large barrier to access, was not struck downuntil 1988.
Posted by: Determinant | March 31, 2014 at 07:34 PM
Lead is a neurotoxin that affects behavior. Childhood lead exposure from gasoline and paint correlates well with crime. The Mother Jones stuff by Kevin Drum is great.
This is a worldwide phenomena. 20 years after the lead goes away the crime goes down.
Posted by: Chris J | March 31, 2014 at 11:05 PM
at first I thought, yeah, yeah, yeah, just some coincidence of factors in the US.
But then looking at the Nevin 2007 paper via the mother jones link,
with the data for several european countries .......
still leaves the possibility open, that since the lead is mainly coming from car traffic
"National trends in average blood lead and the use of lead in gasoline were highly
correlated, with median R2 of 0.94"), which represents the rising wealth in countries, just shifting by a few years from country to country, representing social changes (especially increased mobility resulting in less local social control) and not chemcial intoxication.
and none of the international data show any indication of a peak.
It could be there, but there are no data.
AYnd that makes this correlation a mere co-incidence. Sorry
These data are NOT proof of the thesis.
Posted by: genauer | April 02, 2014 at 01:41 PM