Examination of Social Science Data Research

¶ … Crime and the Depenalization of Cannabis Possession: Evidence from a Policing Experiment" -- published in the Journal of Political Economy, a bimonthly peer-reviewed journal issued by the University of Chicago -- is a fascinating quantitative analysis on the impact of a localized policing "experiment" conducted over the course of seven and a half years. Between April 1998 and January 2006, the south London borough of Lambeth instituted a policy of "depenalization" for cannabis possession and consumption, the LCWS or Lambeth Cannabis Warning Scheme (Adda McConnell Rasul 2014, 1133). Although marijuana still remained illegal, the police in Lambeth made it known that "(i) the possession of small quantities of cannabis for personal consumption was still a recordable offence, but would no longer lead to the individual being arrested; (ii) the primary motivation was to free up police time and other resources to focus on crimes related to other drugs or other non-drug related crimes; (iii) the policy did not alter penalties for cannabis supply" (1135). Adda McConnell and Rasul undertook to examine the overall statistics related to crime that were recorded by the Lambeth police over the course of this seven and a half year period, to determine if this "policing experiment" had any noteworthy effect on crime.

Methods

Adda McConnell and Rasul wanted to examine the police statistics from two different angles. For a start, they wished to find out if the policy that had been stated publicly by Lambeth's police was actually effective. The police had made it clear that they intended to focus their efforts on policing drug crimes related to other more addictive or dangerous drugs instead of cannabis, and on policing non-drug-related crimes. For the first portion of this, Adda McConnell and Rasul were required to assemble all relevant quantitative data to enable them to make an analysis, as to whether Lambeth police had better results by focusing their policing on other drugs. They describe their data collection, measures and instruments, and techniques thus:

We use administrative records obtained from the London Metropolitan Police Service (MPS) to construct a panel data set on crime for all 32 London boroughs for each month from April 1998 until January 2006. This contains information on the number of recorded offences at two fine levels of detail: (i) the number of criminal offences related to any given drug type, e.g. cannabis, heroin, cocaine, etc.; (ii) for each drug type, the specific offence committed: possession, trafficking, intent to supply etc. Such detailed measurement of drug crime allows us to assess the impact of the policy (1136).

Using the full administrative records for the MPS across London as a whole, Adda McConnell and Rasul additionally compile statistics for seven non-drug-related crimes: assault, sexual offences, robbery, burglary, theft, fraud, criminal damage (1137). The goal is to compare police results in Lambeth during the experiment with those in the other 31 London police precincts on all of these variables.

Results

The results presented by Adda McConnell and Rasul are fascinating. Lambeth's public announcement that it will depenalize cannabis (while still recording it as a crime) leads to a "significant increase in cannabis related crime -- 29.3% more in Lambeth relative to the rest of London" during the experiment and afterwards too. In the period after the experiment Lambeth remains 61% higher. They find minimal evidence that the depenalization experiment allows for better policing of hard drugs, but find that it does have a significant and measurable effect on the policing of "non-drug crime" with "significant reudctions in give out of seven other crime types" and causing "total non-drug crime to fall by 9.4% in the long-term relative to the rest of London" (1137-8).

Discussion