Info
National Indicator for Reducing Drug Related Crime Published
Sunday/05/Oct 2008
Some news
for those who followed last years setting of the new Public Service
Agreements and 198 Local Government Indicators with as much nerd-like interest as I did. You'll
recall that apart from the obvious difficulties entailed in setting
up an evaluation framework before a strategy has been written, the
somewhat naive faith in proxy measures of treatment throughput to
measure the effectiveness of national and local approaches to
recovery and the seeming absence of any measures related to
reintegration, employment or housing, that some of the indicators
actually hadn't been defined yet.
National Indicator 38 which sets out to measure improvements in levels of drug related crime is one of the indicators whose methodology for measurement is included in a new publication out for consultation last week from the Department for Communities and Local Government.
The new metrics for this indicator are interesting. What is proposed is to measure the effectiveness of partnerships efforts to reduce drug related crime by comparing the actual criminal activity of a DAT level cohort of people who test positive for class A drugs and people identified through OASys (the criminal justice individual database - Offender Assessment System) with the estimated or forecast criminal activity of that cohort. Essentially what this means is that someone somewhere will be looking at the likelihood of a certain group of people known to police and services offending in the next year. If less than the expected number offend (or are caught offending) then the local partnership is deemed to have done a good job. If more, then the local partnership will have done a bad job.
I am not a criminologist, and me and statistics have at best a love-hate relationship, but this seems a bit of a shaky basis on which to measure the effectiveness of our approach to tackling drug related crime. First of all, as we know not everyone who commits a crime is caught. Secondly, many things other than drug use and treatment affect peoples propensity to commit or not commit offences - so any change in performance against this criteria may not be attributable to the responses of the partnership. Thirdly, the process used to estimate the offending rate of the cohort - the figure that provides the "baseline" for the measure - is 'Response Surface Methodology' - which uses a range of variables to identify a likely outcome. This means that we will be basing our measurement on a proxy built on an estimate based on a series of assumptions informed by a narrow perspective on the relationship between drug use and crime . I'd be very happy to hear others views on this and would like to be convinced it is of more value that it appears at first sight ... please do get in touch if you think you can make this make more sense for me.
The fact that this indicator does not require any additional local data collection is a plus. I'm also grateful to the authors of the work for giving me one of my favourite phrases in any government document so far this year - the priceless and almost poetic "breach is an expression of CJS grip".
You can find out more about this consultation by downloading the new definitions and responding to the DCLG consultation. Responses are due in by 31st October.
National Indicator 38 which sets out to measure improvements in levels of drug related crime is one of the indicators whose methodology for measurement is included in a new publication out for consultation last week from the Department for Communities and Local Government.
The new metrics for this indicator are interesting. What is proposed is to measure the effectiveness of partnerships efforts to reduce drug related crime by comparing the actual criminal activity of a DAT level cohort of people who test positive for class A drugs and people identified through OASys (the criminal justice individual database - Offender Assessment System) with the estimated or forecast criminal activity of that cohort. Essentially what this means is that someone somewhere will be looking at the likelihood of a certain group of people known to police and services offending in the next year. If less than the expected number offend (or are caught offending) then the local partnership is deemed to have done a good job. If more, then the local partnership will have done a bad job.
I am not a criminologist, and me and statistics have at best a love-hate relationship, but this seems a bit of a shaky basis on which to measure the effectiveness of our approach to tackling drug related crime. First of all, as we know not everyone who commits a crime is caught. Secondly, many things other than drug use and treatment affect peoples propensity to commit or not commit offences - so any change in performance against this criteria may not be attributable to the responses of the partnership. Thirdly, the process used to estimate the offending rate of the cohort - the figure that provides the "baseline" for the measure - is 'Response Surface Methodology' - which uses a range of variables to identify a likely outcome. This means that we will be basing our measurement on a proxy built on an estimate based on a series of assumptions informed by a narrow perspective on the relationship between drug use and crime . I'd be very happy to hear others views on this and would like to be convinced it is of more value that it appears at first sight ... please do get in touch if you think you can make this make more sense for me.
The fact that this indicator does not require any additional local data collection is a plus. I'm also grateful to the authors of the work for giving me one of my favourite phrases in any government document so far this year - the priceless and almost poetic "breach is an expression of CJS grip".
You can find out more about this consultation by downloading the new definitions and responding to the DCLG consultation. Responses are due in by 31st October.
|
Something Old Something New - The 2008 Drug Strategy
Friday/29/Feb 2008