Jahanshahi and McVie
Recent changes to stop and search legislation, policy and practice in Scotland has significantly reduced the use of the tactic. In a wider context of rising violence in the UK, this has raised questions over the extent to which the reduction in stop and search has impacted on crime rates, and especially violence.
This paper addresses the following research questions:
· What is the relationship between rates of stop and search and patterns of crime in Scotland?
· Has the introduction of a new code of practice on 11th May 2017 caused a ‘shock’ in terms of this relationship?
· Does this relationship vary according to type of search power used and type of crime studied?
· Is there any variation in the relationship between stop and search and crime rates across Police Divisions? And is this contingent on level of deprivation across divisional areas?
Type(s) of inequality and how inequality is defined:
The paper examines spatial differences in the relationship between search and crime across Scotland’s division, which takes account of the possibility of geographical difference in policing practice. Using the Scottish Index of Multiple Deprivation and other measures of economic stress within Scottish neighbourhoods, we aim to examine whether there is equality in the application of stop and search to neighbourhoods according to their crime rates.
Approach or method used:
We follow the methodological approach used in a recent paper by Tiratelli, Quinton and Bradford (2018 - Does Stop and Search Deter Crime? Evidence From Ten Years Of London-Wide Data) which applies lagged regression models with fixed effects to control for unobserved variables and to address other potential biases. In addition, we apply multi-level analysis to examine spatial differences of the relationship of interest across Scottish divisions.
The preliminary findings show that, with the exception of drug offences, the relation between lagged stop and search of various powers and different categories of crime rates are very small and mostly insignificant which is in line with the findings of current studies.
For drug offences an increase in the rate of drug searches in any one week/month was associated with a significant increase in the rate of drug crime recorded in the following week/month. In addition, multi-level analysis shows significant differences in the relation between total stop and search and total crime rates across various divisions from the positive significant effect in Tayside to the negative insignificant effect in Lanarkshire.
The introduction of the code of practice indicates some significant positive and negative effects on the relationship between stop and search and crime rates, depending on the specific categories of crimes and searches:
• The introduction of the Code of Practice had a positive impact on the relationship between weapon searches in one month and violent crime rates as well as robbery rates in the following month (there is no weekly effect). The positive effects (given the initial effects were negative) imply that weapon searches in any one month was associated with fewer reduction in violent and robbery crime recorded in the following month after the introduction of the code of practice.
• The introduction of the Code of Practice had a negative impact on the relationship between stolen property searches in one month and dishonesty crime rates in the following month (there is no weekly effect). The negative effect (Given the initial effects was positive) implies that stolen property searches in any one month was associated with fewer increase in dishonesty crime rates recorded in the following month after the introduction of the code of practice.
Implication of findings/relevance to policy:
Our research could not find any strong evidence of a widespread deterrence effect of stop and search (either looking at weekly or monthly change) for Scotland which confirm the findings of Tiratelli, Quinton and Bradford (2018) for England. The paper finds a positive link between drug offences and lagged of drug searches which we believe is due to the other policing techniques and activities (such as hot spot policing) which were confounding in our statistical analysis.