Reducing violence and binge drinking in connection with student parties- Findings from a community intervention project in Stockholm Mats Ramstedt1, Håkan Leifman2, Daniel Müller1, Erica Sundin1 Thor Norström 1, 3Presentation at the second European conference on alcohol law enforcement, Stockholm. November 16. 20121STAD (Stockholm Prevents Alcohol and Drug Problems)2CAN (Swedish council for information on alcohol and other drugs)3Institute for social research, Stockholm university
Background • It has become popular for high-school students in Stockholm to arrange their student parties at restaurants in the city, often with help from “Eventbolag” • Spring of 2007: the police noticed a striking increase in violence and binge drinking in connection with these parties • A community intervention inspired by the RBS-program was developed and launched in the spring of 2008– including supervision, collaboration and education involving several actors
Actorsinvolved in the intervention Police: Created a special “student unit” working only with this intervention , visit all parties at least twice, are available and possible to reach during the whole evening. The liquor license board: extended supervision on restaurants having student parties. Tax authorities: special control of the financial accounts of event companies/restaurants if needed.
Otheractorsinvolved Event companies: agreed to follow specific guidelines e.g. to report all parties to the police and to attend information meetings. Restaurants: agreed to ensure safety in restaurants, to work with RBS, to collaborate with local police etc. STAD: served as link between Event companies and restaurants and authorities and disseminated information to students etc.
Aimof the study The present study aims at evaluating the impact of these interventions by assessing whether violence and binge drinking were reduced in relation to the students parties.
Data 1. Indicator of violence related to student parties: Violence-related emergency room visits on weekday nights (22.00-06.00) during the period 1 April – 31 May i.e. when the majority of student parties takes place. 2. Indicator of violence not related to student parties : Violence-related emergency room visits on weekend nights (22.00-06.00) during the period 1 April – 31 May i.e. when no student parties takes place.
Hypotheses: 1. Decline on weekday nights after 2008 (due to intervention) but not on weekend nights (no intervention)during the period with student parties i.e. April-May 2. No difference expected between development on weekday and weekend nights during periods without student parties (and thus no intervention) i.e. Jan- March and June-Sept
Results Emergency room visits (18-20 yrs) in weekday nights (red line) and weekend nights (blue line) Student period i.e. april-may
Emergency room visits (18-20 yrs) weekday nights (red line) and weekend nights (blue line). Not student party period (Jan – March)
Emergency room visits (18-20 yrs)weekday nights(red line) and weekend nights (blue line). Not student party period (June – Sept.)
Statistical analysis The intervention effect was assessed by means of a difference-in-differences estimation (DiD) DiD is a technique often used by economists when evaluating policy changes that have been implemented in a quasi-experimental fashion In short: has the difference in emergency room visits between weekdays and weekends become significantly larger after the intervention and how much?
Findings The estimated DiD-models confirm the impression of this graphical evidence. The estimated intervention effect corresponded to a reduction of 23% in emergency room visits on weekday nights and was statistically significant. In contrast, the estimated intervention effects in jan-march and june-sept were non-significant.
Estimated means of BAC-levels among boys and girls (18-20 yrs) attending student parties according to random breath tests. .
Summary The main finding suggest that the intervention was associated a reduction in violence in terms of a reduction in emergency room visits for young people with 23 %. In contrast, no significant intervention effects were found in during periods before and after periods with student parties i.e. when no intervention was made. Binge drinking among students were reduced during follow-up.
Limitations The number of data points is on the low side The number of violence-related emergency room visits actually related to student parties are not known BAC-levels at student parties were only measured when the intervention had started and it was thus not possible to estimate an effect of the intervention.
Conclusion The evaluation suggests that this type of intervention is a promising measure for preventing violence that is worthy to be continued. This would also provide additional data that are needed for a more conclusive assessment.
For more information: See http://08student.se/ Or contact: STAD: email@example.com firstname.lastname@example.org Police: email@example.com