[ Pobierz całość w formacie PDF ]
//-->Homicide Studieshttp://hsx.sagepub.com/Alcohol and Homicide in Eastern Europe : A Time Series Analysis of SixCountriesElin K. ByeHomicide Studies2008 12: 7DOI: 10.1177/1088767907310851The online version of this article can be found at:http://hsx.sagepub.com/content/12/1/7Published by:http://www.sagepublications.comOn behalf of:Homicide Research Working GroupAdditional services and information forHomicide Studiescan be found at:Email Alerts:http://hsx.sagepub.com/cgi/alertsSubscriptions:http://hsx.sagepub.com/subscriptionsReprints:http://www.sagepub.com/journalsReprints.navPermissions:http://www.sagepub.com/journalsPermissions.navCitations:http://hsx.sagepub.com/content/12/1/7.refs.htmlDownloaded fromhsx.sagepub.comby Anna Dom on October 30, 2012Alcohol and Homicide inEastern EuropeA Time Series Analysis of Six CountriesElin K. ByeNorwegian Institute for Alcohol and Drug Research, OsloHomicide StudiesVolume 12 Number 1February 2008 7-27© 2008 Sage Publications10.1177/1088767907310851http://hs.sagepub.comhosted athttp://online.sagepub.comFew studies have addressed the association between alcohol consumption and homiciderates at the population level in eastern European countries. The aim of the present studywas to test hypotheses on how this association may vary across countries with differentdrinking patterns and for gender specific homicide rates. Time series analysis was usedon annual alcohol consumption and homicide rates for six eastern European countries.The estimates were pooled into two groups of countries with more (Russia and Belarus)and somewhat less (Bulgaria, Hungary, Poland, and former Czechoslovakia) hazardousdrinking patterns. The overall results showed that annual changes in alcohol consumptionwere positively and significantly associated with homicide rates and also indicated thatthe estimates were stronger in countries with a more detrimental drinking pattern. Theresults suggest that alcohol consumption has an effect on homicide rates in EasternEurope and that this effect varies with drinking pattern.Keywords:homicide; alcohol consumption; Eastern Europe; time seriesAlcohol consumption has been found to be an important ingredient in violent inci-dents, and an important factor for the prevalence of violence (Pernanen, 2001; Room& Rossow, 2001). Alcohol-related violence is a serious social problem and a substan-tial threat to public health in many countries (Babor et al., 2003; Room et al., 2002),and it has been estimated that 24% of homicides are attributable to alcohol worldwide(Rehm et al., 2004; World Health Organization [WHO], 2002). In recent years, inter-personal violence has received increased attention both by the public and in research.A number of studies using various designs and methods have demonstrated a positiveassociation between alcohol and violence, and alcohol consumption has been found toAuthor’s Note:This article was written within the framework of the projectAlcohol in Eastern Europe,which is funded by the Swedish Council for Working Life and Social Research (FAS). The data were pro-vided by Mats Ramstedt and Jonas Landberg at SORAD, Sweden. The author feels very grateful toIngeborg Rossow for valuable comments and discussions on earlier drafts of this article. Two anonymousreferees have also provided valuable comments on an earlier version of this article. Correspondence con-cerning this article should be addressed to Elin K. Bye, Norwegian Institute for Alcohol and DrugResearch, POB 565 Sentrum, N-0105, Oslo, Norway; e-mail: ekb@sirus.no.7Downloaded fromhsx.sagepub.comby Anna Dom on October 30, 20128 Homicide Studiesbe a common factor in violent incidents, and an important risk factor for committingviolent acts and for victimization (Lenke, 1990; Pernanen, 2001; Roizen, 1997).Alcohol intake and number of heavy drinking episodes have also been shown toincrease the risk of being involved in violence (Rossow, 1996; Wells, Graham, &West, 2000).Several theories address the relationship between alcohol and violence (Parker &Auerhahn, 1998; Parker & Rebhun, 1995; Pernanen, 1991), and it is suggested thatalcohol intoxication may play a significant role in enhancing and triggering aggres-sive behavior and thereby may increase the risk of violent victimization (Pernanen,1991). Recently, empirical studies have shown that the consequences of alcohol con-sumption are related to the pattern of drinking as well as to the volume (Rehm et al.,1996), and the importance of drinking pattern is an important aspect of the harmcaused by alcohol (Midanik, Tam, Greenfield, & Caetano, 1996; Rehm et al., 1996,2003; Room, Bondy, & Ferris, 1995; Rossow, 1996; Wells, Graham, Speechley, &Koval, 2005). However, intoxication may not only enhance aggressive behavior butalso increase the risk of violent victimization. The likelihood of any bystandersintervening in a violent incident may also be reduced (Pernanen, 1991). Thus, themechanisms underlying the observed alcohol-violence relationship are complex, andsuch complexity may be difficult to assess in studies at the individual level. To whatextent a given increase in overall alcohol consumption may influence violence ratesin society is of considerable interest for public policy, and it has been argued thataggregate time series data is the most feasible approach for this kind of question(Norström & Skog, 2001). Some methodological problems that arise at the individ-ual level, such as selection effects and failure to control for potentially confoundingfactors, are avoided with this method, although aggregate studies do not completelysolve the inference problem of causality.There are at least three possible mechanisms for the alcohol-violence associationat the aggregate level, and the first two mechanisms express a linear association(additive), and the third a semilogarithmic association (multiplicative). First,because studies at the individual level suggest that the association between alcoholand violence is mainly related to acute intoxication (Rossow, 2000), one can hypoth-esize that an increase in total consumption is followed by an increase in the level ofviolent behavior. This is based on the assumption that an increase in total consump-tion would imply an increase in the number of drinking occasions and thereby in thenumber of events of acute intoxication (Greenfield, Midanik, & Rogers, 2000;Horverak, 2006; Rossow, 2007) in which violent behavior may be triggered, more orless irrespective of societal factors (Rossow, 2004). Second, as the proportion ofheavy drinkers increases with an increase in total consumption (Skog, 1980, 1985)the effect may also be due to an increase in the number of heavy drinkers who havea heightened risk of being involved in violent acts (Rossow, 2000). Third, theabsolute effect of changes in alcohol consumption on violent behavior may dependon other factors associated with violence, that is, the effect of alcohol consumption onDownloaded fromhsx.sagepub.comby Anna Dom on October 30, 2012Bye / Alcohol and Homicide in Eastern Europe 9violent behavior may depend on the level of other factors of importance to violence; forexample, the effect of an increase in alcohol consumption on violent behavior may begreater in times of high unemployment or poverty, and it might also depend on the levelof informal social control (Parker & Cartmill, 1998; Rossow, 2001). In this case, the alco-hol effect would be relative, that is, a nominal increase in alcohol consumption wouldimply a relative change in the level of violence. Consequently, we would expect that anincrease in drinking in the population would lead to an increase in the level of violence.In line with these expectations, several time series analyses of aggregate-level data havedemonstrated that an increase in alcohol consumption is followed by an increase in ratesof violence (Bye, 2007; Lenke, 1990; Norström, 1998; Rossow, 2004; Skog & Bjørk,1988). Correspondingly, a reduction in violence has been demonstrated in countries withsudden and large changes in alcohol consumption because of rationing, antialcohol cam-paigns, and strikes (Lenke, 1990; Nemtsov, 1998; Rossow, 2002).Differential Associations Between PopulationDrinking and Violence RatesStudies have demonstrated that in addition to being associated with the level of con-sumption, the alcohol-violence association is also mediated by patterns of drinking andcultural expectations about behavior while drinking (Room & Rossow, 2001). Thisindicates that an increase in per capita consumption of one liter would lead to a largerincrease in violence in countries with an intoxication-oriented drinking pattern, com-pared to countries with a less “explosive” drinking pattern. A comparative study of 14western European countries (European Comparative Alcohol Study [ECAS]) investi-gated whether a one-liter increase in per capita consumption implied a larger increase inviolence in a country where drinking often leads to intoxication (Northern Europe) asopposed to that in countries with a less-explosive drinking pattern (Southern Europe).The results from this study confirmed the hypothesis of a drinking pattern gradient inthe alcohol effect. The strongest association was found in the northern Europeancountries with the most hazardous drinking pattern (13% per liter) and the weakest inthe southern European countries with the least hazardous drinking pattern (6%), withmid-Europe in-between (9%; Rossow, 2001). Correspondingly, in the WHO’s GlobalBurden of Disease project, the relative risk for alcohol’s role in violence and in injurieswas found to vary across countries and subregions by drinking patterns (Rehm et al.,2003). Regarding possible gender differences in alcohol-related violence, studies indi-cate that both offenders and victims of violent behavior are generally more often men(Pernanen, 1991; Rossow, 1996; Rossow & Hauge, 2004). As men drink more oftenand more heavily than women (Babor et al., 2003), it is assumed that the associationbetween total consumption and level of violence is stronger for men than for women.This was also supported in the ECAS study, where changes in alcohol consumptionwere found to have a greater impact on homicide rates for men than on homicide ratesfor women (Rossow, 2001).Downloaded fromhsx.sagepub.comby Anna Dom on October 30, 201210 Homicide StudiesPrevious time series studies of the association between alcohol and violence havemostly been confined to western European countries and North America, and thereis sparse knowledge about this association in other countries, for example, easternEuropean countries. The high levels of alcohol consumption in eastern Europeancountries in combination with detrimental drinking patterns would suggest highlevels of alcohol-related harms as well as expected strong association betweenchanges in alcohol consumption and violence. Previous cross-sectional and one time-series analysis of Russian data have supported this, and demonstrated a significantassociation between alcohol consumption and homicide rates (Pridemore, 2002,2004; Pridemore & Chamlin, 2006).Thus, although a positive association between alcohol and violence has been foundin Western Europe, United States, and Canada, little is known about this association inthe eastern European countries, and how the association might differ between thesecountries and between western European countries. In addition to providing newknowledge about the association between alcohol and violence in several easternEuropean countries, there are several other reasons why a study of the associationbetween alcohol and violence in Eastern Europe is an issue of great interest. Alcoholconsumption and homicide rates are considerably higher in eastern European countriesthan in western European countries, life expectancy is lower, and mortality rates arehigher (Cockerham, 1999). Whereas alcohol consumption is declining in most westernEuropean countries, it is rising in the countries of Central and Eastern Europe. The risein mortality from cardiovascular disease and violence is one of the reasons for the gapin life expectancy between East and West (Bobak & Marmot, 1996).Another issue is the extensive political, social, and economic changes that havetaken place in many of the countries during the years prior to and following thecollapse of the Soviet Union. Many of the eastern European countries face severesocioeconomic challenges such as economic recession, falling living standards, andunemployment (Cockerham, 1999). Alcohol consumption and homicide rates haveincreased in most eastern European countries during the transition period, and it hasbeen shown that countries with higher income and established market economies tendto have lower rates of mortality from homicide than countries with lower income(WHO, 2002). This is clearly seen in Russia where the homicide rate is among thehighest in the world. The homicide rate more than tripled between 1988 and 1994.Recent studies in Russia and Eastern Europe suggest that the primary causes ofincreased alcohol consumption and homicide rates are related to widespread socialproblems because of social stress and disorganization during the transition to a freemarket (Gavrilova, Semyonova, Evdokushkina, & Gavrilov, 2000; Kim & Pridemore,2005; Leon & Shkolnikov, 1998; Pridemore & Spivak, 2003). This is in line with thestructural strain perspective introduced by Durkheim, where rapid social changeswould create anomic conditions; that is, cause strain in individuals who fail in termsof prevailing social expectations, and this strain could increase the risk of both heavydrinking and alcoholism and thus violence (Messner & Rosenfeld, 1997).Downloaded fromhsx.sagepub.comby Anna Dom on October 30, 2012
[ Pobierz całość w formacie PDF ]