the effects of beverage type on homicide rates in russia, 1970–2005

The effects of beverage type on homicide rates in Russia, 1970–2005dar_310 257..262

ANDREW STICKLEY1,2 & YURY RAZVODOVSKY3

1Department of Global Health Policy, Graduate School of Medicine, University of Tokyo,Tokyo, Japan, 2Stockholm Centreon Health of Societies in Transition (SCOHOST), Södertörn University, Huddinge, Sweden, and 3Grodno State MedicalUniversity, Grodno, Belarus

AbstractIntroduction and Aims. Previous research fromWestern Europe and North America has suggested that consuming differenttypes of alcoholic beverage may have differing effects on homicide rates both within and between countries.The aim of this studywas to examine the relation between the consumption of different beverage types and homicide rates in Russia across thelater-Soviet and post-Soviet periods. Design and Methods. Age-standardised male and female homicide data for the period1970–2005 and data on beverage-specific alcohol sales were obtained from the Russian State Statistical Committee (Rosstat).Time series analysis (autoregressive integrated moving average modelling) was used to examine the relation between the sale(consumption) of different alcoholic beverages and homicide rates. Results. Total alcohol consumption and vodka consumptionas measured by sales were significantly associated with both male and female homicide rates: a 1 L increase in overall alcoholsales would result in a 5.9% increase in the male homicide rate and a 5.1% increase in the female homicide rate.The respectivefigures for vodka were 16.4% and 14.3%.The consumption of beer and wine was not associated with changes in homicide rates.Discussion and Conclusions. Our findings suggest that the consumption of distilled spirits has had an especiallydetrimental impact on lethal violence in Russia from at least 1970 onwards. In order to reduce homicide rates in this context,alcohol policy should focus on reducing overall consumption as well as attempting to shift the beverage preference away fromdistilled spirits. [Stickley A, RazvodovskyY.The effects of beverage type on homicide rates in Russia, 1970–2005. DrugAlcohol Rev 2012;31:257–262]

Key words: alcohol, distilled spirits, Russia, homicide, vodka.

Introduction

Although the consumption of alcohol has been linkedwith both aggression and with all forms of violence[1,2], evidence suggests that this relation is conditionalon a number of factors and varies between differentcultures [3], drinking contexts and environments [4],and individuals [5]. At the population level it has beenargued that while more drinking leads to greater vio-lence, this relation is shaped by the drinking pattern asit is stronger in those drinking cultures characterised bymore episodes of acute intoxication [6]. Individual-leveldifferences impacting on the alcohol–aggression/violence relation possibly include the suppression ofanger [5], a predisposition for violence [7], and heavyor problematic drinking [8,9].

One factor which has been examined in relation toboth aggression and violence at different levels ofanalysis is the type of alcoholic beverage consumed.Studies undertaken in both experimental and natural-istic settings have shown that beverage-specific effectson aggression seem to be strongest among those indi-viduals who have consumed distilled beverages[10,11]. As the same result has been obtained aftercontrolling for differences in blood alcohol levels[10], it has been argued that this difference maybe related to differing expectations surrounding thetype of beverage consumed rather than theactual alcohol content of it [12]. Other potentialexplanations may include a differing sensitivity toprovocative cues induced by different alcoholic bever-ages or differences in pharmacologically active

Andrew Stickley PhD, Lecturer, Yury Razvodovsky PhD, Research Scientist. Correspondence to Dr Andrew Stickley, Department of GlobalHealth Policy, Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,Tokyo 113-0033, Japan.Tel: +81 3 5841 3398; Fax:+81 3 5841 3637; E-mail: [email protected]

Received 26 September 2010; accepted for publication 8 February 2011.

bs_bs_banner

R E V I E W

Drug and Alcohol Review (May 2012), 31, 257–262DOI: 10.1111/j.1465-3362.2011.00310.x

© 2011 Australasian Professional Society on Alcohol and other Drugs

substances and their effects between different bever-age types [10].

Beverage-specific effects on violence have also beenobserved at the aggregate level—especially in relation tolethal violence. Studies from across North America[13,14] and from Western Europe [15,16] have high-lighted how different beverage types can have differingeffects on homicide both within and between countries.This has given rise to the suggestion that beverage-specific effects are strongest in the culture in which aparticular beverage predominates [16], that is, that it isnot the alcohol concentration of a particular drinkwhich is important in terms of lethal violence but ratherthe extent to which it is used and its role within thedrinking culture of a specific society [17].

The current study was built on and extended thisearlier work by examining the impact of beverage typeon homicide in Russia during the period 1970–2005.There is reason to believe that examining the relation inthis setting may be particularly instructive. Russia cur-rently has one of the highest homicide rates in the world[18], while research has revealed a close relationshipbetween alcohol and homicide in the country both inthe present and the past [19–21]. As yet however, therehave been no previous attempts to examine if what isbeing drunk, that is, beverage type, has a direct impacton homicide.This is an important omission as findingsfrom neighbouring Belarus have indicated that theremay be beverage-specific effects on homicide in thecountries in this region [22]. Moreover, a recent study,while not examining this issue directly, has neverthelesssuggested that the same may be true for Russia [23].

Another important reason for examining this relationis because of the sharp fluctuations that have occurredin the level and type of alcohol consumed in Russiaacross the study period. Alcohol consumption roseappreciably in Russia and other Soviet countries in the1970s [24] before levelling off in the early 1980s andthen falling sharply in 1985–1987 during Mikhail Gor-bachev’s anti-alcohol campaign [25]. The collapse ofthe Soviet Union and the opening up of the Russianalcohol market in the early 1990s was not only associ-ated with a sharp rise in alcohol consumption [26], butit also led to a change in the beverage mix. According tosurvey data, male consumption of vodka fell sharplybetween 1995 and 2004 while that of beer increasedrapidly [27].

Against this backdrop, determining the exact relationbetween what is being drunk and the occurrence oflethal violence may have important implications forattempts to reduce the high levels of violence which theWorld Health Organization has recently described as a‘major public health and social threat in the RussianFederation’ [28]. The aim of the current study wastherefore to examine the relation between the con-

sumption of different beverage types and homiciderates in Russia across the later-Soviet and post-Sovietperiods, 1970 to 2005.

Methods

Data

Age-adjusted, sex-specific homicide data (rates per1 000 000 of the population) are taken from theRussian Statistical Committee’s (Rosstat) vital statisticsregistration system. Rosstat’s cause of death classifica-tion system has undergone several changes in recentdecades. Until 1988 the cause of death classificationwas based upon the Soviet nomenclature which had alimited number of causes of death in comparison withthe International Classification of Diseases (ICD)system. From 1989 to 1998, Rosstat used a codingscheme that was based on ICD-9. From 1999 a newsystem of coding based on ICD-10 was introduced.Rosstat issued a table of correspondence between itsclassification system and ICD-9 and ICD-10 [29], andit has been claimed that the Russian system of codingwas and is compatible with the ICD codes [30]. Thus,in the current study, Rosstat’s code for homicide (andinjury purposely inflicted by other persons)—174 cor-responds with the ICD-9 codes E960.0–E969.0 andICD-10 codes X85-Y09. The data on per capitabeverage-specific alcohol sales (in litres of pure alcohol)were taken from Rosstat’s annual reports.

Statistical analysis

To examine the relation between changes in the con-sumption of different types of alcoholic beverage andhomicide mortality across the study period, a timeseries analysis was performed using the statisticalpackage ‘Statistica’ (StatSoft Inc., Tulsa, OK, USA).Bivariate correlations between the raw data from twotime series can often be spurious due to commonsources in the trends and due to autocorrelation [31].One way to reduce the risk of obtaining a spuriousrelation between two variables that have commontrends is to remove these trends by means of a ‘differ-encing’ procedure, as expressed in Equation 1:

∇ = − −X X Xt t t 1 (1)

This means that rather than analysing trends we analy-sed to what degree annual changes in per capita alcoholconsumption are related to changes in homicide rates.It was Box and Jenkins [32] who first proposed themethod for the time series analysis that we use. It iscommonly referred to as ARIMA (autoregressive inte-grated moving average) modelling. We used this model

258 A. Stickley &Y. Razvodovsky


specification to estimate the relationship between thetime series homicide rate and beverage-specific alcoholsales in this paper. In accordance with previousaggregate-level studies, we estimated semi-logarithmicmodels with a logged output.The following model wasestimated, as expressed in Equation 2:

∇ = + ∇ + ∇LnH a A Nt t tβ (2)

where ∇ indicates that the series is differenced, H refersto the homicide rate, a indicates the possible trend inhomicide mortality due to other factors than thoseincluded in the model, A represents alcohol sales, b isthe estimated regression parameter, and N is the noise(error) term. The percentage increase in the homiciderate that is associated with a 1 L increase in alcoholsales can be calculated by using the formula:(exp(b1) - 1)*100.

Results

Across the whole period the male homicide rate was 3.2times higher than the female rate (279.5 vs. 86.9 per1 000 000). The average per capita alcohol consump-tion figure was 8.2 L [standard deviation (SD): 2.03 L]with vodka being the drink overwhelmingly consumed(4.44 L, SD 1.5) when compared to wine (2.39 L, SD0.99) or beer (1.33 L, SD 0.58). However, these meanfigures mask differing trends among the beveragesacross the period.While there has been a slight drop invodka sales from 4.84 L in 1970 to 3.88 L in 2005 andwine sales have remained at roughly the same level,

there has been a sharp growth in beer sales—especiallyin recent years. Between 1998 and 2005, the per capitaconsumption figure for beer rose from 1.16 to 3.08 L.As mentioned previously, it is also worth noting thattotal beverage sales have experienced sharp fluctuationsacross the period. Thus, an especially sharp fall wasrecorded in vodka and wine sales in 1984–1987 thatcoincided with Mikhail Gorbachev’s anti-alcohol cam-paign. Similarly, the collapse of the Soviet Union andthe (short-term) ending of the state’s alcohol monopolyin the early 1990s was accompanied by a sharp rise invodka sales.

During the period 1970 to 2005, the male homiciderate in Russia increased by 3.5 times (from 113.7 to401.3 per 1 000 000 of the population) and the femalerate by 2.4 times (from 47.1 to 113.6).There were alsosharp trends in the time series data across the studyperiod.These trends were removed by means of a first-order differencing procedure. After prewhitening of thedata, the cross-correlations between beverage-specificalcohol sales and the homicide mortality time serieswere inspected. This indicated that there was a statisti-cally significant cross-correlation between total percapita alcohol/vodka sales and homicide mortality formen and women at first lags 0 and 1 (Table 1).

At the same time, there were no cross-correlationsbetween prewhitened wine/beer sales and homiciderates: these series were therefore not included in themodel estimations. The outcome of the bivariateARIMA models are presented in Table 2. The esti-mated effects of total alcohol as well as vodka sales onthe homicide rate are clearly statistically significant for

Table 1. Results of the cross-correlation analysis between the prewhitened time series for men and women: effects of beverage-specific percapita alcohol sales on the male and female homicide rates

Lag

Alcohol sales Vodka sales Wine sales Beer sales

r SE r SE r SE r SE

Men-3 -0.239 0.177 -0.144 0.177 -0.188 0.170 -0.062 0.177-2 -0.246 0.174 -0.081 0.174 -0.306 0.174 -0.119 0.174-1 -0.125 0.172 0.344 0.172 -0.357 0.172 -0.167 0.1720 0.372 0.169 0.606 0.167 -0.106 0.169 -0.037 0.1691 0.351 0.172 0.438 0.172 0.106 0.172 -0.087 0.1722 0.284 0.174 0.058 0.174 0.226 0.174 -0.112 0.1743 0.102 0.177 -0.353 0.177 0.196 0.177 -0.246 0.177

Women-3 -0.208 0.177 -0.129 0.177 -0.148 0.174 -0.079 0.177-2 -0.188 0.174 -0.122 0.174 -0.270 0.174 -0.103 0.174-1 -0.606 0.172 0.337 0.172 -0.246 0.172 -0.226 0.1720 0.360 0.169 0.605 0.169 -0.042 0.169 -0.002 0.1691 0.314 0.172 0.411 0.172 0.135 0.172 -0.051 0.1722 0.280 0.174 0.012 0.174 0.286 0.174 -0.110 0.1743 -0.003 0.177 -0.407 0.177 0.129 0.177 -0.177 0.177

Beverage type and homicide in Russia 259


both sexes: a 1 L increase in overall alcohol sales wouldresult in a 5.9% increase in the male homicide rate andin a 5.1% increase in the female homicide rate, while a1 L increase in vodka sales would result in a 16.4%increase in the male homicide rate and a 14.3%increase in the female homicide rate.

Discussion

This study has shown that although overall alcoholconsumption and the consumption of distilled spirits(vodka) were associated with the occurrence of homi-cide in Russia during the later-Soviet and post-Sovietperiods, no effects were observed for the consumptionof either wine or beer. Moreover, these relations wereobserved for both men and women, despite their verydifferent levels of alcohol consumption.

Before discussing these findings, several limitationsof this study must be mentioned. Official alcohol salesdata were used as a measure of alcohol consumptionacross the period. However, the unrecorded consump-tion of alcohol is widespread and relatively high inEastern Europe [23,33] and was commonplace inRussia throughout the study period [24,26]. Bye [23]has highlighted how the effects of this will depend onwhether it was strongly positively or negatively corre-lated with recorded alcohol consumption—or in thepresent case, with the consumption of specific bever-ages, as this can result in either an underestimation oroverestimation of the alcohol relation. Measuringalcohol consumption and its effects in Russia is alsocomplicated by the process of alcohol substitution.Nemtsov has claimed that in response to Gorbachev’santi-alcohol campaign the population intensified theproduction of moonshine (‘samogon’) as early as 1987[34]. If this did happen and moonshine is a substitutefor vodka it also has the potential to distort the natureof the relation we observed between vodka and homi-cide. There may also have been problems with thehomicide mortality data. Although vital statistics dataare regarded as the most reliable measure of homicidein Russia [35] and were considered valid and reliableduring the Soviet period [36], more recent researchsuggests that there may be some problems with their

recording in recent years [37]. Finally, there was also arisk of omitted variable bias in this work. However, arecent study has suggested that alcohol has an indepen-dent effect on violence even when other factors arecontrolled for [38].

This is the first study to examine if there arebeverage-specific effects on homicide in Russia. Thefinding that only vodka was associated with homicideseemingly supports earlier multi-country researchwhich has suggested that beverage-specific effects arestrongest in the cultures in which a beverage predomi-nates [16,23].This would certainly seem to be the casein Russia as vodka continues to be the beverage thatdominates in terms of consumption. Survey data from2001 suggest that over 60% of all alcohol is consumedin the form of spirits [39] and possibly as much as70–80% if illegally produced distilled spirits are alsoconsidered [40].

Although Rossow [16] found no evidence for a spiritsonly effect in her multi-country study examining thealcohol-homicide relation in Western Europe, there areseveral factors which may result in vodka having anespecially detrimental effect in terms of homicide inRussia. Earlier Western research has suggested thatdrinkers who become aggressive are likely to havecertain characteristics including approving of the use ofaggression and to be heavy/problem drinkers who havea high alcohol intake per drinking occasion [9,41].Thisis important as spirits drinkers seem to not only havethe highest levels of consumption amongst drinkers ofdifferent alcohol beverage types [42], but also, thisgroup may contain a greater proportion of heavy drink-ers [42]. If heavy drinkers are responsible for a largeshare of alcohol-related homicides as has been hypoth-esised [13], then this augurs particularly badly forRussia: according to official figures there were nearly2.2 million people registered with alcoholism and alco-holic psychoses in 2005 [43]. However, Nemtsov hasargued that not only is the real figure two to three timeshigher, but there are even more heavy drinkers anddrunkards in Russia—to the extent that they constituteapproximately one-third of the population of adult men[44].This figure is not surprising given his estimate thatadult men consume as many as 155 bottles of vodka

Table 2. Estimated effects of beverage specific alcohol sales on homicide rates

Total alcohol sales Vodka sales

Model Estim. P Model Estim. P

Homicide men 0.1.0 0.057 0.016 1.1.0 0.152 <0.001Homicide women 0.1.0 0.050 0.018 1.1.0 0.134 <0.001



annually, that is, three a week on average. Thisextreme consumption of vodka might also explain whyhe has linked 70% of all homicides in Russia withalcohol [44].

The effects of drinking spirits may also be exacer-bated by the way they are drunk as a heavy episodicdrinking pattern is widespread. Moreover, this patternof consuming alcohol is much more frequent whendrinking spirits than other beverage types among bothmen and women in Russia [45]. According to a recentstudy, 28% of men and 4% of women consume at least200 g of strong spirits (86+ g of pure alcohol) on oneoccasion at least once every 2–3 weeks [45].The heavydrinking of spirits in Russia may result in high levels oflethal violence due in part, to the so-called ‘Mellanbyeffect’, that is, the more rapid rise and higher bloodalcohol concentrations and levels of impairment thatresult from drinking spirits rather than beer [42]. Thismight be important in a Russian context as the weaponof choice among Russian homicide offenders who havebeen drinking (the knife) may suggest a greater impul-siveness of action [46] possibly facilitated by quickerand deeper intoxication. This idea draws support fromearlier research which has shown that heavier levels ofintoxication are connected with an increased immediaterisk of violent behaviour [17].

The findings from the present study have importantimplications for alcohol policy in Russia and othercountries around the world more generally. At thepopulation level as more drinking has been associatedwith greater violence and less drinking with less vio-lence [6], one benefit of reducing overall national salesof alcohol would be to lower the homicide rate [47].The potential benefits of this action were clearly dem-onstrated in Russia during Gorbachev’s anti-alcoholcampaign when a reduction in the per capita consump-tion of ethanol from 14.2 to 10.7 L in 1984 to 1987 wasaccompanied by a fall in the male homicide rate from19.3 to 11.5 per 100 000 [4,6,25].

However, there was also a change in the share ofbeverage sales (between 1980 and 1987) with vodkasales falling by over 10% [25]. Our findings suggest thatattempts to reduce overall consumption should also belinked with efforts through tax policy to shift the bev-erage preference away from distilled spirits. This isespecially important as the sharp rise in mortality(including external causes of death) that has occurredin post-Soviet Russia has been linked to the excessiveconsumption of spirits fuelled by a sharp drop in therelative price of vodka [48]. Further, as the consump-tion of spirits is associated with heavy and problematicdrinkers, the provision of adequate state treatmentfacilities and other treatment options (such as Alcohol-ics Anonymous) for this population of substanceabusers may have beneficial effects in terms of reduced

rates of alcohol-related violence, including homicide ashas been seen in other parts of the world [13,47].

References

[1] Bushman BJ. Effects of alcohol on human aggression: valid-ity of proposed explanations. In: Galanter M, ed. Recentdevelopments in alcoholism, volume 13: alcohol and vio-lence. Hingham, MA: Kluwer Academic Publishers,1997:227–43.

[2] Parker RN, Auerhahn K. Alcohol, drugs, and violence.Annu Rev Sociol 1998;24:291–311.

[3] MacAndrew C, Edgerton RB. Drunken comportment: asocial explanation. Chicago, IL: Aldine, 1969.

[4] Graham K, Leonard KE, Room R, et al. Current directionsin research on understanding and preventing intoxicatedaggression. Addiction 1998;93:659–76.

[5] Norström T, Pape H. Alcohol, suppressed anger and vio-lence. Addiction 2010;105:1580–86.

[6] Room R, Rossow I. The share of violence attributable todrinking. J Subst Use 2001;6:218–28.

[7] Felson RB, Teasdale B, Burchfield KB. The influence ofbeing under the influence: alcohol effects on adolescentviolence. J Res Crime Delinq 2008;45:119–41.

[8] Boyatzis RE. The predisposition toward alcohol-relatedinterpersonal aggression in men. J Stud Alcohol 1975;36:1196–207.

[9] Graham K, Schmidt G, Gillis K. Circumstances whendrinking leads to aggression: an overview of research find-ings. Contemp Drug Probl 1996;23:493–557.

[10] Gustafson R. Male alcohol-related aggression as a functionof type of drink. Aggress Behav 1999;25:401–8.

[11] Murdoch DD, Pihl RO. The influence of beverage type onaggression in males in the natural setting. Aggress Behav1988;14:325–35.

[12] Pihl RO, Smith M, Farrell B. Alcohol and aggression inmen: a comparison of brewed and distilled beverages. J StudAlcohol 1984;45:278–82.

[13] Mann RE, Zalcman RF, Smart RG, Rush BR, Suurvali H.Alcohol consumption, Alcoholics Anonymous membership,and homicide mortality rates in Ontario 1968 to 1991.Alcohol Clin Exp Res 2006;30:1743–51.

[14] Parker RN, Cartmill RS. Alcohol and homicide in theUnited States 1934–1995—or one reason why U.S. rates ofviolence may be going down. J Crim Law Criminol 1998;88:1369–98.

[15] Norström T. Effects on criminal violence of different bev-erage types and private and public drinking. Addiction1998;93:689–99.

[16] Rossow I. Alcohol and homicide: a cross-cultural compari-son of the relationship in 14 European countries. Addiction2001;96 (Suppl 1):S77–S92.

[17] Darke S. The toxicology of homicide offenders and victims:a review. Drug Alcohol Rev 2010;29:202–15.

[18] Chervyakov VV, Shkolnikov VM, Pridemore WA, McKeeM. The changing nature of murder in Russia. Soc Sci Med2002;55:1713–24.

[19] Pridemore WA. Vodka and violence: alcohol consumptionand homicide rates in Russia. Am J Public Health 2002;92:1921–30.

[20] Stickley A, Carlson P. Alcohol and homicide in early 20th-century Russia. Contemp Drug Probl 2005;32:501–25.

[21] Stickley A, Pridemore WA. The social-structural correlatesof homicide in late-tsarist Russia. Br J Criminol 2007;47:80–99.

Beverage type and homicide in Russia 261


[22] Razvodovsky YE. Association between distilled spirits con-sumption and violent mortality rate. Drug-Educ Prev Policy2003;10:235–50.

[23] Bye EK. Alcohol and homicide in Eastern Europe: a timeseries analysis of six countries. Homicide Stud 2008;12:7–27.

[24] Treml VG. Alcohol in the USSR: a statistical study.Durham, NC: Duke University Press, 1982.

[25] Shkolnikov VM, Nemtsov A. The anti-alcohol campaignand variations in Russian mortality. In: Bobadilla JL, Cos-tello CA, Mitchell F, eds. Premature death in the newindependent states. Washington, DC: National AcademyPress, 1997:239–61.

[26] Nemtsov AV. Estimates of total alcohol consumption inRussia, 1980–1994. Drug Alcohol Depend 2000;58:133–42.

[27] Perlman FJA. Drinking in transition: trends in alcoholconsumption in Russia 1994–2004. BMC Public Health2010;10:691.

[28] World Health Organization. Milestones of a global cam-paign for violence prevention 2005: changing the face ofviolence prevention. Geneva: WHO, 2005.

[29] Gavrilova NS, Semyonova VG, Evdokushkina GN, GavrilovLA. The response of violent mortality to economic crisis inRussia. Popul Res Policy Rev 2000;19:397–419.

[30] Razvodovsky YE. Beverage-specific alcohol sale and suicidein Russia. Crisis 2009;30:186–91.

[31] Norström T, Skog OJ. Alcohol and mortality: methodologi-cal and analytical issues in aggregate analyses. Addiction2001;96 (Suppl 1):S5–S17.

[32] Box GEP, Jenkins GM. Time series analysis: forecastingand control. San Francisco, CA: Holden-Day, 1976.

[33] Popova S, Rehm J, Patra J, Zatonski W. Comparing alcoholconsumption in Central and Eastern Europe to other Euro-pean countries. Alcohol Alcohol 2007;42:465–73.

[34] Nemtsov A. Suicides and alcohol consumption in Russia,1965–1999. Drug Alcohol Depend 2003;71:161–8.

[35] Pridemore WA. Measuring homicide in Russia: a compari-son of estimates from the crime and vital statistics reportingsystems. Soc Sci Med 2003;57:1343–54.

[36] Wasserman D, Värnik A. Reliability of statistics on violentdeath and suicide in the former USSR, 1970–1990. ActaPsychiatr Scand 1998;394 (Suppl 1):34–41.

[37] Gavrilova NS, Semyonova VG, Dubrovina E, EvdokushkinaGN, Ivanova AE, Gavrilov LA. Russian mortality crisis andthe quality of vital statistics. Popul Res Policy Rev 2008;27:551–74.

[38] Bye EK. Alcohol and violence: use of possible confoundersin a time-series analysis. Addiction 2007;102:369–76.

[39] Pomerleau J, McKee M, Rose R, Haerpfer CW, Rotman D,Tumanov S. Drinking in the Commonwealth of Indepen-dent States—evidence from eight countries. Addiction2005;100:1647–68.

[40] Pridemore WA, Chamlin MB. A time-series analysisof the impact of heavy drinking on homicide and suicidemortality in Russia, 1956–2002. Addiction 2006;101:1719–29.

[41] Wells S, Graham K. Aggression involving alcohol: relation-ship to drinking patterns and social context. Addiction2003;98:33–42.

[42] Smart RG. Behavioral and social consequences related tothe consumption of different beverage types. J Stud Alcohol1996;57:77–84.

[43] Rosstat. Zdravookhranenie v Rossii 2009: statisticheskiisbornik [The health service in Russia: a statistical anthol-ogy]. Moscow: Rosstat, 2009.

[44] Nemtsov AV. Alkogol’naya istoriya Rossii: noveishii period[A contemporary history of alcohol in Russia]. Moscow:‘Librokom’, 2009.

[45] Pomerleau J, McKee M, Rose R, Haerpfer CW, Rotman D,Tumanov S. Hazardous alcohol drinking in the formerSoviet Union: a cross-sectional study of eight countries.Alcohol Alcohol 2008;43:351–59.

[46] Pridemore WA. An exploratory analysis of homicidevictims, offenders, and events in Russia. Int Crim JusticeRev 2006;16:5–23.

[47] Darke S. Substance use and violent death: a case for the ‘toohard’ basket? Addiction 2009;104:1063–64.

[48] Treisman D. Death and prices: the political economy ofRussia’s alcohol crisis. Econ Transit 2010;18:281–331.



the effects of beverage type on homicide rates in russia, 1970–2005

Documents