European Journal of Public Health

European Journal of Public Health, Vol. 23, No. 2, 340–344 ß The Author 2012. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. doi:10.1093/eurpub/cks079 Advance Access published on 11 July 2012


Alcohol tax, consumption and mortality in tsarist Russia: is a public health perspective applicable?
¨ m1,*, Andrew Stickley2,3,* Thor Norstro
1 Swedish Institute for Social Research (SOFI), Stockholm University, S-109 91 Stockholm, Sweden ¨ derto ¨ rns ho ¨ gskola, 141 89 Huddinge, Sweden 2 Stockholm Centre on Health of Societies in Transition (SCOHOST), So 3 Department of Global Health Policy, Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan ¨ m, Swedish Institute for Social Research, Stockholm University, S-106 91 Stockholm, Correspondence: Thor Norstro Sweden, tel: +46 8 16 23 14, fax: +46 8 15 46 70, e-mail: totto@sofi.su.se

Downloaded from http://eurpub.oxfordjournals.org/ at Florida International University on June 4, 2013

*These authors contributed equally to this work.

Background: The public health perspective on alcohol comprises two main tenets: (i) population drinking impacts on alcohol-related harm and (ii) population drinking is affected by the physical and economic availability of alcohol, where alcohol taxes are the most efficient measure for regulating consumption. This perspective has received considerable empirical support from analyses of contemporary data mainly from Europe and North America. However, as yet, it has been little examined in a historical context. The aims of the present article are to use data from tsarist Russia to explore (i) the relation between changes in the tax on alcohol and per capita alcohol consumption and (ii) the relation between per capita alcohol consumption and alcohol mortality. Methods: The material comprised annual data on alcohol taxes, alcohol consumption and alcohol mortality. The tax and alcohol consumption series spanned the period 1864–1907 and the mortality data covered the period 1870–94. The data were analysed by estimating autoregressive integrated moving average models on differenced data. Results: Changes in alcohol taxes were significantly associated with alcohol consumption in the expected direction. Increases in alcohol consumption, in turn, were significantly related to increases in alcohol mortality. Conclusion: This study provides support for the utility of the public health perspective on alcohol in explaining changes in consumption and alcohol-related harm in a historical context. We discuss our findings from tsarist Russia in the light of experiences from more recent alcohol policy changes in Russia.

state and its intervention, and the strong readiness to find alternative sources of alcohol to those being made less available.6 The current study will thus examine the impact of one important measure that was implemented in tsarist Russia in 1863 to mitigate alcohol-related problems, namely the excise tax system on alcohol. The system meant that alcohol was taxed per gradus, that is, per degree of pure alcohol in each vedro (‘bucket’, equivalent to 12.3 l). The tax increased gradually from 4 kopeks per gradus at its introduction to 11 kopeks per gradus, four decades later.7 Previous assessments of the effects of tax policy in this period have been largely negative. Both historical and contemporary commentators have argued that not only did it fail to decrease drinking8 but that it may have rather been associated with an increase in drunkenness.9 Moreover, although official figures suggest that a fall did occur in alcohol consumption during the existence of the excise duty system,10 it has been argued that it was not due to tax increases. Rather, fluctuations in alcohol consumption arose due to the ‘deeper conditions of the national economy and national life’11—a finding reiterated by tsarist officials themselves at that time11 and subsequently by other researchers.12 However, there is scope for a more rigorous test of the relation between taxes and population drinking; modern statistical tools have not been applied in this context before. But even if such analyses—which will be discussed later in the text—would suggest the existence of a tax effect on consumption, another possibility could also exist. Thus, it has been claimed that as tax increased, so did the growth in contraband alcohol,11 and that the excise system resulted in the illegal sale of (untaxed) vodka on a large scale.6


Iperspective on alcohol’ is consistent with the experience of tsarist
Russia. This perspective has two main tenets (as laid down in, e.g. Bruun et al.).1 The first one is that an increase in per capita alcohol consumption tends to be accompanied by increased rates of alcohol-related harm,2 including chronic outcomes, such as liver cirrhosis and pancreatitis, and various forms of violent death, for example, fatal accidents, homicide and suicide. The second tenet is that per capita consumption responds to changes in the physical and economic availability of alcohol; alcohol taxes are thus generally regarded as the most efficient measure for regulating consumption.3,4 The two main principles have received considerable empirical support, mainly from analyses of post-war data pertaining to Western Europe and North America. However, there is also some evidence from Eastern Europe that seems to provide support for this perspective. The potential for alcohol policy to impact on health in this part of the world was clearly seen during the final years of the Soviet regime. Although Gorbachev’s anti-alcohol campaign—which made alcohol less available and drastically more expensive—was ultimately short-lived (1985–87), it has nevertheless been estimated that it saved more than 1 million lives in Russia alone during these years.5 A more crucial test of the model, though, would be to subject it to data from a context where the cultural and political conditions are little favourable for its success. A feasible candidate for such a test seems to be tsarist Russia with its deep-rooted drinking culture characterized by ‘heavy periodic excesses’,6 scepticism toward the

n the present article, we will explore whether the ‘public health

The tax and alcohol consumption data span the period 1864–1907. but with more normal variation in the Danish alcohol series. did not reveal any significant impact of alcohol consumption. In light of the situation outlined earlier in the text. which includes explanatory variables not considered in the model.18 We used the mortality indicator labelled ‘sudden deaths from drunkenness’. it came out ¨ m’s analysis of this relaas clearly significant. 2013 Methods Data on alcohol consumption and alcohol taxes relate to the whole of the Russian Empire. Population data were compiled from multiple sources21. the fact that the halving of recorded per capita consumption in France during the post-war period was accompanied by a concomitant decrease in cirrhosis mortality. Ultimately. only 14 observations. ‘buckets’) that we converted into litres of 100% alcohol.Applicability of public health perspective model 341 One way of assessing whether the decrease in official consumption is real is to check to what degree it is matched by a decline in alcohol-related mortality. which appeared during the later-tsarist period. Norstro tionship based on a much longer period (1931–80).22 that were cross-checked for accuracy. The alcohol tax series was adjusted for the upward trend in inflation by deflating it with a general price index. A simple differencing procedure was sufficient to remove non-stationary trends. the Baltic States and the part of Russia that is to the west of the Urals).13 By analysing the relation between recorded per capita consumption and alcohol-related mortality during the tsarist period. The presence of strong time trends in the data (figure 1) necessitated some form of filtering to achieve the stationarity required for ARIMA modelling. a fairly weak alcohol effect was expected.7.24 Although having few observations does not induce bias.20 We will refer to this measure as alcohol mortality (alcohol being the sole cause).10 Where possible. The model residuals Downloaded from http://eurpub.23 often referred to as autoregressive integrated moving average (ARIMA) modelling. we will estimate the direct effect of changes in the alcohol tax on alcohol-related mortality. which makes it worthwhile to undertake an analysis. is allowed to have a temporal structure that is modelled and estimated in terms of autoregressive or moving average parameters.org/ at Florida International University on June 4.e.e. the aim of the current study was to use unique data specifically retrieved for the present study. on the basis of Danish timeseries data spanning 1911–24. 1870–94.19 seem to have 10 9 8 7 6 5 4 3 2 1 0 1860 1865 1870 1875 1880 1885 1890 1895 1900 1905 1910 MORTALITY TAX ALCOHOL Figure 1 Trends in alcohol mortality (rate per 100 000).26 Although the alcohol series for the tsarist period is less volatile than the early Danish one. Moldova. An example of this is a study by Skog that estimated the effect of per capita alcohol consumption on suicide. However. that is. whereas the mortality data were available for a more restricted period i. These were obtained from two major works on alcohol. we may thus shed some light on the authenticity of the downward trend in alcohol consumption. yearly changes were analysed. comprised deaths from severe alcohol poisoning. were levied in kopeks per gradus across this period. The number of observations in the mortality series is on the low side (25) compared with the recommended 30–50 observations. Such an analysis is in line with the approach best known from the studies by Cook14 (see ref. these data have been cross-checked with other official data.16 The consumption data relate to the annual per capita consumption of 40 vodka in vedros (i. and per capita alcohol consumption (litres 100%) in Russia . rather than using the raw series. which were supposedly the subject of forensic medical investigation. alcohol tax (kopeks per gradus). pertaining to tsarist Russia to examine (i) the relation between changes in the tax on alcohol and per capita alcohol consumption and (ii) the relation between per capita alcohol consumption and alcohol-related mortality. Differencing reduces the risk of obtaining spurious correlations greatly because an omitted variable is more likely to be correlated with the explanatory variable as a result of common trends than due to the synchronization of yearly changes. this can be compensated by high variability in the input series.15 for a review of studies in this research tradition). makes the downward trend in recorded French drinking more credible. as mentioned earlier. The data were obtained from the ‘Russian Information Anthology. 1896’—issued by the Central Statistical Committee of the Russian Ministry of Internal Affairs. it does contain some changes that are marked. Statistical analysis The data were analysed by using the approach developed by Box and Jenkins. but owing to the marked reduction in consumption due to World War 1. In contrast. Further. For instance. In addition.17 Mortality data pertain to the 50 provinces of European Russia (which covers an area that today coincides with Ukraine. Belarus.25 A priori. that is. the noise term. These fatalities. the outcome will show if the power is sufficient to detect any effect.oxfordjournals. Alcohol taxes. it does entail a loss of power.

5%. a 1% increase in per capita consumption was associated with a 1. and. All models were satisfactory with respect to residual autocorrelation. or the cross-correlations will be contaminated by the autocorrelation of the series. the study period was not identical across the three model estimations. Although this pattern accords with our expectations. negative. it has recently been claimed that none of the previous alcohol systems introduced throughout Russian history have been successful in terms of controlling Russian drinking.20 Indeed.28 The analysis revealed significant correlations at lag 0 (0. This made hardly any difference to the outcome. whereas higher lags were insignificant.69) SE 0.3% increase in alcohol mortality (table 2). and At-1 indicates alcohol consumption lagged 1 year. Finally. with no need to include noise parameters. In such an analysis. SE = 0. the estimated alcohol effect increases to 1. Also. Log–log model estimated on differenced time-series data a: Box–Ljung test for residual autocorrelation.18 P 0. However. At is alcohol consumption at time t. this was assessed empirically through inspecting the cross-correlations between the two series at different lags. we identified a plausible cause by establishing a significant association with the increasing alcohol tax.001)].8.27 We should therefore expect a lag structure between alcohol and mortality. the estimate implies that a 1% tax increase yielded a decrease in consumption of $0. 1870–94 Elasticity Alcohol (weighted) Dummy 1879 Diagnostics Qa(5) 1.67.32 0. the dependent and the explanatory variable were logarithmically transformed) for quantifying the relationships at issue. it should be noted that there was a marked upward shift in the mortality series in 1879.03 (P > 0. first. this was tested with the Box– Ljung Q statistic. Although some commentators maintain that it is probable that the consumption decline in this period was real. one assessing the relation between alcohol taxes and alcohol mortality.0 was used for the analyses.45 3. Discussion During the last three decades of the 19th century.88 (SE = 0.e. the estimated effect of tax changes on per capita alcohol consumption was statistically significant and in the expected direction. the second one assessing the relation between consumption and alcohol mortality. which can be seen by the outcome presented in table 3. Because the mortality data was available for a more restricted time period (1870–94) than the tax and consumption data (1864–1907). population drinking in tsarist Russia was almost halved according to official records. The outcome does support the notion of a genuine decrease in consumption. the estimated effect of consumption on mortality was significant. 1870–94 Elasticity Tax Dummy 1879 Diagnostics Qa(5) –0. As the lag-weights are proportional to the cross-correlations. the model in table 1 was estimated for the period 1867–1904).20) and lag 1 (0.33 5. lastly.32) SE 0.g. it indicates by how many percent the dependent variable is expected to increase if the independent variable increases by 1%.26 0. it was expected that a change in taxes would affect mortality. SPSS 17. this model form seems applicable also for estimating the relation between alcohol consumption and alcohol mortality because it can accommodate the non-linear (accelerating) risk function between alcohol and alcohol mortality. this occurred in parallel with increased alcohol taxes.342 European Journal of Public Health Table 1 Estimated effect (ARIMA model) of alcohol excise tax on per capita alcohol consumption in Russia. Second. As can be seen from table 1. 1864–1907 Elasticity Tax Diagnostics Qa(5) –0.07 P 0. historical experience suggests that attempts by the state to reduce alcohol consumption in this country tend to be undermined by an increased supply of various forms of surrogate alcohol. However. Alcohol mortality may result from acute intoxication and chronic long-term abuse. the increasing trend in alcohol taxes coincides with decreasing trends in alcohol consumption and mortality. that is. 2013 Log–log model estimated on differenced time-series data a: Box–Ljung test for residual autocorrelation.oxfordjournals.31) SE 0.28 we constructed a new alcohol series according to: AWt ¼ ð0:46 Ã At þ 0:67 Ã At À1 Þ=1:13 where AWt denotes the weighted alcohol series at time t. at time t À 1. this was indeed confirmed. This was captured by including a dummy variable that takes the value 0 before 1879 and 1 otherwise.001 Log–log model estimated on differenced time-series data a: Box–Ljung test for residual autocorrelation. [If the dummy variable for the upward shift in mortality (in 1879) is not included. lag 5 Table 2 Estimated effect (ARIMA model) of per capita alcohol consumption (weighted) on alcohol mortality in Russia.29 The resultant effect measure (the elasticity coefficient) is easy to interpret.001 Downloaded from http://eurpub. graphical evidence of this kind is not compelling and thus we proceeded to the stronger tests provided by model estimations based on the differenced data. We used log-log models (i.org/ at Florida International University on June 4. On the basis of these results. lag 5 Table 3 Estimated effect (ARIMA model) of alcohol excise tax on alcohol mortality in Russia. and that things were instead getting worse.027 <0. it is necessary to first remove all structure from the series (pre-whiten). which is documented for several outcomes. even sudden deaths are often associated with a history characterized by excessive drinking. the parameter estimates and their degree of statistical significance remained more or less the same. This is the standard model in demand analyses and is thus a feasible specification for estimating the tax effect on consumption. SE = 0.017 should not differ from white noise. The analysis thus comprised the estimation of three separate models: the first one assessing the relation between alcohol taxes and the per capita consumption of alcohol.07 P 0.001 <0.77 0.40.89 (P > 0. the models reported in tables 1–3 were estimated on somewhat shorter time periods by dropping 3 years at each end (e. implying that the downward trend in registered consumption may have been compensated by increased unrecorded consumption. P < 0. lag 5 As a robustness test. that is.91 (P > 0. Results As can be seen in figure 1.44 5.6 other commentators from that period have suggested that there was no real reduction in the harm caused by alcohol.46.33 0. we identified a significant effect of per capita .9 The present study is the first to analyse historical data with modern statistical techniques to shed some light on this issue.20).

However.oxfordjournals. the current study has suggested that alcohol pricing policy may be an effective vehicle to reduce both consumption and mortality in Russia—with the proviso that it takes place in a more highly regulated market environment where the capacity to acquire cheaper illegal alcohol alternatives is eliminated to a much greater degree. considerable leverage could be expected from moderating the latter. the marked upwards shift in mortality in 1879 is a reminder of the influence of unidentified factors.30. Edwards G.41 At the population level. then the government will have to find a way to regulate the 50% of the alcohol market that it does not currently control in Russia. Examples of such factors include the socio-economic and demographic characteristics of the population under study. although lowering population drinking obviously is expected to impact on rates of harm. the importance of drinking patterns has to an increasing extent come to the fore. References 1 2 Bruun K. it can be argued that the public health perspective on alcohol seems to apply to the experience in tsarist Russia. the sharp growth in both the consumption of ‘samogon’ (moonshine) and alcohol surrogates in recent years. it is necessary to point out a number of potential limitations of the study. ¨ m T. if any. Study limitations Before discussing the implications of our findings. Drug Alcohol Rev 2005.  Our results highlight the potential utility of the public health perspective in terms of understanding the tax-alcohol consumption relation in a historical context. was associated with a fall in alcohol mortality. During the period 1990–94. Indeed. Russia’s leading alcohol researcher. Further. Alexandr Nemtsov.42 This study has highlighted how looking at the past may be a useful mechanism to guide policy in the present—despite the multitude of changes that take place in a given society across time. it does not seem likely in the present case. which has been driven by their lower price. This suggests that if price is to be used as a weapon in an attempt to lower drinking in the contemporary period.40 Second. This points to the necessity of wider interventions. have yielded sensible results with estimates consistent across countries and time periods. for example. what can be regarded as the inverse of the process that has been depicted in this article occurred about 100 years later in post-Soviet Russia. the likely presence of unrecorded consumption mentioned earlier may have biased the estimate of the alcohol effect on mortality upwards (if there was a positive correlation between recorded and unrecorded consumption) or downwards (if there was a negative correlation between recorded and unrecorded consumption).31 and there is little reason to believe that Russian data would fare worse. which in turn led to a decrease in alcohol mortality. has argued that ‘Russian statistics were reliable for their time’. Although the existence of such a factor cannot be precluded. as always in this kind of work.Applicability of public health perspective model 343 consumption on alcohol mortality.34. Policy implications On the basis of our findings suggesting that increased alcohol taxes were associated with lowered per capita consumption. to assess which. which in turn. it should also be kept in mind that we performed the analyses on differenced data.24:537–47. whereas the mortality of. falling alcohol consumption and lowered mortality. however. . 1975. This finding seemingly underpins recent research indicating continuity in a harmful drinking culture in Russia across time36 and which has suggested that what happens in the past may also hold important policy lessons for the present.32 and the mortality data provided by the Central Statistical Committee was the most complete during these years in terms of the number of deaths recorded. which reduces the risk of omitted variable bias. those with higher education in fact improved. Lumio M. this is manifested in the way that the association between per capita consumption and rates of harm tends to be stronger in countries where binge drinking and drinking to intoxication characterize the drinking pattern. it has been little researched in a historical context. it is striking that the experience from Gorbachev’s anti-alcohol campaign echoes the relation we observed between stricter alcohol policy. this study has shown that increases in alcohol taxes were associated with a fall in alcohol consumption. There is little information available. To this end. First. A highly relevant finding in the present context is that the sharp mortality increase after 1990. and minimal and stepped price increases (depending on alcohol concentration) for all beverages containing 28% alcohol or more by volume. They also suggest that taxes may be a useful mechanism to reduce the high levels of alcohol consumption in contemporary Russia— provided the alcohol market can be properly regulated. However. Ramstedt M. Helsinki: Finnish Foundation for Alcohol Studies.33 Nevertheless. It can further be argued that the results from the three separate analyses support each other so that the joint findings are stronger than their constituent parts. Mortality and population drinking: a review of the Norstro literature. Specifically. which supports the public health perspective on alcohol. Lastly.35 Thus. two important qualifications need to be noted in this context. affected less privileged groups.org/ at Florida International University on June 4. 2013 Key points  Although much contemporary empirical evidence has been gathered from Europe and North America. Further. previous analyses based on western European nineteenth century time-series data on alcohol consumption and mortality. Writing recently about statistical data from the nineteenth century.39 also points to the dangers of using alcohol tax in a market that is not fully regulated.  Using data from historical Russia. and the marked reduction in the latter would hardly have occurred in the absence of lowered population drinking. Alcohol control policies in public health perspective. Downloaded from http://eurpub. As Russia exhibits the devastating combination of a high level of consumption and detrimental drinking patterns.37 Our findings taken together with these more contemporary experiences suggest that alcohol tax policy may be one mechanism by which the exceptionally high rates of alcohol consumption underpinning the current mortality crisis in Russia can be tackled. such as improving the poor everyday living conditions that in part seem to underpin harmful drinking or at least its fatal effects. the shift is consistent with what one should expect from a change in recording practice.3 implying that the harm per litre of alcohol is elevated in such countries.38 Although this is to be welcomed. Conflicts of interest: None declared. there exist a large number of factors that may have affected the outcome but that are not considered in the analyses due to a lack of data. as yet. An obvious concern when working with historical records is the quality of the data relied on. above all. a prerequisite for an omitted factor to bias our estimates is that—in addition to affecting the outcome—it is also synchronized with the explanatory variable in the annual changes. of these forms of bias is most likely. et al. the real price of vodka dropped by almost 80% spurring a sharp rise in alcohol consumption and concomitant mortality. the Russian government has recently introduced a minimum price of 89 roubles for a standard bottle of vodka. but we have not been able to locate any more specific information on this.

In: Statisticheskii vremennik Rossiiskoi Imperii.18:281–331.org/ at Florida International University on June 4. Ryabushinskii. 10 Osipov NO. The composition of surrogate alcohols 39 McKee M. Testing causal hypotheses about correlated trends: pitfalls and remedies. Alcohol and suicide in Russia. cirrhosis.29:1884–8. 17 Mironov BN. Petrograd: K. 12 Dmitriev VK. Alcohol and suicide in Scandinavia. Alkogolizm i bor’ba s nim [Alcoholism and the fight against it]. editor.45:586–8. Addiction 2011. Russia: alcohol yesterday and today. ¨ m T. Addiction 2005. Casswell S. Bobak M. et al.88:1189–93. Petersburg: P. St. 1916:14–94. London: P. Jukkala T. Tsentral’nyi statisticheskii komitet Ministerstva Vnutrennikh Del. Norstro 1956-2005: Evidence for the continuation of a harmful drinking culture across time? J Stud Alcohol Drugs 2011..344 3 4 5 6 7 8 9 European Journal of Public Health [Calendar for Physicians of All Departments in 1916]. Obozrenie nasil’stvennykh i vnezapnykh smertei v 1870-74 godakh [A review of violent and sudden unexpected deaths in 1870-74]. Su zcs S. P. 14 Cook P. . Alcohol Alcohol 2010. Jenkins GM. A contemporary history of alcohol in Russia. 23 Box GEP. Redwood City: Addison-Wesley. Petersburg: Tsentral’nyi statisticheskii komitet M. San Diego. Washington: National Academy Press.oxfordjournals. Per capita alcohol consumption and total mortality: an analysis of 30 Norstro historical data. St. 20 Sikorskii IA. Letter from Russia: minimal price for vodka established in Russia from 1 January 2010. In: Norstro postwar Europe: consumption. Obzor glavneishikh dannykh po demografii i sanitarnoi statistike Rossii [A review of the main data on demography and public health statistics of Russia].72:341–7. et al. consequences and policy responses in 15 European countries. St. 2013 28 Wei WWS. 1882:XCVII-CIX. Zaigraev GG. Tobler AL. Vinnaya monopoliya v Rossii [The alcohol monopoly. New York: Oxford University Press. Introduction to time-series analysis and forecasting: with applications of SAS and SPSS. Nicholson A.96(Suppl):S5–17. Caetano R. Ltd.100: 2270–8. 34 Skog OJ. No. Effectiveness and cost-effectiveness of policies and programmes to reduce the harm caused by alcohol.63:307–19. 40 Nemtsov A. In: Kazennaya prodazha vina [The state sale of alcohol]. Addiction 2001. The effect of liquor taxes on drinking. Alcohol and risk of physical harm. 26 Norstro 27 Petersson B.15:565–607. 2011. 00 ´ rva ´ ry S. Volume II. Inc.373:2234–46. The widening gap in mortality by educational level in the Russian Federation. 2000. 32 Nemtsov A. Voprosy nervno-psikhicheskoi meditsiny 1901. editors.83:553–9. 1870-1894 and 36 Stickley A. Alcohol and mortality: methodological and analytical issues 35 Norstro in aggregate analyses. particularly sudden un¨ . 15 Wagenaar AC. Am J Public Health 2006. Anderson P. Kriticheskie issledovaniya o potreblenii alkogoliya v Rossii [Critical research on the consumption of alcohol in Russia]. ¨ m T. King & Son. ¨ derto ¨ rns ho So ¨ kinen IH. Sociol Res 2010. Trends in population drinking. 1990. Econ Transit 2010. Seriya II. Babor T. 1918. Am J Public Health 2010. 1879 (and all subsequent years until 1898). Edwards G. Addiction 1993. Addiction 2003. Lancet 2009. Petrograd: Pravda. number 14: Population trends in European Russia in 1870]. Kalendar’ dlya vrachei vsekh vedomstv na 1916 god Downloaded from http://eurpub. Addiction 2002. 2002:49–82. ¨ m T. Sa consumed in Russia. Alcoholism and drunkenness in Russia: ways to come out of the crisis situation.97:1413–25. Series 2. New Brunswick: Rutgers Center of Alcohol Studies. Death and prices: the political economy of Russia’s alcohol crisis. In: Bulatov PN. et al. Alcohol and public policy. The relationship of average volume of alcohol consumption and patterns of drinking to burden of disease: an overview. Tom II. Oxford: Oxford University Press. ¨ m T. Vinnaya monopoliya. Addiction 1988. 16 Zolotarev AM. 2nd edn. Huddinge. vypusk 19. ¨ m T. St. St. Rikker. 1987. Segal BM. Moscow: V. Analysis of the role of alcohol in mortality. V. Alcohol in 13 Leifman H. Contemp Drug Probl 1988. 1910:LXXVIIICXVI. Seriya II. Sweden.100:146–9. Potreblenie vina v Rossii v 1904-1909 godakh [Alcohol consumption in Russia in 1904-1909].91:339–44. 38 Jargin SV. Stockholm: Almqvist & Wiksell. L. editors. et al. Wages and Prices in Imperial Russia. Skog OJ. [Russian statistical yearbook 1909]. ¨ m T. Petersburg.96: 1293–9. editor. and auto accidents. Graham K. Fuhr DC. Time series analysis: forecasting and control. The role of alcohol in the Russian mortality crisis. 1980-2001. Petersburg: Tsental’nyi statisticheskii komitet Ministerstva Vnutrennikh Del. 11 Fridman MI. Suicide mortality of Eastern European regions before and after the 33 Ma Communist period. Petersburg. Alcohol and suicide in Denmark 1911-24—experiences from a ‘natural experiment’. 41 Rehm J.6:537–41. 1916. Soikin. drinking patterns. London: Holden-Day. vypusk 14: Dvizhenie naselenie v evropeiskoi Rossii za 1870 god [Statistical almanac of the Russian Empire. In: Holder HD. Chisholm D. 1909.69: 47–72. Effects on alcohol tax and price policies on morbidity and mortality: a systematic review. Gerstein DR. Russian drinking: use and abuse of alcohol in pre-revolutionary Russia. Some notes on the state sale monopoly and subsequent prohibition of vodka in Russia. 1896]. 2010. Simpson JY.106: 31 Norstro 1957–65.23: witnessed death. Alkogolizm v Rossii v XIX stoletii i bor’ba s nim [Alcoholism in Russia in the nineteenth century and the fight against it]. Time series analysis: univariate and multivariate methods. In: Moore MH. Room R. 1995: 82–113. 19 Zverinskii V.98:1209–28. 24 Yaffe RA. 22 Statisticheskii vremennik Rossiiskoi Imperii. 1897. Nemtsov AV. 1976. 18 Sbornik svedenii po Rossii. Russ Rev 2010. Alcohol-related human losses in Russia in the 1980s and 1990s. D. S. Addiction 1996. Soc Sci Med 2006. in middle-aged men in Malmo 259–63. Alcohol: no ordinary commodity. In: Ezhegodnik Rossii 1909 g. Research and public policy. 1896 [Russian Information Anthology. 1703–1913. Alcohol Clin Exp Res 2005. Sweden: ¨ gskola. 21 Novosel’skii SA. 1911. 1900:1–77. 29 Anderson P. 1981:255–85. Alcohol Alcohol 1988. 42 Murphy M. Petersburg: Tsentral’nyi statisticheskii komitet Ministerstva Vnutrennikh Del. P. 37 Treisman D.49:3–18. 25 Skog OJ. Istoricheskii ocherk vzimaniya piteinykh sborov v Rossii [A historical essay on the collection of potatory taxes in Russia]. St. The alcohol monopoly in Russia]. CA: Academic Press. Alcohol and public policy: evidence and issues. Komro KA.

Sign up to vote on this title
UsefulNot useful