Extreme temperature and extreme violence: evidence from Russia

Otrachshenko, Vladimir; Popova, Olga; Tavares, José Article — Published Version Extreme temperature and extreme violence: evidence from Russia Economic Inquiry Provided in Cooperation with: John Wiley & Sons Suggested Citation: Otrachshenko, Vladimir; Popova, Olga; Tavares, José (2020) : Extreme temperature and extreme violence: evidence from Russia, Economic Inquiry, ISSN 1465-7295, Wiley Periodicals, Inc., Boston, USA, Vol. 59, Iss. 1, pp. 243-262, https://doi.org/10.1111/ecin.12936 This Version is available at: https://hdl.handle.net/10419/266710 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ EXTREME TEMPERATURE AND EXTREME VIOLENCE: EVIDENCE FROM RUSSIA VLADIMIR OTRACHSHENKO, OLGA POPOVA and JOSÉ TAVARES We study the relationship between extreme temperatures and violent mortality, employing novel regional panel data from Russia. We find that extremely hot temperatures increase violent mortality, while extremely cold temperatures have no effect. The impact of hot temperature on violence is unequal across gender and age groups, rises noticeably during weekends, and leads to considerable social costs. Our findings also suggest that better job opportunities and lower vodka consumption may decrease this impact. The results underscore that economic policies need to target vulnerable population groups to mitigate the adverse impact of extreme temperatures. (JEL Q54, I14, K42) “For now, these hot days, is the mad blood stirring.” William Shakespeare, Romeo and Juliet, Act 3, Scene 1 I. INTRODUCTION “[T]he prime time for murder is clear: summertime,” states The New York Times. Heightened social interactions and the presence of biological and psychological triggers that prompt violence partially explain why, “in the summer months, the bad guys tend to be deadliest” (Lehren and Baker 2009). Global climate *The authors thank Jason Lindo (co-editor), three anonymous referees, Richard Frensch, Ali Kutan, Igor Makarov, Milena Nikolova, Mariola Pytlikova, and participants at the ASSA 2019 meeting in Atlanta, IOS/APB/EACES summer academy in Tutzing, and research seminars at IOS Regensburg, Curtin University, and the Laboratory for Economics of Climate Change at HSE Moscow for valuable comments. The authors acknowledge the support from Russian Science Foundation (RSCF) grant no. 19-18-00262. Otrachshenko: Senior Researcher, Leibniz Institute for East and Southeast European Studies (IOS), Landshuter Str. 4, Regensburg, 93047, Germany. Graduate School of Economics and Management, Ural Federal University, Yekaterinburg, Russian Federation. Far Eastern Federal University, Vladivostok, Russia Federation. E-mail votra- [email protected] Popova: Senior Researcher, Leibniz Institute for East and Southeast European Studies (IOS), Landshuter Str. 4, Regensburg, 93047, Germany. Graduate School of Economics and Management, Ural Federal University, Yekaterinburg, Russian Federation. CERGE-EI, a joint workplace of Charles University and the Economics Institute of the Czech Academy of Sciences, Prague, Czech Republic. E-mail [email protected] Tavares: Professor, Nova School of Business and Economics, Campus de Carcavelos, 2775-405, Carcavelos, Portugal. Centre for Economic Policy Research (CEPR), London, UK. E-mail jtav[email protected] change is persistently debated in policy circles and in the media. Beyond the physical changes in the Earth’s environment, it is important to examine possible changes in human behavior having social and economic consequences. Documenting the empirical link between rising temperatures and specific social consequences is a crucial, and quite demanding, task. The most ubiquitous weather-behavior linkage advanced in biology and psychology is the relationship between uncomfortable temperatures and aggressive individual behavior.1 In the first review of psychological literature, Anderson (1989) claimed that the temperature–aggression relationship is complex, and laboratory experiments fail to provide robust evidence on it. As suggested by the author, more field and within-country studies are needed to document more precisely whether uncomfortable temperature and aggression have a Jor U-shaped relationship. In a J-shaped form, only hot temperatures prompt aggression, while cold temperatures do not affect it, while in a 1. For reviews, see Anderson (1989), Anderson et al. (2000), and Hsiang, Burke, and Miguel (2013). A growing body of research also suggests that climate change fosters conflict and warfare. For instance, Burke et al. (2009) and Hsiang, Meng, and Cane (2011) show that weather shocks plausibly impact political stability. Burke and Leigh (2010) and Bruckner and Ciccone (2011) document that weather shocks appear to lead to democratization. Dell, Jones, and Olken (2012) show that adverse temperature shocks increase the probability of irregular leader transitions (i.e., coups). ABBREVIATIONS RusFMD: Russian Fertility and Mortality Database WHO: World Health Organization 243 Economic Inquiry (ISSN 0095-2583) Vol. 59, No. 1, January 2021, 243–262 doi:10.1111/ecin.12936 Online Early publication August 18, 2020 © 2020 The Authors. Economic Inquiry published by Wiley Periodicals LLC on behalf of Western Economic Association International. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 244 ECONOMIC INQUIRY U-shaped form, both cold and hot temperatures increase aggression. In economics, the literature has so far focused on the heat-crime relationship and barely examined whether uncomfortably low temperatures affect violent and aggressive individual behavior. The first economic study of the impact of extreme temperatures on crime was presented by Ranson (2014), who analyzed U.S. historical data and uncovered a positive association between hot weather and crime, extrapolating long-term effects for different scenarios. Recently, Blakeslee and Fishman (2018) have also shown that hot temperature increases violent crime in India. By examining the impact of both extremely hot and extremely cold temperatures, our paper offers an important insight to the ongoing discussion in the literature regarding the shape of the temperature–crime relationship. We fill this gap in knowledge by examining the consequences of extreme temperatures in Russia, a country with a wide range of observed temperatures and one of the highest incidence rates of violence.2 This paper examines the impact of uncomfortably hot and cold days on violent mortality and its unequal incidence across gender and age groups by exploring a dataset on temperature and violence across 79 regions of the Russian Federation between 1989 and 2015. We draw on the cultural, geographic, and climatic diversity of the Russian Federation to estimate the likely impact of an additional high and low temperature day on violent acts. These are individual violent acts leading to death occurring in the course of daily life interactions, not acts driven by political or social unrest. The relevance of our results cannot be escaped, especially as the effects of change may become more acute, and as suggested by other studies, the impact of hot days on mortality may be greater in developing countries.3Violent acts by individuals are hard to predict, so that any 2. According to Soares and Naritomi (2010), Russia is burdened by the largest present value social cost of violence from reduced life expectancy as a share of GDP, immediately after Latin America and the Philippines. Our unique dataset allows us to examine violence perpetrated against women and against men across age groups, on weekdays, and on weekends. 3. Burgess et al. (2017) repeat the exercise reported in Deschênes and Greenstone (2011) for India and find that an additional day with temperatures exceeding 36∘C leads to a rise in the annual mortality rate in India that is about seven times higher than for the United States. These are computed relative to a day in the 22–24∘C range. For the effects of temperature on mortality, see also Karlsson and Ziebarth (2018) for Germany and Otrachshenko, Popova, and Solomin (2017, 2018) for Russia. information that helps us reduce victimization is important and valuable.4 Our paper makes three distinct contributions to the temperature–crime literature. We are the first to examine the temperature–violence relationship in Russia, an upper-middle-income economy with a wide range of observed temperatures and the institutional context that is substantially different from existing studies on the United States. Moreover, we are the first to document that severe cold temperatures do not impact the violent mortality. We also outline in detail the mechanisms behind the relationship between temperature and violence, examine the heterogeneity of the impact by different population groups, and quantify the socioeconomic costs of the temperature–violence linkage. As such, our paper fundamentally revises insights from earlier papers and offers an original contribution to the literature. Though only the “tip of the iceberg,” evaluating the impact on murders overcomes, in part, the underreporting of physical violence and associated consequences, including psychological violence, the latter naturally also being important.5We find that days with average temperatures above 25∘C lead to an increase in both female and male victims, while days with lower temperatures do not affect violent mortality.6 The likelihood of victimization during weekends, as opposed to workdays, rises noticeably for females, suggesting different contexts for the emergence of violence. Our findings also suggest that in regions with higher unemployment and with greater consumption of spirits, the likelihood of victimization during hot days is greater. This suggests that improving economic conditions may help to mitigate the harmful effects of temperature shocks. The remainder of the paper is organized as follows. Section 2 details the conceptual framework underlying the relationship between 4. Hereinafter by “victimization” we mean the process of being victimized or becoming a victim. 5. Cerqueira and Soares (2016) show results indicating that the total welfare cost of homicides in Brazil corresponds to about 78% of Gross Domestic Product, and the yearly welfare cost is about 2.3%. 6. Deschênes and Greenstone (2011) document the relationship between daily temperatures and annual mortality rates, with both relationships exhibiting nonlinearities, with significant increases at the extremes of the temperature distribution. The estimates in Deschênes and Greenstone (2011) suggest that climate change will lead to an increase in the age adjusted US mortality rate of 3% by the end of the twentyfirst century. OTRACHSHENKO, POPOVA & TAVARES: WEATHER AND VIOLENCE IN RUSSIA 245 violence and weather and discusses the earlier literature. Sections 3 and 4 describe the background of our study and data. Methodology and estimation results are presented in Sections 5 and 6, respectively. The last section offers conclusions. II. CONCEPTUAL FRAMEWORK AND RELATED LITERATURE A. Conceptual Framework The impact of weather on violence can be explained from both the supply and demand sides. On the supply side, weather may affect the behavior of potential criminals. On the demand side, weather may affect the behavior of potential victims. Below we outline the possible channels on the supply and demand side in more detail. On the supply side there are three theories that may explain the impact of weather on violence, including economic theory of rational decision making, biological theory, and contact theory. Becker’s model—the canonical model of crime—implies a decrease in crime on hot and cold days, if heat/cold increases the cost of supply of crime. In Becker (1968), an individual’s decision to commit a crime is based on rational consideration of the costs and benefits of the act. In this model, the weather is an input that affects the probability of successfully completing a crime and the probability of escaping undetected thereafter. This explanation has empirical support in cases of cold temperatures, since cold weather associated with natural obstacles to violent crime such as lower mobility due to snow drift, closed doors and windows, and so on (Ranson 2014; Vrij, Van Der Steen, and Koppelaar 1994). However, explaining violence also requires going beyond strictly rational explanations as it may occur as an impulse, not just the result of a search for greater individual utility. The biological explanation is summed up in Simister and Cooper (2005), who suggest that the human body reacts to both extremely cold and extremely hot temperatures by producing stress hormones, including adrenaline, noradrenaline, and testosterone. This leads to the expansion of the blood vessels, increased heart rate and blood pressure, stimulated respiration, focused attention, and heightened anxiety. These same bodily effects are also present when the human body and human brain need to mobilize for action and possible aggression, during stressful or dangerous situations. Noradrenaline is also associated with higher anger levels (Simister and Cooper 2005). Moreover, the interaction of noradrenaline and testosterone fosters aggression (Kemper 1990). Generally, existing epidemiological studies suggest that hot temperatures are more closely related to hormone activation than cold temperatures, as warm clothes reduce the body stress stemming from the cold (Anderson 1989). Another biological link between weather and violence is psychological. Anderson (1989) suggests that violent crime and aggressive behavior during extreme temperatures are driven by an emotional or instinctive state of arousal of the nervous system. Anderson (1989) argues that both extremely hot and cold temperatures are uncomfortable for the human body, facilitating aggression. However, while the relationship between hot temperatures and violence is supported by early laboratory experiments, as in Baron and Bell (1976), findings related to cold temperatures are inconclusive.7 A third possible explanation for the relationship between heat and aggression is an increased frequency in social contacts (Anderson 1989). As people spend more time outside, get together in larger numbers, and go on vacations, opportunities for violent interactions increase. However, Anderson (1987) and Rotton and Frey (1985) find no empirical support for the interactionsviolence explanation and suggest that the impact of extreme temperatures on aggression and crime is not necessarily mediated by the frequency of social contacts. On the demand side, weather may reduce violence by making potential victims more cautious during extreme days. Existing research suggests that people call for police service more often during hot weather (Auliciems and DiBartolo 1995; Brunsdon et al. 2009; Cohn 1993) and less often during cold weather (LeBeau and Corcoran 1990), walk faster during extremely hot (Rotton, Shats, and Standers 1990) and extremely cold days (Liang et al. 2020). The increased level of adrenaline in the blood on hot days may also help a potential victim to either run away from a threat or to use aggression against it (Simister and Cooper 2005). Temperature may also affect 7. This may stem from contradictory effects of neurotransmitters: while the increase in the level of serotonin during cold weather slows aggression down (Reis 1974), another neurotransmitter, acetylcholine, triggers aggression (Myers 1974). 246 ECONOMIC INQUIRY violence via changes in law enforcement. For instance, in periods of hot ambient temperature, police officers tend to be more aggressive toward suspects (Vrij, Van Der Steen, and Koppelaar 1994) and more likely to issue traffic citations (Ryan 2020). However, police officers are also more likely to be assaulted or killed during hot days (Annan-Phan and Ba 2019). All explanations cited guide our empirical analysis and are naturally interrelated. In fact, according to Pakiam (1981), a multiple causation theory of crime prevails, whereby “anthropological-biological, socio-economic and physical environmental causes are possible, with a crime finally being triggered by appropriate psychological and physiological changes” (p. 185). While we cannot separate the abovediscussed mechanisms from each other, the theoretical underpinnings suggest the following hypotheses regarding the impact of hot and cold temperatures on violence: H1a: Extremely hot temperature increases violence. H1b: Extremely cold temperature reduces violence. B. Existing Literature External conditions, including weather conditions, have been shown to affect human judgment and facilitate aggression. For instance, in periods of hot ambient temperature, judges make stricter decisions (Heyes and Saberian 2019), strikes and job quits are more frequent (Simister and Cooper 2005),8drivers sound their horn more often (Kenrick and MacFarlane 1986), and even baseball pitchers hit batters more often (Larrick et al. 2011; Reifman, Larrick, and Fein 1991). Early psychological and epidemiological studies suggest a positive correlation between hot temperatures and crime in the United States (Anderson 1987; Rotton and Frey 1985).9Rotton and Frey (1985) find a positive correlation between hot temperatures and assaults, while Anderson (1987) suggests that this correlation is stronger for violent crimes against other persons—e.g., murders, rapes, and assaults, than 8. In the 1960s “U.S. government officials noted that riots were more likely to occur in warmer weather, and subsequent analysis confirmed this relationship” (Dell, Jones, and Olken 2014, 768, who refer to Carlsmith and Anderson 1979; U.S. Riot Commission 1968). 9. For a review of psychological literature on temperature and crime, see Cohn (1990) and Anderson (1989). for violent crimes against property—e.g., robbery, burglary, larceny, and motor vehicle theft. Furthermore, Rotton and Cohn (2004) find that in air-conditioned locations aggravated assaults are not as likely during hot weather. In economics, several studies examine the temperature–crime relationship in the United States.10 Jacob, Lefgren, and Moretti (2007) and Ranson (2014) have uncovered a positive association between hot weather and different types of crime. Ranson (2014) also suggests that most violent and nonviolent crimes are significantly reduced during cold weather (below 10∘F), while murders are not affected. The literature is also inconclusive regarding the effects of precipitation. Ranson (2014) points out that precipitation does not affect murders, rapes, robbery, or larceny, decreases assaults, and increases manslaughter, burglary, and vehicle thefts. Jacob, Lefgren, and Moretti (2007) find that higher precipitation is associated with a reduction in violent crime. Recently, in studying the impact of pollen concentration on the U.S. crime, Chalfin, Danagoulian, and Deza (2019) use temperature and precipitation as additional controls. The authors find that precipitation reduces both violent and property crime. Using data from India, Blakeslee and Fishman (2018) suggest that hot temperature (above 32∘C) increases violent crime and precipitation decreases it, while property crime is unaffected by either temperature or precipitation. The circumstances behind violent acts differ widely, but it is reasonable to consider whether victimization falls more heavily on specific gender and age groups. Multicountry studies show that 15%–75% of all violence against women is perpetrated by a spouse or domestic partner, and is more prevalent on weekends, when family interactions increase (Aizer 2010; GarciaMoreno et al. 2006; Hidrobo and Fernald 2013; Hindin, Kishor, and Ansara 2008).11 For Russia, Volkova, Lipai, and Wendt (2015) estimate 10. Burke, Hsiang, and Miguel (2015) conduct a metaanalysis and suggest that extreme temperature and precipitation increase the likelihood of interpersonal and intergroup conflict. 11. Gantz, Bradley, and Wang (2006) and Card and Dahl (2011) find that emotional cues associated with the results in games of professional football in the U.S. increase the rate of at-home violence by men against their female partners. Economic difficulties and excessive alcohol consumption also increase the incidence of domestic violence (Aizer 2010; Bobonis, González-Brenes, and Castro 2013; Carpenter and Dobkin 2011; Hidrobo and Fernald 2013; Luca, Owens, and Sharma 2015). OTRACHSHENKO, POPOVA & TAVARES: WEATHER AND VIOLENCE IN RUSSIA 247 that 70%–80% of serious violent crimes, and 30%–40% of murders are committed in the family, with upwards of 10,000 women killed by their close partner. Also, analyzing the impact of weather on violence against females may have important implications for human capital and economic development. For instance, to avoid violence female students in India often compromise on the quality of their education by choosing lower-ranked colleges over higher-ranked ones if a travel route to the latter school is safer (Borker 2017). For different reasons, related or unrelated to gender, older individuals may suffer differential rates of victimization. Otrachshenko, Popova, and Solomin (2017) investigate the impact of days with hot temperature on all mortality causes, as well as cardiovascular-caused mortality, and respiratory-caused mortality. They find that the adult, but not elderly, are the most affected, and people over 60 are relatively less affected. Using data from the 1990s, Soares and Naritomi (2010) find that the incidence of violence in Latin America is concentrated in prime age.12 The same is true for the United States (Levitt 1999), while Russia has a later age profile, with groups around 40–45 the most victimized.13 There is, however, evidence that older women may be especially targeted. Miguel (2005) finds that in Tanzania negative income shocks are associated with a large increase in the murder, by relatives, of elderly women, but not other population groups. Using data for 73 countries, Soares (2006) estimates that each year of life expectancy lost to violence corresponds, on average, to 3.8% of GDP. In Russia, the difference between male and female life expectancies may be a factor, as well as the evolution of relative health status between males and females, with the latter seeing their health degrading more rapidly over time. Cerqueira and Soares (2016) also point out that incorporating heterogeneities such as age and gender has important effects on the estimated welfare cost of deadly violence, leading to a 23% upward correction in total costs. 12. Cerqueira and Soares (2016) explore data from Brazil and find that men in their 20s are about ten times more exposed to homicide than women of similar age. Also, men in their 20s are three times more likely to be victims of homicide than men in their 40s. 13. In Russia, men in their 20s are about five times more exposed than women of the same age, and as exposed as men in their 40s. These are authors’ calculations based on the Russian Fertility and Mortality Database (RusFMD 2016). Therefore, the literature suggests the following hypotheses regarding the heterogeneity of results by gender, age, and work days/weekends: H2a Extreme temperatures affect violence against women more strongly than violence against men. H2b Extreme temperatures affect violence against middle-aged and elderly more strongly than violence against younger individuals. H2c The impact of extreme temperatures is stronger over weekends as compared to week days. III. BACKGROUND Russia is an upper-middle income economy with the largest territory in the world and a population over 143 million. After the collapse of the Soviet Union, in 1991, Russia had faced the transitory period till 2000s. This period was characterized by economic reforms and institutional changes, high crime and homicide rates, and high mortality (Kaminski 1996; Schleifer and Treisman 2005).14 Currently, the homicide rate in Russia is remains the highest in Europe and is almost twice as high as in the United States (United Nations Office on Drugs and Crime 2019). Russian regions are homogeneous in terms of official language, legislation, and law enforcement, but are heterogeneous in terms of climate, homicide rates, and socioeconomic conditions. Figure 1 presents the average annual violent mortality rates per million of population in Russian regions in 1989–2015. As shown in this figure, in the northern European part of Russia and in the most Asian part, the homicide rates are higher than in the central and southern European parts. Homicide rates also differ by gender and age. Figure 2 presents the violent mortality rates per million of population across gender and age groups in 1989–2015. As shown in this figure, violent mortality is higher among the prime age population, and the homicide rates are higher among males. The climate of Russia can be predominantly classified as continental in the European part of Russia and subarctic in the Asian part. Summers are warm to hot and winters are very cold in most regions. In the European part of Russia, northern and central regions have mostly dry and 14. To ensure that the transition period does not drive our results, in our robustness checks, we estimate the model for the post-transition period only. The results do not change. 248 ECONOMIC INQUIRY FIGURE 1 Map of Average Annual Violent Mortality Rate in Russia Per Million, 1989–2015 Source: Authors’ construction. Computations are based on data from the Federal Statistical Service of the Russian Federation and the Russian Fertility and Mortality Database (RusFMD 2016). Violent mortality rate is measured per million persons. FIGURE 2 Average Annual Violent Mortality Rate across Gender and Age Groups Per Million—Russia—1989–2015 Source: Authors’ computations based on data from the Federal Statistical Service of the Russian Federation and the Russian Fertility and Mortality Database (RusFMD 2016). Violent mortality rate measured per million persons of the corresponding age group is presented. OTRACHSHENKO, POPOVA & TAVARES: WEATHER AND VIOLENCE IN RUSSIA 249 FIGURE 3 Average Monthly Temperature in June–August in Russia, 1990–2012 Source: Authors’ construction based on data from the Climate Change Knowledge Portal of the World Bank. sunny summers and cold winters with frost and snowfall. In the southern European part, summers are hot and dry, and winters are very cold, except for the areas around the Black Sea, which have mild winters with frequent rainfall. The Siberian part of Russia is known for its extreme weather with very cold winters and hot summers that are short and wet. The coldest place in the central Siberian part is Oymyakon, located in the Sakha Republic, where the winter temperature can be below −55∘C in January. Generally, most regions have even precipitation across seasons, except for East Siberia and the Far East, where winter is dry compared to summer. According to the World Bank the annual average temperature in Russia from 1990–2012 was −5.4∘C. The warmest month is July, with the average monthly temperature 15.1∘C, while the coldest month is January, with −25.2∘C. As shown in Figure 3, the average summer temperature in Russia from 1990–2012 was about 13∘C with an upward trend. IV. DATA We use annual data on violent mortality rates in 79 regions of the Russian Federation for the period from 1989 until 2015 from the Russian Federal State Statistics Service and the Russian Fertility and Mortality Database (RusFMD 2016).15 According to the 15. Our dataset includes all regions of the Russian Federation with the exception of autonomous districts that International Statistical Classification of Diseases and Related Health Problems by the World Health Organization (WHO), violent death is defined as a death from homicide and injury purposely inflicted by other persons, including legal execution.16 The data on average daily temperature and precipitation are collected from 518 meteorological ground stations and are weighted by an inverse distance square from the nearest population settlement within a 200 km radius. The settlements within a region are then weighted based on their population. Ground stations that are closer to settlements with a larger population thereby receive the largest weight. This approach gives us the weather experienced by an average person in a region (Dell, Jones, and Olken 2014; Hanigan, Hall, and Dear 2006).17 Figure 4 shows the distribution of days with a particular mean daily temperature in Russia from are included in larger territorial units, that is, the KhantyMansi Autonomous District—Yugra and the Yamalo-Nenets Autonomous District, which are part of a larger Tyumen oblast, the Nenets Autonomous District, which is a part of the Arkhangelsk oblast. Also, data for the Chechen Republic are not available due to the military conflicts there in the 1990s. 16. The death penalty has been indefinitely suspended and not executed in Russia since 1996. According to archival data, in the period 1991–1996, 163 persons were executed; this is 0.07% of the total violent mortality in Russia during this period. 17. An alternative approach is to use the area-weighted weather data, which gives “the average weather experienced by a place” (Dell, Jones, and Olken 2014, 751). As suggested by Dell, Jones, and Olken (2014), this approach is less preferred for countries with large scarcely populated regions, e.g. the US and Russia. 250 ECONOMIC INQUIRY FIGURE 4 Distribution of Days across Temperature Ranges in Russia, 1989–2015 Source: Authors’ computations. Notes: The intervals in White, (19∘C, 22∘C] and (22∘C, 25∘C], the most comfortable temperature limits, are used as default. The intervals in Black, (25∘C, 28∘C] and above 28∘C, show the extremely hot temperature. 1989 to 2015. As shown in this figure, the temperature spectrum is divided into 3∘centigrade intervals. For the empirical analysis, these intervals are constructed for each region and each year. Each interval presents the frequency of days with a particular temperature within a region and year. In Figure 4 the white bars stand for the frequency of days with the (19∘C, 22∘C] and (22∘C, 25∘C] temperature ranges, which are the most comfortable temperature limits and used as default. The black bars stand for the frequency of days with the (25∘C, 28∘C] and above 28∘C temperature ranges, showing the extremely hot temperature. Only two thirds of the regions have experienced days above 28∘C, and the average number of days with such temperature is 0.97 per year in our sample. Thus, in our analysis, we combine the days above 25∘C into one interval. On average, a day with the mean temperature above 25∘C has 33.2∘C during the day and 20.1∘C during the night. Thus, we consider the days with above 25∘C as extremely hot. Regarding the extremely cold days, an average day with the mean temperature below −23∘C has −25.6∘C during the day and −36.3∘C during the night.18 18. We also split the temperature bin below −23∘Cinto the [−23∘Cand−26∘C) and below −26∘C temperature bins. The results are similar and are available upon request. The data on mean daily precipitation within a region and a year are divided into terciles: [0 mm, 10 mm), [10 mm, 20 mm), and between 20 mm and above. The precipitation interval [0 mm, 10 mm) is used as a default. In case of both temperature and precipitation, the numbers of days per year is standardized to 365 days. Table 1 presents summary statistics on homicide rates and the number of days with extreme temperature and precipitation. As shown in this table, the average number of days with temperature above 25∘C is 5.13 per region per year in our sample. Overall, the 25th percentile of regions has experienced 0.93 days above 25∘C per year and the 75th percentile of regions has experienced 5.52 such days per year during the period under study. Days below −23∘C are more frequent. There are 9.5 such days on average per year with 0.63 days in the 25th percentile of regions and 12.22 days in the 75th percentile of regions. There are also on average 2.05 days with extreme precipitation. V. METHODOLOGY To examine the impact of weather on violent mortality, we follow the econometric approach suggested by Deschênes and Greenstone (2011), Burgess et al. (2017), Otrachshenko, Popova, and OTRACHSHENKO, POPOVA & TAVARES: WEATHER AND VIOLENCE IN RUSSIA 257 TABLE 5 Impact of Temperature on Total Homicide by Gender Female Male Coef. S.E. Coef. S.E. −23∘C and below −0.01 0.22 −0.51 0.82 −23∘C−20∘C−0.30 0.35 −1.02 0.94 −20∘C−17∘C−0.14 0.37 −0.88 1.38 −17∘C−14∘C 0.02 0.28 0.30 1.10 −14∘C−11∘C−0.47 0.39 −1.55 1.48 −11∘C−8∘C−0.24 0.18 −0.60 0.58 −8∘C−5∘C 0.00 0.22 −0.61 0.85 −5∘C−2∘C−0.21 0.23 −0.62 0.93 −2∘C1 ∘C 0.03 0.16 0.07 0.68 1∘C4 ∘C−0.11 0.18 −0.94 0.80 4∘C7 ∘C−0.23 0.17 −0.55 0.71 7∘C10 ∘C−0.10 0.17 −0.51 0.72 10∘C13 ∘C−0.02 0.15 −0.56 0.58 13∘C16 ∘C 0.04 0.10 0.12 0.34 16∘C19 ∘C−0.07 0.11 −0.09 0.44 Above 25∘C 0.32*** 0.07 0.91*** 0.27 10 mm 20 mm −0.12 0.16 −0.90 0.61 20 mm 100 mm 0.35 0.49 1.40 1.87 Time fixed effects Yes Yes Regional fixed effects Yes Yes Regional linear trends Yes Yes Number of observations 2,120 2,122 Rsq-within 0.75 0.74 Notes: Coef. and S.E. stand for coefficients and robust standard errors that are clustered at a regional level. All regressions are weighted by the corresponding population. The temperature interval (19∘C, 25∘C] and the precipitation interval [0 mm, 10 mm) are used as defaults. ***Significant level at 1%. with temperature below −23∘C on homicide for either gender. D. Years of Life Lost by a Victim According to McCollister, French, and Fang (2010), the total social costs of criminal acts consist of tangible and intangible costs. In the case of murders, tangible costs include victim costs (a present value of life time earnings), criminal system costs (i.e., police protection cost, legal and adjudication costs, and the convicted perpetrators’ correction costs), and crime carrier costs (productivity losses associated with perpetrators of crimes). Intangible costs include corrected risk-of-homicide costs that are willingness to pay to prevent violence. According to McCollister, French, and Fang (2010), the total social costs of one murder in the United States are about 9 million USD in 2008 prices. We compute years of life lost by a victim that are due to the impact of one hot day (above 25∘C). That is, how many years a victim would live if she/he were not murdered. This measure is equivalent to victim costs suggested by McCollister, French, and Fang (2010) and contributes 8.2% to the total social costs (McCollister, French, and Fang 2010). Our estimates should be considered as a lower bound of the total social costs. The results are in Table 8, in which columns 1–3 correspond to the estimated number of deaths of females, males, and both genders based on the impact of 1 day with temperature above 25∘C (hot) from Table 6. Columns 4–6 stand for the years of life lost by a victim of a particular age group. Those columns are based on the statistics of the WHO (2016) on life expectancy of particular age groups. Columns 7–9 stand for the total number years of life lost due to the impact of one hot day. Table 8 shows that the greatest number of victims associated with 1 day above 25∘Cis among adult and mature females and males—10 and 8 female victims, and 23 and 25 male victims, respectively. The greatest total number of years lost is observed among the adult females and males, which is the most economically active and reproductive age group. Overall, we find that the total number of years lost among all age groups is 642 and 1,579 for females and 258 ECONOMIC INQUIRY Table 6 Impact of One Day with Temperatures above 25∘C and below −23∘C by Gender and Age Group Both Gender Female Male Coef. S.E. Coef. S.E. Coef. S.E. Impact of 1 day with temperatures above 25∘C All ages 0.60*** 0.15 0.32*** 0.07 0.91*** 0.27 Young (15–24) 0.44** 0.16 0.06 0.09 0.80*** 0.27 Adult (25–44) 0.76*** 0.23 0.46*** 0.13 1.06** 0.42 Mature (45–59) 1.19*** 0.30 0.55*** 0.16 1.93*** 0.52 Old (60+) 0.48*** 0.16 0.36** 0.13 0.74** 0.31 Impact of 1 day with temperatures below −23∘C All ages −0.29 0.50 −0.01 0.22 −0.51 0.82 Young (15–24) −0.41 0.41 −0.05 0.16 −0.76 0.71 Adult (25–44) −0.47 0.73 −0.13 0.31 −0.69 1.20 Mature (45–59) −0.83 0.94 −0.20 0.44 −1.44 1.54 Old (60+)−0.32 0.47 0.05 0.30 −0.91 0.89 Number of observations 2,120 2,120 2,122 Notes: The estimated coefficients on the above 25∘C bin and the below −23∘C bin for a particular age and gender group are from Equation (1). Coef. and S.E. stand for coefficients and robust standard errors that are clustered at a regional level. Regressions for each gender and age group are estimated separately. Each regression includes all temperature and precipitation bins, regional and year fixed effects, and regional time trends, and is weighted by the corresponding population. Full results are available from the authors upon request. **Significant level at 5%; ***significant level at 1%. Table 7 Impact of a Work Day and a Weekend Day with Temperature above 25∘C by Gender and Age Group Both Gender Female Male Coef. S.E. Coef. S.E. Coef. S.E. Impact of a work day with temperatures above 25∘C All ages 0.75*** 0.18 0.37*** 0.09 1.18*** 0.31 Young (15–24) 0.74*** 0.21 0.02 0.13 1.41*** 0.37 Adult (25–44) 0.94*** 0.29 0.53*** 0.18 1.37** 0.53 Mature (45–59) 1.34*** 0.35 0.59** 0.23 2.22*** 0.58 Old (60+) 0.60** 0.21 0.46** 0.18 0.91** 0.41 Impact of a weekend day with temperatures above 25∘C All ages 1.48** 1.55 0.89*** 0.25 2.13** 0.93 Young (15–24) 0.69 0.45 0.31 0.23 1.02 0.79 Adult (25–44) 1.82** 0.83 1.40*** 0.39 2.20 1.43 Mature (45–59) 3.50*** 1.02 1.66*** 0.61 5.60*** 1.71 Old (60+) 1.04** 0.49 0.63 0.40 1.91* 1.04 Number of observations 2,120 2,120 2,122 Notes: The estimated coefficients on the above 25∘C bin for a particular age and gender group are from Equation (1). Coef. and S.E. stand for coefficients and robust standard errors that are clustered at a regional level. Regressions for each gender and age group are estimated separately. Each regression includes all temperature and precipitation bins, regional and year fixed effects, and regional time trends, and is weighted by the corresponding population. Full results are available from the authors upon request. *Significant level at 10%; **significant level at 5%; ***significant level at 1%. males, respectively. We also compute the average number of years lost per victim. According to our results, if not killed, a female victim would live an additional 26.75 years, while a male victim would live 25.06 years more. Thus, even though there are more males than females among the victims at all age groups, except for the elderly, females have a greater cost in terms of years lost per se. VII. CONCLUSION The importance of climate change is hard to exaggerate. However, precise estimates of the social consequences of high temperatures are rare. We rely on heretofore unused data from a novel Russian dataset covering three decades of information on temperatures and violent mortality to estimate the impact of hot OTRACHSHENKO, POPOVA & TAVARES: WEATHER AND VIOLENCE IN RUSSIA 259 TABLE 8 Estimated Number of Victims of One Day with Temperatures above 25∘C by Gender and Age Group (1) (2) (3) (4) (5) (6) (7) (8) (9) Estimated Number of Deaths Years of Life Lost Person-Years of Life Lost Age Groups Female Male Both Gender Female Male Both Gender Female Male Both Gender Young (15–24) 1* 9 9 57.2 45.8 51.5 57.2* 412.20 463.50 Adult (25–44) 10 23 33 38.6 29.2 33.9 386 672 1,118.70 Mature (45–59) 8 25 33 25.3 18.2 21.8 202 455.00 718 Old (above 60) 6 6 12 17.2 12.4 14.8 54 40.18 177.60 Total 24 63 87 642 1,579 2,478 Years of life lost per death 26.75 25.06 28.48 Notes: * is based on nonsignificant coefficient. Columns 1–3 are computed by multiplying the estimated impact of 1 day above 25∘C from Table 6 by the average regional population of each gender and age group during the 1989–2015 period. Columns 4–6 stand for the number of years of life lost by each gender and age group computed as a difference between life expectancy and the upper age limit of a particular age group. Columns 7–9 are computed by multiplying columns 1–3 and 4–6. and cold temperatures on violence. We show that the temperature–violence relationship in Russia is J-shaped, that is, extremely hot temperatures increase violent mortality, while extremely cold temperatures have no effect. In contrast, the nonviolent mortality is affected by most temperature ranges. This finding suggests that the mechanisms behind the temperature–violence relationship go beyond the biological explanations that may drive the effect of temperature on nonviolent mortality and may include economic and social factors too. We also uncover that most victimization occurs on weekends for individuals aged between 45 and 59 years old. Males are more often victimized, especially men between 45 and 59 years old, but females are significantly more victimized on weekends. The consistent relevance of economic conditions, including unemployment and alcohol consumption, as intermediating factors for the impact of weather on violence suggests that we are capturing an important social mechanism that mediates how temperature translates into violence. Our findings suggest that higher unemployment and more widespread vodka consumption may increase the impact of extreme temperatures on violence. However, it is important to note that socioeconomic conditions are only possible mediating channels since they are themselves affected by the weather conditions. Overall, our results imply that the intensity and type of social interactions, which vary across age and gender, and between work days and weekdays, are important factors determining the social cost of temperature shocks. The importance of these results cannot be overstated, especially as the effects of climate change may become more acute. Our findings might be interesting to policy makers in other regions and countries. First, we find that victimization due to both hot and cold temperature shocks might be mitigated by improving job opportunities in a region. Also, regulating spirits consumption may help to mitigate the harmful impact of weather shocks, though only cold ones. The increasing relevance of climate change, and the vulnerability of developing countries to its effects, calls for further work on the social determinants of victimhood and the gender and age inequalities it generates. REFERENCES Aizer, A. “The Gender Wage Gap and Domestic Violence.” American Economic Review, 100(4), 2010, 1847–59. https://doi.org/10.1257/aer.100.4.1847. Anderson, C. A. “Temperature and Aggression: Ubiquitous Effects of Heat on Occurrence of Human Violence.” Psychological Bulletin, 106(1), 1989, 74–96. https://doi .org/10.1037/0033-2909.106.1.74. Anderson, C. A., K. B. Anderson, N. Dorr, K. M. DeNeve, and M. Flanagan. “Temperature and Aggression.” Advances in Experimental Social Psychology, 32, 2000, 63–133. https://doi.org/10.1016/b978-012278805-5/50011-0. Anderson, C. A. “Temperature and Aggression: Effects on Quarterly, Yearly, and City Rates of Violent and Nonviolent Crime.” Journal of Personality and Social Psychology, 52(6), 1987, 1161–73. https://doi.org/10.1037/ 0022-3514.52.6.1161. Annan-Phan, S., and B. A. Ba. “Hot Temperatures, Aggression, and Death at the Hands of the Police: Evidence from the U.S.” SSRN Electronic Journal, 2019. https:// doi.org/10.2139/ssrn.3356045. Auliciems, A., and L. DiBartolo. “Domestic Violence in a Subtropical Environment: Police Calls and Weather in Brisbane.” International Journal of Biometeorology, 39(1), 1995, 34–9. https://doi.org/10.1007/ BF01320891. Baron, R. A., and P. A. Bell. “Aggression and Heat: The Influence of Ambient Temperature, Negative Affect, and a Cooling Drink on Physical Aggression.” Journal of Personality and Social Psychology, 33(3), 1976, 245–55. https://doi.org/10.1037/0022-3514.33.3.245. Becker, G. S. “Crime and Punishment: An Economic Approach.” Journal of Political Economy, 76(2), 1968, 169–217. https://doi.org/10.1086/259394. 260 ECONOMIC INQUIRY Blakeslee, D. S., and R. Fishman. “Weather Shocks, Agriculture, and Crime: Evidence from India.” Journal of Human Resources, 53(3), 2018, 750–82. https://doi .org/10.3368/jhr.53.3.0715-7234R1. Bobonis, G. J., M. González-Brenes, and R. Castro. “Public Transfers and Domestic Violence: The Roles of Private Information and Spousal Control.” American Economic Journal: Economic Policy, 5(1), 2013, 179–205. https:// doi.org/10.1257/pol.5.1.179. Borker, G. “Safety First: Perceived Risk of Street Harassment and Educational Choices of Women.” Job Market Paper, 2017. Accessed March 19, 2020. https://girijaborker .files.wordpress.com/2017/11/borker_jmp.pdf. Bruckner, M., and A. Ciccone. “Rain and the Democratic Window of Opportunity.” Econometrica, 79(3), 2011, 923–47. https://doi.org/10.3982/ECTA8183. Brunsdon, C., J. Corcoran, G. Higgs, and A. Ware. “The Influence of Weather on Local Geographical Patterns of Police Calls for Service.” Environment and Planning B: Planning and Design, 36(5), 2009, 906–26. https://doi .org/10.1068/b32133. Burgess, R., O. Deschenes, D. Donaldson, and M. Greenstone. “Weather, Climate Change and Death in India.” Mimeo, 2017. Burke, M. B., E. Miguel, S. Satyanath, J. A. Dykema, and D. B. Lobell. “Warming Increases the Risk of Civil War in Africa.” Proceedings of the National Academy of Sciences of the United States of America, 106(49), 2009, 20670–4. https://doi.org/10.1073/pnas.0907998106. Burke, M., S. Hsiang, and E. Miguel. “Climate and Conflict.” Annual Review of Economics, 7(1), 2015, 577–617. https://doi.org/10.1146/annurev-economics-080614115430. Burke, P. J., and A. Leigh. “Do Output Contractions Trigger Democratic Change?” American Economic Journal: Macroeconomics, 2(4), 2010, 124–57. https://doi.org/ 10.1257/mac.2.4.124. Card, D., and G. B. Dahl. “Family Violence and Football: The Effect of Unexpected Emotional Cues on Violent Behavior.” Quarterly Journal of Economics, 126(1), 2011, 103–43. https://doi.org/10.1093/qje/qjr001. Carlsmith, J. M., and C. A. Anderson. “Ambient Temperature and the Occurrence of Collective Violence: A New Analysis.” Journal of Personality and Social Psycology, 37(3), 1979, 337–44. https://doi.org/10.1037/ 0022-3514.37.3.337. Carpenter, C., and C. Dobkin. “The Minimum Legal Drinking Age and Public Health.” Journal of Economic Perspectives, 25(2), 2011, 133–56. https://doi.org/10.1257/jep .25.2.133. Cerqueira, D., and R. R. Soares. “The Welfare Cost of Homicides in Brazil: Accounting for Heterogeneity in the Willingness to Pay for Mortality Reductions.” Health Economics, 25(3), 2016, 259–76. https://doi .org/10.1002/hec.3137. Chalfin, A., S. Danagoulian, and M. Deza. “More Sneezing, Less Crime? Health Shocks and the Market for Offenses.” Journal of Health Economics, 68, 2019, 102230. https://doi.org/10.1016/j.jhealeco.2019 .102230. Cohn, E. G. “Weather and Crime.” British Journal of Criminology, 30(1), 1990, 51–64. https://doi.org/10.1093/ oxfordjournals.bjc.a047980. . “The Prediction of Police Calls for Service: The Influence of Weather and Temporal Variables on Rape and Domestic Violence.” Journal of Environmental Psychology, 13(1), 1993, 71–83. https://doi.org/10.1016/ S0272-4944(05)80216-6. Dell, M., B. F. Jones, and B. A. Olken. “Temperature Shocks and Economic Growth: Evidence from the Last Half Century.” American Economic Journal: Macroeconomics, 4(3), 2012, 66–95. https://doi.org/10.1257/ mac.4.3.66. . “What Do We Learn from the Weather ? The New Climate–Economy Literature.” Journal of Economic Literature, 52(3), 2014, 740–98. https://doi.org/ 10.3386/w19578. Deschênes, O., and M. Greenstone. “Climate Change, Mortality, and Adaptation: Evidence from Annual Fluctuations in Weather in the US.” American Economic Journal: Applied Economics, 3(4), 2011, 152–85. https://doi .org/10.1257/app.3.4.152. Deschênes, O., and E. Moretti. “Extreme Weather Events, Mortality, and Migration.” The Review of Economics and Statistics, 91(4), 2009, 659–81. https://doi.org/10 .1162/rest.91.4.659. Gantz, W., S. D. Bradley, and Z. Wang. “Televised NFL Games, the Family, and Domestic Violence,” in Handbook of Sports and Media, edited by A. A. Raney and J. Bryant. Mahwah, NJ: Erlbaum, 2006, 365–82. Garcia-Moreno, C., H. Jansen, M. Ellsberg, L. Heise, and C. Watts. “Prevalence of Intimate Partner Violence: Findings from the WHO Multi-Country Study on Women’s Health and Domestic Violence.” Lancet, 368(9543), 2006, 1260–9. https://doi.org/10.1016/ S0140-6736(06)69523-8. Graff Zivin, J., and M. J. Neidell. “Temperature and the Allocation of Time: Implications for Climate Change.” Journal of Labor Economics, 32(1), 2014, 1–26. https:// doi.org/10.1086/671766. Graff Zivin, J., S. M. Hsiang, and M. Neidell. “Temperature and Human Capital in the Short and Long Run.” Journal of the Association of Environmental and Resource Economists, 5(1), 2018, 77–105. https://doi.org/10 .1086/694177. Hanigan, I., G. Hall, and K. B. G. Dear. “A Comparison of Methods for Calculating Population Exposure Estimates of Daily Weather for Health Research.” International Journal of Health Geographics, 5(1), 2006, 38. https://doi.org/10.1186/1476-072X-5-38. Heyes, A., and S. Saberian. “Temperature and Decisions: Evidence from 207,000 Court Cases.” American Economic Journal: Applied Economics, 11(2), 2019, 238–65. https://doi.org/10.1257/APP.20170223. Hidrobo, M., and L. Fernald. “Cash Transfers and Domestic Violence.” Journal of Health Economics, 32(1), 2013, 304–19. https://doi.org/10.1016/j.jhealeco.2012 .11.002. Hindin, M. J., S. Kishor, and D. L. Ansara. “Intimate Partner Violence among Couples in 10 DHS Countries: Predictors and Health Outcomes.” DHS Analytical Studies, (18), 2008. https://dhsprogram.com/pubs/pdf/AS18/ AS18.pdf. Hsiang, S. M., M. Burke, and E. Miguel. “Quantifying the Influence of Climate on Human Conflict.” Science, 341(6151), 2013, 1235367. https://doi.org/10.1126/ science.1235367. Hsiang, S. M., K. C. Meng, and M. A. Cane. “Civil Conflicts Are Associated with the Global Climate.” Nature, 476(7361), 2011, 438–41. https://doi.org/10.1038/ nature10311. Jacob, B., L. Lefgren, and E. Moretti. “The Dynamics of Criminal Behavior: Evidence from Weather Shocks.” The Journal of Human Resources, 42(3), 2007, 489–527. https://doi.org/10.2307/40057316. Kaminski, B., ed.. Economic Transition in Russia and the New States of Eurasia. M.E. Sharpe, 1996. Karlsson, M., and N. R. Ziebarth. “Population Health Effects and Health-Related Costs of Extreme Temperatures: Comprehensive Evidence from Germany.” Journal of OTRACHSHENKO, POPOVA & TAVARES: WEATHER AND VIOLENCE IN RUSSIA 261 Environmental Economics and Management, 91, 2018, 93–117. https://doi.org/10.1016/j.jeem.2018.06.004. Kemper, T. D. Social Structure and Testosterone.New Brunswick, NJ: Rutgers University Press, 1990. Kenrick, D. T., and S. W. MacFarlane. “Ambient Temperature and Horn Honking: A Field Study of the Heat/Aggression Relationship.” Environment and Behavior, 18(2), 1986, 179–91. https://doi.org/10 .1177/0013916586182002. Larrick, R. P., T. A. Timmerman, A. M. Carton, and J. Abrevaya. “Temper, Temperature, and Temptation: HeatRelated Retaliation in Baseball.” Psychological Science: A Journal of the American Psychological Society/APS, 22(4), 2011, 423–8. https://doi.org/10.1177/ 0956797611399292. LeBeau, J. L., and W. T. Corcoran. “Changes in Calls for Police Service with Changes in Routine Activities and the Arrival and Passage of Weather Fronts.” Journal of Quantitative Criminology, 6(3), 1990, 269–91. https:// doi.org/10.1007/BF01065411. Lehren, A. W., and A. Baker. “The Dark Side Of Summer: More Killings.” The New York Times, 2009. Accessed April 25, 2017. http://www.nytimes.com/2009/06/19/ nyregion/19murder.html. Levitt, S. D. “The Limited Role of Changing Age Structure in Explaining Aggregate Crime Rates.” Criminology, 37(3), 1999, 581–98. https://doi.org/10.1111/j .1745-9125.1999.tb00497.x. Liang, S., H. Leng, Q. Yuan, B. W. Wang, and C. Yuan. “How Does Weather and Climate Affect Pedestrian Walking Speed during Cool and Cold Seasons in Severely Cold Areas?” Building and Environment, 175, 2020, 106811. https://doi.org/10.1016/j.buildenv.2020.106811. Lindo, J., C. Myers, A. Schlosser, and S. Cunningham. “How Far Is Too Far?: New Evidence on Abortion Clinic Closures, Access, and Abortions.” Journal of Human Resources, 2019, 1217. https://doi.org/10.3368/jhr.55.4 .1217-9254R3. Luca, D. L., E. Owens, and G. Sharma. “Can Alcohol Prohibition Reduce Violence against Women?” American Economic Review, 105(5), 2015, 625–9. https://doi.org/ 10.1257/aer.p20151120. Martens, W. J. M. “Climate Change, Thermal Stress and Mortality Changes.” Social Science and Medicine, 46(3), 1998, 331–44. https://doi.org/10.1016/S02779536(97)00162-7. McCollister, K. E., M. T. French, and H. Fang. “The Cost of Crime to Society: New Crime-Specific Estimates for Policy and Program Evaluation.” Drug and Alcohol Dependence, 108(1–2), 2010, 98–109. https://doi.org/ 10.1016/j.drugalcdep.2009.12.002. Miguel, E. “Poverty and Witch Killing.” Review of Economic Studies, 72(4), 2005, 1153–72. https://doi.org/10.1111/ 0034-6527.00365. Myers, R. D. “Neurochemical Mechanisms of Temperature Regulation and Food Ingestion,” in Neurahumoral Coding of Brain Function, edited by R. D. Myers and R. R. Drucker-Colin. New York: Plenum Press, 1974, 99–114. Otrachshenko, V., O. Popova, and P. Solomin. “Health Consequences of the Russian Weather.” Ecological Economics, 132, 2017, 290–306. https://doi.org/10.1016/j .ecolecon.2016.10.021. . “Misfortunes Never Come Singly: Consecutive Weather Shocks and Mortality in Russia.” Economics and Human Biology, 31, 2018, 249–58. https://doi.org/ 10.1016/j.ehb.2018.08.008. Pakiam, J. E. “The Effects of Weather on Crime.” International Journal of Comparative and Applied Criminal Justice, 5(2), 1981, 179–86. https://doi.org/10.1080/ 01924036.1981.9688728. Park, J., J. Goodman, M. Hurwitz, and J. Smith. “Heat and Learning.” American Economic Journal: Economic Policy, 12(2), 2020, 306–39. https://doi.org/10.1257/pol .20180612. Park, R. J. “Hot Temperature and High Stakes Performance.” Journal of Human Resources, 2020. https://doi.org/10 .3368/jhr.57.2.0618-9535r3. Ranson, M. “Crime, Weather, and Climate Change.” Journal of Environmental Economics and Management, 67(3), 2014, 274–302. https://doi.org/10.1016/j.jeem.2013.11 .008. Reifman, A. S., R. P. Larrick, and S. Fein. “Temper and Temperature on the Diamond: The Heat-Aggression Relationship in Major League Baseball.” Personality and Social Psychology Bulletin, 17(5), 1991, 580–5. https://doi.org/10.1177/0146167291175013. Reis, D. J. “The Chemical Coding of Aggression in Brain.” in Neurahumoral Coding of Brain Function, edited by R. D. Myers and R. R. Drucker-Colin. New York: Plenum Press, 1974, 125–50. Rotton, J., and E. G. Cohn. “Outdoor Temperature, Climate Control, and Criminal Assault: The Spatial and Temporal Ecology of Violence.” Environment and Behavior, 36(2), 2004, 276–306. https://doi.org/10 .1177/0013916503259515. Rotton, J., and J. Frey. “Air Pollution, Weather, and Violent Crimes: Concomitant Time-Series Analysis of Archival Data.” Journal of Personality and Social Psychology, 49(5), 1985, 1207–20. https://doi.org/10.1037/00223514.49.5.1207. Rotton, J., M. Shats, and R. Standers. “Temperature and Pedestrian Tempo.” Environment and Behavior, 22(5), 1990, 650–74. https://doi.org/10.1177/ 0013916590225005. Russian Fertility and Mortality Database. Russian Fertility and Mortality Database. Moscow, Russia: Center for Demographic Research, 2016, Accessed January 15, 2017. http://demogr.nes.ru/index.php/ru/demogr_ indicat/data. Ryan, M. E. “The Heat: Temperature, Police Behavior and the Enforcement of Law.” European Journal of Law and Economics, 49(2), 2020, 187–203. https://doi.org/ 10.1007/s10657-020-09646-6. Schleifer, A., and D. Treisman. “A Normal Country: Russia after Communism.” Journal of Economic Perspectives, 19(1), 2005, 151–74. https://doi.org/10.1257/ 0895330053147949. Schmidheiny, K., and S. Siegloch. “On Event Studies and Distributed-Lags in Two-Way Fixed Effects Models: Identification, Equivalence, and Generalization.” ZEW Discussion Paper 20-017. 2020. Accessed April 10, 2020. http://ftp.zew.de/pub/zew-docs/dp/dp20017.pdf. Simister, J., and C. Cooper. “Thermal Stress in the U.S.A.: Effects on Violence and on Employee Behaviour.” Stress and Health, 21(1), 2005, 3–15. https://doi.org/10 .1002/smi.1029. Soares, R. R. “The Welfare Cost of Violence across Countries.” Journal of Health Economics, 25(5), 2006, 821–46. https://doi.org/10.1016/j.jhealeco.2005.10 .007. Soares, R. R., and J. Naritomi. “Understanding High Crime Rates in Latin America: The Role of Social and Policy Factors,” in The Economics of Crime: Lessons for and from Latin America, edited by R. Di Tella, S. Edwards and E. Schargrodsky. Chicago: University of Chicago Press, 2010, 19–55. U.S. Riot Commission. Report of the National Advisory Commission on Civil Disorders. New York: 1968. United Nations. “Integration of the Human Rights of Women and a Gender Perspective: Violence against Women,” 2006. Accessed April 5, 2017. https://documents- 262 ECONOMIC INQUIRY dds-ny.un.org/doc/UNDOC/GEN/G06/104/47/PDF/ G0610447.pdf?OpenElement. United Nations Office on Drugs and Crime. “Global Study on Homicide. Homicide, Development and the Sustainable Development Goals.” Unodoc, 2019. Accessed January 20, 2020. https://www.unodc.org/unodc/en/dataand-analysis/global-study-on-homicide.html. Volkova, O., T. Lipai, and S. Wendt. “Domestic Violence in Rural Areas of Russia and Australia.” International Letters of Social and Humanistic Sciences, 55, 2015, 101–10. https://doi.org/10.18052/www.scipress.com/ ILSHS.55.101. Vrij, A., J. Van Der Steen, and L. Koppelaar. “Aggression of Police Officers as a Function of Temperature: An Experiment with the Fire Arms Training System.” Journal of Community & Applied Social Psychology, 4(5), 1994, 365–70. https://doi.org/10.1002/casp.2450040505. World Health Organization. “Global Health Observatory Data Repository.” 2016. Accessed April 14, 2017. http://apps .who.int/gho/data/?theme=main&vid=61360. SUPPORTING INFORMATION Additional supporting information may be found online in the Supporting Information section at the end of the article. Data S1 Supporting information