Use your study guide, lectures, notes, and other online resources (Google is always your friend if used wisely) to answer the following question. Y

Abstract The heat hypothesis states

that hot temperatures can in- crease aggressive motives and behaviors. Although alterna- tive explanations occasionally account for some portion of the observed increases in aggres- sion when temperatures are high, none are sufficient to ac- count for most such heat ef- fects. Hot temperatures in- crease aggression by directly increasing feelings of hostility and indirectly increasing ag- gressive thoughts. Results show that global warming trends may well increase violent-crime rates. Better climate controls in many institutional settings (e.g., prisons, schools, the workplace) may reduce aggression-related problems in those settings.

Keywords temperature; aggression; vio- lence; global warming

I pray thee, good Mercutio, let’s retire; The day is hot, the Capulets abroad, And, if we meet, we shall not ’scape a

brawl,

For now, these hot days, is the mad blood stirring.

—William Shakespeare, Romeo and Juliet, Act 3, Scene 1

Does excessive heat increase violence? Social commentators have long noted effects of weather on human behavior and have used heat-related imagery in their works (e.g., Cicero, 106–32 B.C.; Siouxsie and the Banshees, in their song “92°,” 1986). Empirical methods were first applied to this theory in the middle 1700s. Montesquieu (1748/1989) noted that “you will find in the northern climates peoples who have few vices, enough virtues, and much sincerity and frankness. As you move to- ward the countries of the south, you will believe you have moved away from morality itself: the live- liest passions will increase crime . . . “(p. 234). In the late 1800s and early 1900s, a number of Euro- pean and North American scholars found that rates of violent crime in- creased during the hottest times of the year, and were higher in re- gions with hotter climates (Ander- son, 1989). Perhaps Shakespeare was right.

The heat hypothesis states that hot temperatures increase aggressive motivation and (under some condi- tions) aggressive behavior. The heat effect is the observation of higher rates of aggression by people who are hot relative to people who are cooler. Methodological difficulties and the lack of modern statistical analyses in early studies made cau- sal statements risky, but causal is- sues are crucial. For example, more assaults occur during the summer months than during other months, but this could be a spurious artifact of differences in the daily activities people perform at different times of the year. Perhaps people are out- side more during the summer, in- creasing the opportunity for con- flicts. Routine activities associated with summer may increase assault rates, and heat-induced discomfort may play no direct causal role in this increase. Such mediated, or in- direct, heat effects are important in their own right, of course.

MODERN STUDIES OF THE HEAT HYPOTHESIS

Modern studies (i.e., post-1950) address these methodological is- sues in several ways. The research can be classed into three broad cat- egories: (a) field studies, all of which focus on some form of ag- gression; (b) laboratory studies with a focus on aggression; and (c)

Heat and Violence Craig A. Anderson1

Department of Psychology, Iowa State University, Ames, Iowa

33CURRENT DIRECTIONS IN PSYCHOLOGICAL SCIENCE

Copyright © 2001 American Psychological Society

laboratory studies with a focus on aggression-related variables, such as hostile feelings, beliefs, and arousal.

The results can be characterized with four summary statements. First, periodic claims that observed heat effects result solely from arti- factual processes have, to date, proven false. Second, the ongoing search for conditions under which excessive heat may cause a decline in aggression has largely failed. Third, there has been a growing re- alization that other aggression- related processes sometimes ob- scure, exaggerate, or modify the heat effect. Fourth, a simple ver- sion of the heat hypothesis (e.g., Berkowitz, 1993)—that people get cranky when uncomfortable—has proven surprisingly robust to all challenges. In short, excessive heat appears to cause increases in ag- gression in many settings.

In investigating the relation be- tween heat and aggression, my stu- dents and I have relied on a well- worn philosophical approach know as triangulation. This in- volves examining competing ex- planations of the heat effect from multiple perspectives. Because the weaknesses of one particular meth- odology differ from those of other methodologies, an explanation of observed heat effects that works across different methodologies is less likely to be invalid than expla- nations that work only for one or two methods. For example, changes in routine activities may be able to explain summer in- creases in violent crime, but cannot account for the finding that base- ball pitchers are more likely to hit batters with a pitched ball on hot days than on cool days (Reifman, Larrick, & Fein, 1991). The parsi- monious explanation is that heat- induced discomfort increases ag- gressive inclinations on the baseball field and in other natural- istic settings.

Field Studies of Heat and Aggressive Behavior

Field studies may be categorized according to whether they compare aggression rates (usually violent- crime rates) across geographic re- gions that are similar in many re- spects but differ in climate, or whether they compare aggression rates in one geographic region but across time periods that differ in temperature.

Studies Comparing Geographic Regions Data consistently show that vio-

lent-crime rates are higher in the South than in other regions of the United States. Similar patterns ap- peared in the older European stud- ies (Anderson, 1989).

The heat hypothesis is only one of several explanations of the U.S. version of the hot-region effect. One alternative explanation is that, for some reason, a culture of vio- lence (e.g., Nisbett, 1993) devel- oped in the U.S. South, and that this cultural difference has been passed on to present-day inhabit- ants. Reasons given for this cul- tural development differ among scholars; analyses of who settled the South, the institution of slav- ery, and the effects of being a fron- tier or a herding economy have all been offered. Nonetheless, claims that Southern culture accounts for the observed high violent-crime rate in hotter regions of the United States are contradicted by recent analyses of violent-crime rates in 260 U.S. cities (Anderson, Ander- son, Dorr, DeNeve, & Flanagan, 2000). Latent variable statistical techniques were used to estimate the effect of temperature on vio- lent-crime rate, while statistically controlling for the Southernness, population size, and socioeco- nomic status of the cities. This same analysis estimated the effect of Southernness on violent-crime rate, while statistically controlling

for the temperature, population size, and socioeconomic status of the cities. As shown by the positive path coefficient for the link be- tween temperature and violent crime in Figure 1 (.43), temperature was significantly and positively re- lated to violent-crime rate. That is, hotter cities were more violent than cooler cities even after city-to-city differences in Southernness, popu- lation size, and socioeconomic sta- tus were statistically controlled. However, the path coefficient for the link between Southernness and violent crime (.14) was not reliably different from zero (no effect), cast- ing further doubt on the claim that a Southern culture of violence is the sole or primary cause of higher vio- lent-crime rates in hotter U.S. cities.

Studies Comparing Time Periods Field studies comparing aggres-

sion rates in hotter versus cooler time periods also support the heat hypothesis. For example, there are about 2.6% more murders and as- saults in the United States during the summer than other seasons of the year; hot summers produce a bigger increase in violence than cooler summers; and violence rates are higher in hotter years than in cooler years even when various statistical controls are used (Ander- son et al., 2000). Other time-period studies provide consistent results. Aggression—as measured by as- sault rates, spontaneous riots, spouse batterings, and batters be- ing hit by pitched baseballs—is higher during hotter days, months, seasons, and years.

Several studies have examined the heat hypothesis with time peri- ods even shorter than days. Some have found increases in assaults, rapes, and domestic violence at hotter temperatures (Anderson et al., 2000). Studies that have mea- sured temperature at the exact time that aggressive behaviors occurred

34 VOLUME 10, NUMBER 1, FEBRUARY 2001

Published by Blackwell Publishers Inc.

have also yielded the standard heat effect. Kenrick and MacFarlane’s (1984) classic study in Phoenix, Arizona, found that aggressive horn honking increased at hotter temperatures, but only for drivers without air-conditioned cars. More recently, Vrij, van der Steen, and Koppelaar (1994) conducted a field experiment in which Dutch police officers performed in a simulated burglary scenario under hot or comfortable conditions. Hot offic- ers reported more aggressive and threatening impressions of the sus- pect, and were more likely to draw their weapon and shoot the suspect (with laser training weapons), rela- tive to officers in the cool condi- tion.

Summary of Field Studies Field studies consistently find

positive associations between un- comfortable heat and aggression. Most field studies are correlational, so causal interpretation must be

tempered by the possibility that unknown extraneous variables caused a spurious relation between heat and aggression. However, the two major challenges to the heat hypothesis—changes in routine ac- tivities and Southern culture—do not fare well from a broad perspec- tive. Each can account for a few findings, but neither can account for the broad array of heat effects. The consistency of findings across many settings and methods pro- vides strong support for the causal version of the heat hypothesis, even from correlational studies. Furthermore, the few experimental and quasi-experimental field stud- ies lend considerable support to the causal interpretation.

Laboratory Studies and Aggressive Behavior

Mixed Results Lab studies of the heat hypoth-

esis have yielded somewhat mixed

results. The negative-affect escape model (Anderson, 1989; Anderson et al., 2000; Baron & Bell, 1976) pos- tulates that excessive heat increases aggression when the total amount of negative affect a person experi- ences is in the low to moderate range (the fight response), but that excessive heat decreases aggres- sion when total negative affect gets too high (the flight response). In other words, if other aspects of a particular situation (such as being insulted) also produce negative af- fect, then further increases in nega- tive affect caused by hot tempera- ture should (according to this model) lead to escape behavior in- stead of aggressive behavior. In brief, hot temperatures should pro- duce a decline in aggression in situations that have other negative- affect-producing factors present. A meta-analysis (i.e., an analysis combining results across all rel- evant studies) yielded some sup- port for the standard heat effect (hot temperatures increased ag- gression) in lab settings that had few extraneous negative-affect- producing factors present. How- ever, there was little support for the predicted decrease in aggres- sion when extraneous negative- affect-producing factors were present (Anderson et al., 2000). Many early lab studies, especially those that used kerosene heaters, suffered from potential suspicion problems. That is, some partici- pants in those studies may have be- come suspicious about the “true” purpose of the study, and may therefore have behaved in an arti- ficial way. Boyanowsky (1999) re- cently discussed other method- ological problems with early lab studies and provided experimental evidence that when people’s atten- tion is not focused on temperature (as in most naturalistic settings), hot temperatures increase aggres- sion even when additional nega- tive-affect-producing factors (such as being insulted) are present.

Fig. 1. Latent variable model of effects of temperature and Southernness on violent crime, controlling for population and socioeconomic status (SES). Positive path co- efficients (e.g., the .43 above the line connecting “Temperature” to “Violent Crime”) indicate a positive relation between the variables linked by that path. Negative path coefficients would indicate a negative relation between the linked variables. A path coefficient of zero would indicate that the two variables are totally unrelated. Solid lines linking two variables indicate that the associated path coefficient is reliably different from zero (i.e., is statistically significant). The dashed line indicates that the link between the two variables is not reliably different from zero. Adapted from Anderson, Anderson, Dorr, DeNeve, and Flanagan (2000).

35CURRENT DIRECTIONS IN PSYCHOLOGICAL SCIENCE

Copyright © 2001 American Psychological Society

One recent experiment using more subtle ways to manipulate temperature than kerosene heaters succeeded in creating conditions that yielded both heat-induced in- creases and decreases in aggres- sion. Two factors were involved: an ambiguous provocation, followed by multiple opportunities to retali- ate. Under these conditions, my colleagues and I found an initial heat-induced increase in aggres- sive retaliation, followed by a de- crease (Anderson et al., 2000). One explanation of this pattern involves two separate processes. The initial outburst of aggression may have been the result of heat-induced in- creases in aggressive inclinations (via hostile affect and cognition). The later decrease may have been the result of a social justice norm; hot participants may have decided that the initial retaliation was suf- ficient. Of course, in most natural settings, the initial outburst will it- self provoke an aggressive re- sponse from the victim, initiating an escalating cycle of retaliatory aggression.

Summary of Laboratory Studies of Aggressive Behavior In affectively neutral and posi-

tive circumstances, hot tempera- tures cause increases in aggression. Recent lab studies show that even in affectively negative circum- stances, heat causes increases in initial retaliatory aggression.

Laboratory Studies and Aggression-Related Variables

Heat effects on affective, cogni- tive, and arousal variables have proven quite consistent. Exposure to hot temperatures increases heart rate, endorsement of aggressive at- titudes and beliefs, and feelings of hostility, all the while decreasing feelings of arousal and comfort. The heat-induced increase in en- dorsement of aggressive attitudes

and beliefs looks, at first glance, like a cognitive priming effect, au- tomatically increasing the accessi- bility of aggressive thoughts. How- ever, hot temperatures do not automatically prime aggressive thoughts, at least not in the same way that viewing pictures of guns does (Anderson, Anderson, & Deuser, 1996). Thus, the effects of heat on attitudes and beliefs are in- direct, most likely mediated by more direct effects of heat on hos- tile affect. Uncomfortably warm temperatures also produce biases in the interpretation of observed social interactions. Specifically, heat seems to increase the likeli- hood that ambiguous social inter- actions will be interpreted as hav- i n g a g g r e s s i v e c o m p o n e n t s (Anderson et al., 2000). Finally, heat stress decreases performance on many cognitive tasks.

PSYCHOLOGICAL PROCESSES UNDERLYING

THE HEAT EFFECT

Numerous fascinating psycho- logical processes might be in- volved in the typical effect of high temperatures on aggression and violence. The simplest and most powerful ones all revolve around the “crankiness” notion. Being un- comfortable colors the way people see things. Minor insults may be perceived as major ones, inviting (even demanding) retaliation. This notion is compatible with several well-established theories in social psychology, including Berkowitz’s (1984) cognitive neo-association theory and Zillmann’s (1983) exci- tation transfer theory. Our own General Affective Aggression Model (GAAM; e.g., Anderson et al., 2000) explicitly incorporates the key aspects of these earlier models, including the crankiness notion.

GAAM also includes social in- teraction processes that play a key role in the genesis of violent behav-

ior. Specifically, GAAM highlights the fact that any social interaction involves at least two people. Fur- thermore, aggression can come about through fairly automatic processes (i.e., impulsively) as well as through careful planning. My colleagues and I believe that most heat-induced increases in aggres- sion, including the most violent be- haviors, result from distortion of the social interaction process in a hostile direction. Heat-induced dis- comfort makes people cranky. It in- creases hostile affect (e.g., feelings of anger), which in turn primes ag- gressive thoughts, attitudes, prepa- ratory behaviors (e.g., fist clench- ing), and behavioral scripts (such as “retaliation” scripts). A minor provocation can quickly escalate, especially if both participants are affectively and cognitively primed for hostility by their heightened level of discomfort. A mild insult is more likely to provoke a severe in- sult in response when people are hot than when they are more com- fortable. This may lead to further increases in the aggressiveness of responses and counterresponses. An accidental bump in a hot and crowded bar can lead to the trad- ing of insults, punches, and (even- tually) bullets.

NEW RESEARCH DIRECTIONS

Many of the basic pieces of this puzzle have been found, but sev- eral are still missing. Though re- search on the heat hypothesis has been carried out for many years, my colleagues and I believe that the hardest work lies ahead and that the missing pieces are likely to be found in future laboratory stud- ies. Additional work is needed to answer the following key questions.

1. Does excessive heat bias percep- tions in ongoing social interac- tions?

36 VOLUME 10, NUMBER 1, FEBRUARY 2001

Published by Blackwell Publishers Inc.

2. Do people in hot conditions— and who are therefore physi- ologically aroused (i.e., have an increased heart rate) but psy- chologically unaroused (i.e., feel lethargic)—misattribute some of their heat-based arousal to mi- nor provoking social events?

3. Do the cognitive effects of heat stress interfere with normal mechanisms for inhibiting ag- gression?

4. How do escape motives influ- ence the heat effect? The nega- tive-affect escape model speci- fies that escape motives should play a major role. It predicts that under some circumstances, in- creases in heat-induced discom- fort will increase the desire to escape more than the desire to retaliate, and therefore will re- duce aggression if escape is in- compatible with aggression. However, no research has ex- plicitly pitted escape motives against aggressive motives.

5. Do social justice processes un- derlie the finding that excessive heat can at first increase and later decrease aggression?

CURRENT IMPLICATIONS

A broad view of the research— triangulation—suggests that in many settings hot temperatures cause increases in aggression. There are conditions that limit the generality of this conclusion, but the overall pattern of data is im- pressive and convincing.

The implications of this general conclusion are many. Consider the finding that hot years produce in- creases in violent-crime rates. If this heat effect is truly caused by heat-induced increases in aggres- sive motivation, then increased violence can be added to the list of negative social consequences of global warming. Figure 2 illus-

trates just how much of an increase can be expected, based on several estimates of the true relation be- tween temperature and U.S. mur- der and assault rates, at several es- timated levels of global warming. For example, using the best esti- mate of how much the violent- crime rate will increase for each 1 °F increase in temperature (i.e., 4.58), we see that a 2 °F increase in average temperature predicts an increase of about 9 more murders or assaults per 100,000 people, or more than 24,000 additional mur- ders and assaults per year in a population of 270 million.

There are numerous institu- tional settings in which aggression is a problem and in which tempera- ture can be controlled. Schools, prisons, and a wide variety of workplaces are good targets for in- tervention. Research on the effects of better climate control in such set- tings might well show that the additional costs are outweighed by the benefits—better learning, lower incarceration costs, less property damage, and increased productivity.

Recommended Reading

Anderson, C.A., Anderson, K.B., Dorr, N., DeNeve, K.M., & Flana- gan, M. (2000). (See References)

Berkowitz, L. (1993). (See Refer- ences)

Geen, R.G. (1990). Human aggression. Pacific Grove, CA: Brooks Cole.

Reifman, A.S., Larrick, R.P., & Fein, S. (1991). (See References)

Acknowledgments—I thank Kathryn Anderson, Brad Bushman, and Kristina DeNeve for their helpful comments on this article.

Note

1. Address correspondence to Craig A. Anderson, Department of Psychol- ogy, W112 Lagomarcino Hall, Iowa State University, Ames, IA 50011-3180; e-mail: [email protected]

References

Anderson, C.A. (1989). Temperature and aggres- sion: Ubiquitous effects of heat on occurrence of human violence. Psychological Bulletin, 106, 74–96.

Anderson, C.A., Anderson, K.B., & Deuser, W.E. (1996). Examining an affective aggression framework: Weapon and temperature effects on aggressive thoughts, affect, and attitudes.

Fig. 2. Estimated effect of global warming on murders and assaults in the United States. The graph shows the estimated increase in the murder-assault rate and in the number of murders and assaults (for a population of 270 million) based on three estimates of the relation between temperature and violence. From Anderson, Ander- son, Dorr, DeNeve, and Flanagan (2000).

37CURRENT DIRECTIONS IN PSYCHOLOGICAL SCIENCE

Copyright © 2001 American Psychological Society

Personality and Social Psychology Bulletin, 22, 366–376.

Anderson, C.A., Anderson, K.B., Dorr, N., DeNeve, K.M., & Flanagan, M. (2000). Temperature and aggression. In M. Zanna (Ed.), Advances in ex- perimental social psychology (Vol. 32, pp. 63– 133). New York: Academic Press.

Baron, R.A., & Bell, P.A. (1976). Aggression and heat: The influence of ambient temperature, negative affect, and a cooling drink on physi- cal aggression. Journal of Personality and Social Psychology, 33, 245–255.

Berkowitz, L. (1984). Some effects of thoughts on anti- and prosocial influences of media events: A cognitive-neoassociation analysis. Psycho- logical Bulletin, 95, 410–427.

Berkowitz, L. (1993). Aggression: Its causes, conse- quences, and control. New York: McGraw Hill.

Boyanowsky, E. (1999). Violence and aggression in the heat of passion and in cold blood. Interna- tional Journal of Law and Psychiatry, 22, 257–271.

Kenrick, D.T., & MacFarlane, S.W. (1984). Ambient temperature and horn-honking: A field study of the heat/aggression relationship. Environ- ment and Behavior, 18, 179–191.

Montesquieu, C. (1989). The spirit of the laws (A. Cohler, B. Miller, & H. Stone, Trans.). New York: Cambridge University Press. (Original work published 1748)

Nisbett, R.E. (1993). Violence and U.S. regional culture. American Psychologist, 48, 441–449.

Reifman, A.S., Larrick, R.P., & Fein, S. (1991). Tem- per and temperature on the diamond: The heat-aggression relationship in major league baseball. Personality and Social Psychology Bul- letin, 17, 580–585.

Vrij, A., van der Steen, J., & Koppelaar, L. (1994). Aggression of police officers as a function of temperature: An experiment with the Fire Arms Training System. Journal of Community and Applied Social Psychology, 4, 365–370.

Zillmann, D. (1983). Arousal and aggression. In R. Geen & E. Donnerstein (Eds.), Aggression: Theoretical and empirical reviews, Vol. 1. Theoret- ical and methodological issues (pp. 75–101). New York: Academic Press.

38 VOLUME 10, NUMBER 1, FEBRUARY 2001

Published by Blackwell Publishers Inc.