Anthropogenic climate change is increasing average temperatures and the number of days each year with extreme heat. This is likely to affect the way people behave and the economic choices they make: how and when will depend on where in the world they live.
The earth’s average temperature is rising rapidly. The latest Intergovernmental Panel on Climate Change (IPCC) report concludes that the globe is likely to warm by 1.5°C over pre-industrial levels by as soon as 2030, a decade earlier than previously expected. Hot temperature extremes over land could become three to ten times more frequent by the middle of the century, depending on the scenario for global warming.
But harmful extreme temperatures are already occurring. Several coastal sub-tropical areas have already recorded wet-bulb temperatures (a measure that combines temperature and humidity) of 35°C and above (Raymond et al, 2020). Sustained temperatures at or above this mark are physically intolerable and potentially deadly (Sherwood and Huber, 2010).
Increased exposure to high temperatures will affect lives in many other, more subtle ways. This includes the effect of heat on people’s attitudes towards taking risks (their ‘risk aversion’) and how they make decisions with future payoffs (their ‘time preferences’). These economic preferences are fundamental for investment decisions, such as our propensity to save, open new businesses, and invest in our health and children’s education.
Research on this topic is still in its infancy, but results suggest that high temperatures are likely to reduce people’s tolerance of risk and increase their impatience. A likely pathway for these effects is through cognition. It is now well established that extreme heat diminishes cognitive ability, which is important for optimal economic decision-making.
Why are risk and time preferences important?
Our lives are full of big and small decisions. We are continually balancing the benefits of present and future payoffs. For example, by saving, we give up buying things we want today for future purchases. Similarly, we are often faced with risky choices, which may have long-term benefits at short-term costs – for example, investing savings in a risky business now in the hope that it will be a success in months and years to come.
Our willingness to take risks and our level of patience influence our future wellbeing. Economic theory predicts that a certain amount of risk-taking and investing in the future is optimal economic behaviour, so long as we believe that there will be a future. For example, increased risk aversion is estimated to reduce the probability of self-employment, a key indicator of entrepreneurship (Ekelund et al, 2005); and impatient individuals are less likely to take up savings products (Ashraf et al, 2006).
Understanding how temperatures affect these choices is an increasingly important issue.
What might be the effects of high temperatures on economic preferences?
In the short term, temperature changes may affect people’s choices through three potential channels: cognition; sleep; and mood.
There is a significant link between cognition and economic preferences. Studies show that higher scores in mathematics are associated with greater patience when faced with the choice between future and immediate rewards (lower time discount rates). Mathematical skills are also associated with choosing more risky bets during lottery games (lower risk aversion) (Benjamin et al, 2013; Falk et al, 2018).
At the same time, higher temperatures have been shown to deplete cognition significantly. After all, our brain is the most heat-sensitive organ in our body: its temperature can rise by up to 2.5°C after intense physical activity in high temperatures (Kiyatkin, 2007). This could explain why the rate of learning and skill formation in schools decreases with the number of hot days experienced (Park et al, 2021).
Better sleep, the second potential channel, is associated with lower discount rates and less loss aversion, which is when people prefer to avoid losses rather than acquiring gains of equivalent value (Nofsinger and Shank, 2019). This is likely to be related to the role of sleep in the production of insightful and inferential thinking (Deak and Stickgold, 2010).
A study using over seven million night-time sleep records across 68 countries shows that higher night-time temperatures decrease sleep duration by delaying its onset. These effects are substantially larger for residents of lower-income countries (Minor et al, 2020).
An aggressive mood is associated with higher financial risk-taking (Cueva et al, 2015). Induced changes in hormones associated with increased aggression, such as cortisol and testosterone, have been shown to shift financial investments toward riskier assets. Anger is also associated with riskier behaviour while driving (Deffenbacher et al, 2003).
Uncomfortably hot temperatures can significantly increase an individual’s propensity for aggression (Anderson, 1989). This has been documented in multiple populations, including American football players (Craig et al, 2016), prison inmates (Mukherjee and Sanders, 2021) and university students (Almås et al, 2020).
Overall, research on cognition, sleep and mood suggests that the effects of high temperatures on time and risk preferences are difficult to predict. The sleep and cognition channels predict that higher temperatures increase risk aversion. But if aggression were the operating channel, higher temperatures would decrease risk aversion and increase risk-taking.
All three channels predict a significant increase in impatience: during periods of very hot weather, people tend to put a higher value on current payoffs than future ones.
What is the evidence from economic research?
Economists have not yet established whether there is a causal relationship between temperature and economic preferences. The main challenge is that people who experience high temperatures are on average different from those who don’t. For example, poorer households tend to live in buildings that are less protected from heat, and individuals with less wealth tend to work in industries with higher exposure to dangerous temperatures. In turn, these differences could be associated with economic preferences.
Due to the difficulties of addressing these challenges, most evidence comes from laboratory experiments. One such study finds that people operating in warmer temperatures are less likely to gamble, more likely to rely on quick processing and more likely to perform poorly on cognitive tasks (Cheema and Patrick, 2012). Another suggests that high ambient temperatures lead individuals to pursue high-riskoptions (Wang, 2017).
The most recent study is a laboratory experiment conducted among university students in Berkley, California, and Nairobi, Kenya. It finds no evidence that high heat affects risk-taking, time preferences or cognition (Almås et al, 2020). But the authors do find that heat affects people’s willingness to destroy others’ earnings in a game known as joy of destruction. This is consistent with the temperature-aggression hypothesis.
The strength of the laboratory approach is the ability to allocate temperatures randomly, overcoming potential discrepancies in real-world data. The disadvantages are the typically small sample sizes, the assumed immediate nature of the temperature effects, and the limited subject pool. Subjects are often young university students of relatively high socio-economic status and are not usually representative of wider populations.
Equally important are the timing and duration of exposure to heat. For example, if heat affects economic preferences via sleep quantity and quality, laboratory studies, which cannot readily randomise overnight sleep temperatures, will fail to detect the effect. Similarly, if temperature effects are the result of cumulative exposure to heat, 20 minutes or even an hour of exposure to high temperatures may be insufficient to induce a change in preferences.
For these reasons, field studies are also necessary. Research using a large sample of Indonesian men and women of all ages and socio-economic status makes use of variations in outdoor temperatures caused by the quasi-random allocation of survey dates (Escobar et al, 2021). The results suggest that high night-time temperatures increase impatience the next day, as well as violations of expected utility theory – a framework used by economists to analyse decisions when outcomes are uncertain. The dominant channel behind these effects is depleted cognitive functioning, mathematical skills in particular.
The importance of night-time temperatures is especially relevant because global warming is causing night-time temperatures to increase at a greater rate than day-time temperatures across most of the earth’s landmass (Cox et al, 2020).
Who is most likely to be affected?
More evidence is needed before we can conclude how higher temperatures will affect economic decision-making. But current research suggests that people will become less patient and less rational in their willingness to take risks with rising temperatures.
What’s more, any effects will be spread unevenly. The effects are especially likely to differ between high- and low-income countries. There are many possible reasons why, but one crucial difference between countries is their ability to adapt to higher temperatures.
If cognition is the main operating channel, as some findings suggest, the negative effects of temperature on time and risk preferences are likely to be larger in developing countries and low-income households. This is because the negative effects of temperature on learning and cognition are larger for poorer countries, and air conditioning offsets most of the disruptive impacts of heat on learning, as shown in US studies (Park et al, 2020).
Rising incomes mean that the coverage of air conditioning across the developed world is becoming widespread in homes, office spaces and public transport, protecting people from the harmful impacts of heat. There is evidence for this in the 75% decline in heat-induced fatalities in the United States since the 1960s, which is almost entirely due to residential air conditioning (Barreca et al, 2016).
Conversely, in the developing world, the capacity to use technology to adapt to a warming climate is significantly lower. The data indicate that 13% of dwellings in the world are without access to electricity (World Bank, 2021). Further, the energy infrastructure would have to be upgraded to cope with the demands that high energy use imposes on the system. Finally, households need to have the purchasing power to acquire air conditioning units. They also need to be able to afford the electricity bills.
For these reasons, only 8% of the 2.8 billion people living in the hottest parts of the world own air conditioning units (International Energy Agency (IEA), 2018). Families in low-cost dwellings are especially vulnerable, as their houses tend to be built with materials that store heat rather than insulate them from high outdoor temperatures (Naicker et al, 2017).
As the evidence across multiple contexts grows, we will be more able to predict whose economic preferences are affected by rising temperatures. The effort is necessary: the answers are not trivial. Optimal and rational decision-making shapes the way we think about the future, the investments we make in education and health, and even our openness to adoption of innovative technologies such as new vaccines.
Where can I find out more?
- The effects of temperature on economic preferences: working paper by Escobar et al, 2021.
- Destructive behaviour, judgment, and economic decision-making under thermals tress: working paper by Almås et al, 2020.
- Influence of warm versus cool temperatures on consumer choice: journal article by Cheema and Patrick, 2012.
- An empirical study of the impacts of ambient temperature on risk taking: journal article by Wang, 2017.
Who are experts on this question?
- Ingvild Almås, Stockholm University
- Maximilian Auffhammer, University of California, Berkeley
- Michelle Escobar, Monash University
- Solomon Hsiang, University of California, Berkeley