Seasonal temperature variability and economic cycles

Produced as part of the Climate information for adaptation CCCEP research programme theme

In many countries, gross domestic product (GDP) varies by several per cent from quarter to quarter. For example, in the UK between 1990 and 2018 total production in the second quarter was on average about 5 per cent smaller than in the fourth quarter. Economists have long suspected that temperature is one driver of these fluctuations but their models typically have focused on only one country and on proximate factors such as preferences and production technologies, without attributing those proximate factors to fundamental drivers. Because the influence of temperature has not been explicitly examined, it has remained unknown – but as climate change is projected to change the seasonality of temperature, the question appears very relevant.

In this paper, the author assembles a new set of quarterly and six-monthly temperature and GDP data for 98 countries and develops a new estimation strategy to attribute observed fluctuations in GDP to changes in temperature.

The results show a much larger diversity of seasonal cycles of GDP around the world than previously reported and suggest that differences in temperature between summer and winter can explain a major part of the observed differences in GDP between these seasons. This finding is in contrast to previous work, which concluded that temperature plays at most a minor role for seasonal cycles of GDP. For a small sample of European economies, the effect of temperature can be attributed to sectors in which production is relatively exposed to ambient temperature.

Key points for decision-makers

• The author’s data cover the period 1990–2018, using a global dataset of quarterly GDP covering 98 countries, a dataset on quarterly gross value added (GVA) for 35 European economies, and a dataset on temperature and precipitation.
• Previous findings on quarterly fluctuations in GDP suggest a peak in the fourth quarter, possibly associated with consumption around Christmas, and a trough in the second or third quarter, possibly associated with mid-year vacations. However, these are observed in only about half of all countries in the global sample and primarily in the Northern hemisphere.
• Countries with greater temperature differences between summer and winter also tend to have larger seasonal economic cycles. This pattern can be observed in both the Northern and Southern hemispheres and holds when differences between countries in terms of GDP per capita, sectoral specialisation, religion, latitude, and many other control variables are taken into account.
• In Europe, the effect of temperature on the seasonality of GDP is largest and most significant in sectors exposed to ambient temperatures, such as agriculture, construction and manufacturing.
• Projections made by climate models suggest that with climate change, seasonal temperature variability will decrease at higher latitudes and increase at lower latitudes. These projections, together with the results of this new paper, suggest that climate change will lead to a reallocation of economic production across seasons, increasing or decreasing seasonal differences in GDP.