Future heat stress to reduce people's purchasing power

Abstract

With increasing carbon emissions rising temperatures are likely to impact our economies and societies profoundly. In particular, it has been shown that heat stress can strongly reduce labor productivity. The resulting economic perturbations can propagate along the global supply network. Here we show, using numerical simulations, that output losses due to heat stress alone are expected to increase by about 24% within the next 20 years, if no additional adaptation measures are taken. The subsequent market response with rising prices and supply shortages strongly reduces the consumers' purchasing power in almost all countries including the US and Europe with particularly strong effects in India, Brazil, and Indonesia. As a consequence, the producing sectors in many regions temporarily benefit from higher selling prices while decreasing their production in quantity, whereas other countries suffer losses within their entire national economy. Our results stress that, even though climate shocks may stimulate economic activity in some regions and some sectors, their unpredictability exerts increasing pressure on people's livelihood.

Fig 1. Global mean temperature anomaly and heat stress-induced direct output losses.

A,C: Temporal evolution of A global mean temperature anomaly relating to pre-industrial level and C heat stress-induced direct output losses for the four climate models, HadGEM2-ES (blue), IPSL-CM5A-LR (green), MIROC5 (red), GFDL-ESM2M (orange) and RCP2.6 (dashed) and RCP6.0 (dotted), and ensemble median (black line). The higher temperature anomaly of climate model IPSL-CM5A-LR compared to the other climate models is due to the relatively lower model-internal pre-industrial temperature. B,D: Regional maps of B absolute and D relative annual direct output loss due to heat stress based on the respective regional projected median for 2020–2039. Regions with an absolute or relative direct annual output loss below USD 1bn or 0.2% of baseline (unperturbed) production are depicted in light purple.

Fig 2. Total production losses and gains.

A: Temporal evolution of global total production change for the four climate models, HadGEM2-ES (blue), IPSL-CM5A-LR (green), MIROC5 (red), GFDL-ESM2M (orange), and RCP2.6 (dashed) and RCP6.0 (dotted), and ensemble median (black line). B: Regional map of total production gains and losses based on the respective regional projected median for 2020–2039 and the full GCM–RCP ensemble. Quantities are given relative to the baseline (unperturbed) production.

Fig 3. Consumption and total production change per country for the time period 2020–2039.

The area of each dot is proportional to the corresponding country’s baseline (unperturbed) production. The dot colours denote the geographic regions (see S2 Table). Quantities are given relative to the baseline (unperturbed) production and consumption, respectively.

Fig 4. Consumption and expenditure change.

A,C: Temporal evolution of A global consumption change and C global expenditure change for the four climate models, HadGEM2-ES (blue), IPSL-CM5A-LR (green), MIROC5 (red), GFDL-ESM2M (orange), and RCPs 2.6 (dashed) and 6.0 (dotted), and ensemble median (black line). B,D: Regional maps of B annual consumption change and D annual expenditure change based on the respective regional projected median for 2020–2039 and the full GCM–RCP ensemble. Quantities are given relative to the baseline (unperturbed) consumption and expenditure, respectively.

Fig 5. Total production, consumption, and expenditure change for economic networks of 2012, 2002, and 1992.

Temporal evolution of A total production changes, B consumption change and C expenditure change for economic networks of 2012 (blue), 2002 (orange) and 1992 (red). Ensemble members and median are depicted in light and thick lines, respectively.

Fig 6. Results are robust against choice of consumption price elasticity and temperature-productivity-uncertainty.

Median of RCPs-GCMs-Ensemble and likely range (16.7 to 83.3 percentiles) of heat stress-induced changes in consumer expenditure and consumption in 2039 relative to unperturbed baseline for A different consumption price elasticities and B change in intensity of productivity reduction due to heat stress.