Living with a hotter world

Views From the Marsh - David Owen

This week saw us publish a detailed note, looking at the patterns of natural disasters we are seeing globally. Some of the observations we made for the US are summarised later in this post, behind the paywall.

However, we thought it worth highlighting one aspect of a changing climate that is often not included in estimates of the costs of a warmer and wetter world and more frequent and extreme climate related events; namely, the potential impact on GDP, because of runs of very hot days.

The chart below from an earlier research piece, we update each year, confirms what we all know; drilled down by region, European summers really are getting hotter.

This looked at changes in so-called cooling day indexes for July, for around 1,500 NUTS-3 regions of the EU-27. These track days when the average temperature is 24°C, or more, probably implying peak daily temperatures of 30°C, in some cases well over 30°C.

The upward slope of the regression line confirms the general rule for much warmer summers, compared to 1990; although there are some regions - 136 altogether - that have bucked the trend, including some of the hotter NUTS-3 regions of the EU.

To our knowledge, it was our former colleague Leo Doyle who first published a study weather adjusting UK GDP data, breaking the potential impact on sectors, some 30 years ago.

Back then, it received relatively little interest. But that was some 30 years ago.

More recently, we have seen back of the envelope calculations assuming that for every day when the temperature exceed 32°C, halve a day’s output is lost, which is never regained.

Hence a run of days when temperatures exceed 32⁰C, such as in 2023 and perhaps this year, could in some countries’ cases (say Southern Europe) reduce GDP during that period (say Q3) by 1%, or more.

Of course, one cannot insure oneself as a business against the effects of extreme temperatures, in the way that one can get insurance for flooding and wildfires – although, in a hotter or wetter world, this could also become prohibitively expensive, or ultimately in the case of some companies or households (say, because of flood risk) not available at all.

Now researchers from the ECB, making further use of the NUTS-3 regional EU figures that we highlighted earlier, have updated an earlier working paper for publication in the European Economic Review. This shows the potential long-term costs of repeated extreme temperature, droughts and flooding, on GDP (see here), and deserves to be widely read.

To quote the authors “Output is 1.5 percentage points lower two years after a summer heatwave, 3 percentage points lower four years after a drought and 2.8 percentage points lower four years after a flood. Both lower regional population and lower labour productivity per hour worked contribute to the medium-term decline.”

But there is also the finding that when it comes to flooding, richer regions with fiscal space, and access to credit and finance, can build back better. Migration away from impacted regions might also imply the need for intra-country or intra-region fiscal transfers.

Importantly, the finding of a 1.5 percentage loss in output 2 years after a summer heatwave, is broadly in line with those back of the envelope calculations referred to earlier. As we continue to stress ESG is not the same as pricing in the risks associated with a changing climate better.

Now on to some observations on the US.