Graph of climate related deaths 1900-2019

Craig Kelly and climate-related deaths

Three reasons why Craig Kelly’s recent climate change post is misleading and dangerous.


On 8 February, Craig Kelly criticised a BBC report about future climate related deaths. He presented a graph which he said showed the report was “complete unmitigated BS”.


Mr Kelly’s report is misleading and dangerous. His graph doesn’t show all the data, he doesn’t appear to have examined the data apart from this graph, and he has ignored the conclusions of experts.

The claims in detail

Mr Kelly comments on a BBC report about Mark Carney, formerly head of the Bank of Canada and of the Bank of England, and now United Nations envoy for climate action and finance. Mr Carney predicts that climate crisis will cause a similar number of deaths as the Covid pandemic every year by mid-century unless action is taken. (I’ve posted a screen shot of the whole post at the end.)

Mr Kelly calls this “complete unmitigated BS …. just another scare campaign designed to hand more power to the global elites and the big wbankers.”

He bases these assessments on this graph of “Climate-related death risk, 1920-2020”, which shows the death toll falling rapidly. He comments: “world mortality rates from climate disasters have plunged. The data evidences that the world is heading for mortality rates close to zero from climate disasters.”

“Where are the fact checkers?” he asked. So that was my invitation! 🙂

Three reasons why Mr Kelly’s post is misleading

1. He cherry picked the data

The information shown in the graph comes from the EM-DAT (International Disaster Database). The database lists disasters related to climate (floods, droughts, etc) and other causes (e.g. famines) and estimates the death tolls, financial costs, etc.

The database commences in 1900, but Mr Kelly has only shown the data from 1920 onwards. That is interesting because the decade before his graph had much fewer fatalities, not much above the death toll in the last 40 years.

Mr Kelly has cherry-picked the data and not shown you it all!

I have downloaded the database, and the full graph is shown at the top of this post. While there has been a fall in climate-disaster deaths over the 120 years, it isn’t a very clear trend.

Instead, the reality has been a few disastrous events in China, India and Bangladesh in the period 1928-1965, with fewer fatalities either side of that. There is insufficient information here to draw conclusions. Further analysis is required.

2. He didn’t explain the causes

Many factors impact on the number of climate related deaths. The severity of storms, floods and droughts is one factor, but other factors include:

  • whether the climate event occurs near a populated area, or not;
  • whether the population is well prepared via warnings and safety information;
  • the effectiveness of protection measures such as dams, sea-walls, levies, etc;
  • the options available for escape.

It isn’t a coincidence that the worst disasters have occurred in less developed countries, mostly in highly populated Asia. As these countries have developed, their populations have become more protected and fatalities have been reduced.

The high death toll in the period 1928-1965 was caused predominantly by 6 disasters in China, India and Bangladesh. In each of these events – 3 floods and 3 droughts – more than a million people died. The death toll in such events is very much affected by the existence, or not, of water infrastructure – levees to contain flooding and dams to relieve drought.

For example, flooding on the Yangtze River has been a periodic problem for centuries. Levees built to contain flooding have on occasions (e.g. in the disastrous 1937 flood) made the problem worse when they were overtopped. Since that time, China has invested heavily in improving the levee system and building dams, and so reduced the death toll.

(EM-DAT estimates that 3.7 million people died in this event, but other estimates vary from 0.3 million to 2 million. If a lower figure was correct, this would change the graph significantly.)

So it is likely that the major cause of the rise in climate-related deaths in the early 20th century was increasing population in vulnerable areas, and the major cause of the subsequent decline is the improved flood and drought infrastructure in those Asian countries.

(Imagine thinking the decrease in road deaths is because there are less cars on the road, when in fact there are more cars on the road, but road deaths are decreasing because of safer roads and cars, greater public awareness and stronger policing. It is similar here.)

The graph reflects the past but probably doesn’t represent the future, because the situation is different now. Mr Kelly’s post doesn’t consider this.

3. He didn’t report what the experts say

There are two major sources of expert information on current and future impacts of climate disasters – insurance companies and climate scientists.

Insurance companies and the banking and finance industry employ actuaries to make sophisticated risk assessments of potential threats to their businesses. Mark Carney, who Mr Kelly dismisses as “a big end of town banker from central casting”, is reflecting what the banking and insurance sector has calculated.

For example, Deloitte reports that the US insurance industry expects climate change to have a major impact on insurance premiums and risk. The Insurance Council of Australia has similar expectations.

The risk to human life comes in several forms:

  • Direct loss of life caused by floods, droughts, bushfires, storms, heat exposure, etc.
  • Indirect deaths from various diseases (e.g. diarrhoea, malaria, childhood undernutrition, respiratory issues, etc)
  • Dislocation, forced emigration and loss of income and, possibly, climate-related conflict.

There are quotes from the business community at the end of this post.

The right way to assess climate change

This database of deaths is a poor way to measure the impacts of climate change.

  • The time scale is too short and the graph too variable to show a clear trend. (This is elementary statistics wisdom.)
  • There are too many factors likely to affect the mortality rate (warning systems, preparedness, rescue infrastructure, location of natural disaster) to draw direct conclusion without a lot of work to separate out the impacts of the different factors. (This too is elementary statistics wisdom.)
  • The database is incomplete.

Meanwhile the science is clear. Natural disasters are increasing in number, globally, in the US and in Australia, as predicted by climate science. Improved warning, protection and disaster procedures are keeping the death toll down, but they can only do so much.


Mr Kelly’s post is misleading:

  1. His graph doesn’t present all the data.
  2. He hasn’t explained the causes of the rise and fall in climate-related fatalities and missed that it is other factors than climate change that best explain his graph. The graph doesn’t predict the future because of this.
  3. He has ignored (and impugned) the experts who say climate-related fatalities, illness, dislocation and economic cost are all on the increase.
  4. Ignoring these realities is therefore foolish and inhumane.

We should be listening to Mark Carney, not calling him names.

Quotes from the business community

“The world’s biggest companies face $1 trillion in climate change risks.” (CDP)

“The evidence on climate risk is compelling investors to reassess core assumptions about modern finance. ….. our understanding of how climate risk will impact both our physical world and the global system that finances economic growth. These questions are driving a profound reassessment of risk and asset values. …. In the near future – and sooner than most anticipate – there will be a significant reallocation of capital.” (Larry Fink, Chairman and CEO, BlackRock)

“The Bank of England has said up to $20tn (£16tn) of assets could be wiped out if the climate emergency is not addressed effectively.” (Mark Carney in the Guardian)

“Unmitigated climate change and extreme weather events will have significant health impacts, including respiratory issues, the spread of diseases and premature deaths. Climate change and extreme weather events will also create major productivity losses, particularly in industries that require workers to be outside. Migration forced by climate change has already displaced an average of 26.4 million people per year globally between 2008 and 2015. By 2050, climate change will force 50 to 700 million people to emigrate.” (Ceres)

Screenshot of Craig Kelly post

Leave a Reply

Your email address will not be published. Required fields are marked *