The world has become more resilient to disasters, but investment is needed to save more lives

Deaths from disasters have fallen, but we need to build even more resilience to ensure this progress doesn't reverse.

In 1970, more than 300,000 people died when a strong cyclone hit the coast of Bangladesh.1 In 1985, another storm caused 15,000 deaths. Just six years later, another killed 140,000.

Fast-forward to 2020. Bangladesh was hit by cyclone Amphan, one of the strongest storms on record in the Bay of Bengal. The death toll was 26 - barely visible on the chart below, compared to these very deadly disasters.

That's 26 too many deaths, and the cyclone also caused huge amounts of damage: millions of people were displaced, and there were large economic losses. But tens - possibly hundreds - of thousands of lives were saved through early warnings, evacuations, and increased resilience. People in Bangladesh are much better protected from disasters than they were a few decades ago.

This development is part of a longer-term and widespread success in reducing humanity's vulnerability to storms, floods, earthquakes, and other hazards.

Number of deaths from storms

Storms include tornado, hail, thunderstorm, sand storm, blizzards and extreme wind events

The world has become more resilient to disasters

Bangladesh is not an isolated example. We can observe long-term improvements in the world's resilience.

Here, I will look at data published by the International Disaster Database, EM-DAT, which stretches back to 1900. In the chart below, I've shown the number of deaths from disasters, given as the decadal average. This is helpful as there is a lot of volatility in disasters from year to year.2 You can also explore this data annually.

The number of people killed in disasters has fallen a lot over the last century. That's despite there being four times as many people. That means the decline in death rates has been even more dramatic.

Decadal average: Annual number of deaths from disasters

Disasters include all geophysical, meteorological and climate events including earthquakes, volcanic activity, landslides,drought, wildfires, storms, and flooding. Decadal figures are measured as the annual average over the subsequent ten-yearperiod.

Single events that caused over a million deaths were not rare in the first half of the 20th century. These events were mostly floods or droughts and were often linked to agricultural shocks that caused hunger and starvation. There are many similar examples in humanity's history before the year 1900. A strong El-Niño weather event in the 1870s led to severe famines across China, India, and Brazil, killing tens of millions.

To be clear, deaths haven't declined so steeply because disasters are becoming less frequent or intense. This data also doesn't mean climate change isn't happening or isn't worsening weather events. The main reason that fewer people are dying is that we've gotten better at protecting ourselves and each other.

As we'll see later, the toll of disasters depends on our physical environment, economic resources, political systems, technological advances, and cooperation.

Massive productivity improvements have made agricultural systems far more resilient to shocks. Changes in political systems have reduced the risk of famine. As my colleagues Joe Hasell and Max Roser have shown, famines are rare in well-functioning democracies.

Weather forecasts have improved dramatically over the past 50 years, and populations can better prepare for storms, floods, droughts, and wildfires. Many more countries have early warning systems.

And overall, people worldwide are richer than they were a century ago. It's often the poorest who are most vulnerable to disasters. Having an earthquake-proof home, access to electricity and clean water, air conditioning or heating, enough money to absorb shocks in energy or food prices, and resources to recover after a disaster means you can protect yourself. Billions of people have gained access to these basic resources over the last century.

Many researchers and organizations have noted this reduction in vulnerability to various types of disasters. The World Meteorological Organization found that deaths from climate and weather-related disasters decreased almost 3-fold from 1970 to 2019. Researchers Giuseppe Formetta and Luc Feyen studied a different disaster database - Munich RE's NatCatSERVICE - and found a large reduction in global vulnerability to disasters from the 1980s to 2010s.3

"No such thing as a natural disaster"

Three factors determine the risk of damage when a hazard hits.

First, the characteristics of the hazard itself. Is it a flood, drought, hurricane, or heatwave? What's its magnitude, speed, or power? And how long does it last? Is it a 30-minute downpour or a 5-day deluge of heavy rainfall?

Second, the number of people or the amount of infrastructure exposed to the hazard. Does the earthquake strike a densely populated city or a rural area? How many people live on a coastline inundated by storm surges or sea level rise? How much stuff - buildings, bridges, roads, and other infrastructure - is in harm's way?

Third and finally, the vulnerability of those who are exposed. A heatwave in Dubai will be less harmful than one in New Delhi because most people in Dubai have air conditioning. A strong earthquake in a country with quake-resistant infrastructure will be less damaging than the one that struck Haiti in 2010. Vulnerability is often strongly linked to income: poorer countries and communities tend to have fewer resources to protect themselves and respond afterward.

Disaster risk, then, sits at the center of all three. An increase in any of these dimensions increases the risk, while a reduction lowers it.

This is why you might hear the phrase, "There's no such thing as a natural disaster". Hazards only become disasters when they impact societies and people. A hurricane, for example, is not a disaster until it hurts or kills people or destroys homes in its path.

How does this framework explain the dramatic decline in deaths from disasters over the last century?

The hazard component has not declined. It has probably increased - on several dimensions - because of climate change.4

Exposure has not declined either. There are more than four times as many people on the planet than a century ago. People have migrated to cities, often along coastlines, where events such as storm surges, cyclones, and flooding are more likely. That means more people in harm's way. What has reduced, though, is acute exposure to some events. Better weather prediction and early warning systems mean people can evacuate before a hazard hits.

What has declined is vulnerability: a long list of improvements have made communities less vulnerable. Our agricultural systems are more productive and recover from damaging events. Political systems allow for national and international support before or after a hazard strikes. People live in better buildings. Some have heating or air conditioning to protect them from extreme temperatures.

How to reduce disaster risk going forward

How should we think about these three factors going forward?

First, we will never be able to stop hazards completely. There will never be an end to earthquakes, cyclones, or extreme rainfall. What we can determine is how much worse they get in a changing climate. Hazards at 3°C of warming will be worse than at 2°C, which will be worse than at 1.5°C. We need to reduce our carbon emissions and avoid the worst impacts of climate change. The better we do here, the less exposed and vulnerable we will be.

But that alone won't be enough. Regardless of how successful we are in reducing our emissions, climate change is already here, and the world will get warmer even on our most ambitious pathways. Societies will need to be more resilient to these changes, and they should be protected from events unrelated to climate change - the types of events that our ancestors were exposed to 50, 100, or 1,000 years ago.

To reduce exposure, we must understand when and where hazards are most likely to hit. We must understand which regions are most vulnerable to sea-level rise or wildfires. People are still moving into areas at serious risk in the future, exposing many of them to disruption and damage.

Increasing the coverage of early warning systems will help. When a hazard is imminent, people can evacuate and their exposure can be temporarily reduced. According to the World Meteorological Organization, around one-third of the world - mostly in the poorest countries - does not have these systems. I've recently written about this.

Finally, there is a lot of room for reducing vulnerabilities. Poverty makes people most vulnerable; that's why I argue that lifting people out of poverty is an essential solution to reducing climate risks. Investing in infrastructure, making agriculture more productive, and building strong political governance are all vital. Poorer countries tend to have higher death rates from disasters, as the chart below shows. (Note that this is on a logarithmic scale.)

Death rates from disasters vs. GDP per capita, 2021

Death rates are measured per 100,000 people. Disasters include all geophysical, meteorological and climate eventsincluding earthquakes, volcanic activity, landslides, drought, wildfires, storms, and flooding. GDP figures are adjusted forcross-country price differences.

This relationship is even stronger for the total number of people affected by disasters, which includes those left homeless, injured, or requiring assistance. People in lower-income countries are much more vulnerable to disasters such as drought, which affects many more people. Poorer infrastructure also means they're slower to respond and recover from disasters, leaving more people affected.

Richer countries can help and have committed to doing so with adaptation funds within the Paris Agreement on climate change.

One way to build more resilient communities is to learn from other countries, like Bangladesh, that have saved many lives with the right preparation and response. The problem is that we tend to see the large and fatal events where things have gone wrong and miss the small events where people were protected. Disasters hit the news; averted ones don't.

Thankfully, some researchers are trying to change this. The project "Disasters Avoided", led by a team of disaster risk experts, tries to highlight case studies of events where disasters were prevented. As Ilan Kelman, one of the authors, writes: "We frequently see headlines about disasters. But where are the headlines covering the good news of lives saved and damage averted when disasters do not happen?"

To keep up with escalating climate change, we will need to move faster

Over the last century, we have outpaced the impacts of climate change on natural disasters. Deaths have fallen despite climate change because we've built more resilient societies.

Think of it like being in a race. Climate change has been jogging while we've been running. We've mostly stayed ahead, but there is no guarantee that things will stay that way. Slow down, and we'll be overtaken. Stay at the same pace, and we'll probably still be overtaken as the impacts of climate change accelerate.

If we fail to invest in protection measures and development trends slow down, then the progress we've made over the last century could easily reverse, and disaster deaths could start to rise again.

What's key, though, is that the direction of that trend - a continued fall or a reversal - is up to us.

None of this will happen on its own. Bangladesh's success was driven by local communities and investment in early warning systems. Chile and Japan's resilience to earthquakes came from architects, engineers, and governments upholding strict building standards. The dramatic decline in famine came from technological revolutions in agriculture and populations pushing for political rights and accountability.

International cooperation and support will be needed to ensure that the poorest and most vulnerable are not left behind.

Endnotes

1. Note that there are some uncertainties around these estimates, especially when the indirect impacts of the disasters are considered. All estimates are in the range of hundreds of thousands of people.
2. One caveat to keep in mind about this data. Data quality and completeness are lower the further we go back in time. I'll cover these issues in more detail in an upcoming article. It's mostly large events captured in the first half of the 20th century because smaller ones were often not recorded or reported. That means many smaller events - and the deaths they caused - are missing.
3. Formetta, G., & Feyen, L. (2019). Empirical evidence of declining global vulnerability to climate-related hazards. Global Environmental Change, 57, 101920.
4. The latest report from the Intergovernmental Panel on Climate Change (IPCC) notes that the frequency and intensity of heatwave events have likely increased due to climate change. Heavy precipitation events have likely increased, particularly across Europe, North America, and Asia. Agricultural droughts have likely increased in some regions.