How researchers get heard
Abstract lines

Climate Change & Indefensibly Big Numbers

Climate change is a potential widespread catastrophe, and in some cases an actual living catastrophe.

Our species uses big numbers as one of its primary signifiers of catastrophe: deaths; property damage; lost economic growth; etc.

It’s understandable, then, that some scientists have gravitated to using big numbers to bring into focus the catastrophe of the recent Australian wildfires. One of the biggest and most reported of those numbers: the fires have killed more than 1 billion animals.

That’s an arrestingly big number — almost perfect in its Holocaust-like dimensions for signifying catastrophe and its rhyme with the numbers corporations and ministerial budgets trade in.

But is it accurate?

And does its accuracy even matter, if it serves the larger purpose of getting people on board with taking action against climate change?

I listen to Tim Harford’s podcast More or Less: Behind the Stats, which every week goes behind statistics in the news and asks: are these accurate, and does that matter?

In the latest episode, he interviewed Kate Parr, a professor of earth, ocean and ecological sciences at the University of Liverpool, on the 1 billion claim.

The number, it turns out, comes from a 13-year-old population density study of animals per hectare of New South Wales uncleared bush. (The survey was funded by WWF-Australia; WWF is currently one of my clients.)

The survey estimated that there were about 170 animals in each hectare of New South Wales bush: 18 mammals, 21 birds and 130 reptiles.

For the fires, the same University of Sydney scientists who conducted the survey simply multiplied 170 by the number of hectares burned to arrive at their “animals killed number” — first issuing an estimate of 480 million dead animals, then over 1 billion as the fires spread.

Parr outlined to Harford a wide range of problems with all of this:

  • It’s unclear where the data came from in the original study;
  • The researchers didn’t have data for the density of every species, so they estimated those numbers from “closely related species, which introduces more error”;
  • The researchers had estimates for all reptile species but not individuals — which matters because 3/4 of the animals estimated to have died were reptiles;
  • Wildlife in New South Wales is adapted to fire — it’s part of the ecology, and while some species will die, most will avoid the fire through flying away (if you’re a bird) to burrowing and entering a torpor state for three weeks (if you’re a reptile or small mammal), as well as becoming more active at night to adapt to loss of cover;
  • Studies from previous deadly fires in Australia (such as the 2009 Black Saturday fire in Victoria that killed 173 people) show surprising resilience in some species — certainly not decimation.

After listening to Parr, the 1 billion figure started to sound a lot less defensible to me — and more like what scientists accuse marketers of routinely doing.

Parr admitted to Harford that the current Australian fires are “unprecedented” in the combination of their scale, size and the extreme temperatures New South Wales is experiencing. “It means we haven’t seen anything like this, so this is a major event for these animals,” she said.

Harford reached out for comment to Dickman, who sent Harford this statement: “While it is true that in many or even most fires the animals will escape, the current fires are unprecedented in intensity, speed and scale. The consensus among those who have witnessed the fires is that escapees will be few and far between.”

But that’s not science. That’s a scientific expert eyeballing an event and offering an assessment floating untethered from data and the literature.

Parr herself told Harford that she has rarely seen vertebrates killed by fires she has witnessed in Australia and Africa. Which eyewitness account are we to believe?

As Harford gently concludes at the end of the segment: “We may never know the truth. Without a baseline census to refer to, we will struggle to know how many animals have died.”

Except that very few of us outside of science are really struggling.

We count on science to make that struggle for us, and make it loudly. To be the buzzkill for non-replicated studies on ocean acidification and its impacts on fish behavior, or claims about climate change causing an insect apocalypse, which might also be due to light pollution. We count on science to come back two years from now and tell us: hey, we were right on the money about those instant claims of wildlife fatalities, or we were way off.

Thought leadership is about making the things you see as an expert visible to others who don’t see them and who will benefit from them. It might deploy in reaction to an event, but it isn’t reactive. And it’s more than just a number or an opinion that cuts against all the data we already have, and assumes the present is unique.

In an era of increasing certainty about climate change and what should be done about it, we will continue to see a lot of big numbers and big scientific claims in the context of climate change, all seeking to motivate action. Perhaps even more, now that the “1 billion killed” big number proved irresistible for media.

Are the Australian wildfires a catastrophe? Absolutely.

Do we need an indefensible number to reinforce that message? Perhaps. Hard to test that now.

Is putting forward that indefensible big number good for science? It is even good for fighting climate change? You tell me.

Before you answer, perform this commutation test: Replace “indefensible big number” with “non-replicable science.”

Because isn’t that big number just bad science in a really small package?