
Tim Harford investigates heat deaths, the carbon footprint of eggs, and ironing out Wales
Loading summary
A
Hello and welcome to More or Less your weekly guide to the numbers in the news. And in Wales, I'm Tim Harford. This week, Conservative backbenchers were conspicuous by their absence in a recent Prime Minister's questions. Running scared or a run of bad luck? We'll run the numbers. Are white eggs really more environmentally friendly than brown eggs? If so, why? And our listeners ask if Wales is actually bigger than England. Or once you take into account all the lovely crinkly bits. But first, some hot takes on the Internet have left some of us feeling cold. Posts comparing deaths from high temperatures in the US and Europe. Then for some reason comparing that to US gun deaths swept across the Twittersphere
B
as summer comes into view. Always remember that more Europeans regularly die in heat waves than Americans die in gun violence.
A
The tweet was illustrated by a bar chart showing that almost 70,000 Europeans died because of heat in 2024, compared to around 2,000Americans. The chart also shows that almost 50,000Americans died due to gun violence. Now, we're not going to waste time comparing gun deaths in a country of 330 million people to a continent of 740 million. That would be silly, and not just because a gun death is much easier to define and arguably prevent than a heat death. But it is eyebrow raising that the US has so few heat deaths when compared to Europe. After all, a lot of the US is hot, really hot. They have deserts, they have Death Valley, we have the Wye Valley. So what's going on? Well, some other Twitterers, tweeters, twits, believe they have the answer. Aircon economists estimate that if Europeans used AC as much as Americans do, it would save up to 100,000 European lives every year. Your heat related deaths are higher and it's hotter in the us. It's empirical. You have the data.
C
Ooh.
A
To add some calm into proceedings is our ever clement climate correspondent, Lizzie McNeil.
D
Hello, Lizzie.
B
Hi, Tim. Should we get the big bit of information out of the way first?
A
Sure.
B
First, we are talking about deaths from when it's too hot, so not when it's cold or when it's ambient. Secondly, you cannot compare these stats.
A
But comparing these stats is what we do.
B
I know it is, but we can't compare the figures in the viral graph because they're not really counting the same thing. You see, the US and Europe estimate and record heat deaths differently. Europe uses excess mortality, whereas the US used whether heat appeared as a contributing or underlying factor on death certificates.
A
So Europe looks at how many more people died than would have been expected to die and ties all that to heat. Whereas the US just counts cases where heat was explicitly named as a factor.
B
Exactly.
A
Well, are the two figures correct with were there just under 70,000 excess deaths in Europe due to summer heat?
B
So the figure is from a peer reviewed paper. It's from a Spanish study that tried to quantify how many Deaths occur across 32 European countries due to heat. And the team used daily regional temperature and mortality records from 2015 to 2019 as the baseline for how many deaths they'd expect to see. They then looked at the temperature observations and forecasts and weekly mortality for each year from 2022 to 2024 to calculate how many more people died due to HOT. And yep, they found that across those 32 countries, there were 67,000 excess deaths in 2024.
C
Wow.
A
Okay, so these are all excess deaths. We don't know for sure whether the
D
deaths were directly caused by the heat.
A
Is that a better way of counting deaths than looking at death certificates?
B
So the Centres for Disease Control and Prevention seem to think so, as they've announced they're going to try and start using excess death data too. The problem with heat deaths is that it's generally not people dying of something measurable like heat stroke or exposure. It's generally vulnerable people with pre existing health conditions whose condition is exacerbated due to the heat. That's why we sometimes see an effect that's somewhat morbidly known as harvesting, where mortality is higher than expected levels during heat waves, then lower than expected levels in the following months, as some of the vulnerable people who might have died later die a bit sooner.
A
So a rise in mortality during a heat wave could mean these people had
D
pre existing conditions or gravely ill, and
A
they might have been expected to die
D
within a few months.
A
And the heat wave meant they died sooner than they otherwise would. Yeah, well, that's the European figure. What about the American figure?
B
Yeah. So as mentioned, the Centers for Disease Control CDC currently only counts deaths when heat is explicitly mentioned on the death certificate.
A
Well, that might be a problem because from what you've told me about heat deaths, many of them are due to underlying conditions. So you might have like a heart
D
attack or a respiratory failure.
B
Yeah, and so that's what would be recorded on the death certificate. Now, heat does sometimes get recorded, for example, if someone died of heat stroke, but for other more complex cases it requires an investigation. States also fill out death certificates differently and just don't have the staffing power to assess every death during hot weather to ascertain whether heat played a part.
A
But what headline number do they report?
B
So the CDC reports that heat plays a part in around 2,300 deaths a year in the US.
A
And has anybody in the US tried to measure excess deaths during hot periods?
B
So the CDC is apparently currently working on a methodology for that. A group of researchers have also looked at data from the late 90s to early 2000s, and they estimate that the number of excess deaths during hot periods were more like 10,000 a year.
A
Overall, it sounds as though the European figures might be over counting heat deaths and the US ones are under counting.
B
Yep. But at this stage, it might be time to introduce some cool and refreshing context. Ultimately, this is a policy argument, sometimes childish and sometimes thoughtful, about whether Europe should just get serious about installing more air conditioning.
A
It does make sense that air conditioning could be beneficial, even though if you look at excess mortality in the us, it still goes up during heat waves.
B
Yeah, absolutely. I think most scientists would agree that aircon is beneficial to keep a constant temperature, so people aren't subjected to sudden extremes they don't have time to acclimatise to. It's this jump in temperature that causes the body to become overstressed and it's the stress that often exacerbates existing conditions. However, there are many countries, for example Italy, which have widespread AC use but large excess mortality during heat waves. So it's definitely not a silver bullet.
A
Thank you, Lizzie. You, our dear loyal listeners, are often inspired to get in touch by the most worthy of causes, disinformation, democratic accountability, the fiscal sustainability of the state itself, and eggs. Not just any old eggs, white eggs. A surprising number of you emailed more or less@BBC.co.uk to ask us to look into a recent decision by the supermarket chain Sainsbury's to stop selling brown eggs in its own brand ranges. They are switching to white eggs only. The justification for this, they say, is that white eggs have a 12.7% lower carbon footprint. Can this really be true? You asked what is it about the colour of an egg that gives it a smaller carbon footprint? To be honest, we are not really eggmen, so we are someone who is.
C
My name's Gary Ford. I'm a policy advisor at the British Free Range Egg Producers Association. I'm comfortable being referred to as the
A
eggman, which makes me the walrus. That's fine.
C
So, eggs, essentially, the only difference between white eggs and brown eggs is the shell colour. The contents are the same.
A
Eggs is eggs, essentially. However, in terms of the colour, there is an important
C
earth it's simple from the colour of the hen laying, the egg point of view, because white layers lay white eggs and brown layers or brown coloured birds lay brown eggs.
A
Fancy that. Every day's a school day on Radio 4. Now for the commercial breeds used in industrial scale egg production. White hens lay white eggs and brown hens lay brown. Gary says There are about 40 million laying hens in the UK, about 80% of which are brown commercial breeds. We're a brown egg nation, although it hasn't always been that way.
C
As recent as the 1970s and before, we were predominantly white egg consumers and it was very unusual to see a brown egg.
A
Brown egg dominance is not a universal phenomenon. In the us, Asia and some Europe, white egg rules supreme. But in the UK, brown eggs were
C
seen as a premium in the 1970s into the 1980s. So supply pivoted to meet that increasing consumer demand for brown eggs and hence where we are today. But over the last few years we have seen a significant increase, typically 5% year on year growth in the number of white layers in the national laying flock.
A
Okay, let's get to the carbon footprint. Eggs have a low carbon footprint compared to other forms of animal protein, and most of it is determined by the food that the chickens eat.
C
80% of the carbon footprint of an egg is made up of the feed that we give the hen. And the big driver of the carbon content is soya. That's a protein, it's a soya bean grown in North America, South America, China, and that is the main driver to the carbon footprint.
A
The carbon footprint of the soya itself massively depends on the land that the soy is grown on and what it replaced. For example, if, if the soya crop is replacing rainforest, the carbon footprint of that soya is high. However, these calculations are subject to debate and recalculation, and they change all the time. As for the rest of the carbon footprint of an egg, a big component is the rearing of the young bird to egg laying age. If a young bird goes on to lay more eggs over its productive lifetime and does so eating less food, then it will likely have a lower carbon footprint. And that calculation is where the white hens have an edge.
C
Generally, they will be eating slightly less than a brown laying hen, and that's partly a reflection of the fact that it's a slightly smaller bird laying a slightly smaller egg. So a smaller animal takes less feed, they produce slightly more eggs over a given time frame compared to a brown layer, just slightly different, slightly up. But the main difference is that the White layer will stay in lay longer.
A
Now, again, all these numbers are variable and they're based on the length of time that the hens lay at a commercially viable rate. They do keep on laying afterwards, but
D
at a slower rate, which means that
A
they are unlikely to be given the chance. Once you've got the numbers for all of this, you stick them into a carbon calculator to work out the egg carbon footprint. And there's another issue here.
C
There's a plethora of carbon calculators out there, all giving different results.
A
A recent review found that one carbon calculator might produce results three times higher than another. All of this means that the exact difference in the carbon footprint of a kilogram of white eggs compared to a kilogram of brown eggs has a pretty wide range. Another analysis with different hen farmers found white eggs had a carbon footprint between 6 to 8% lower, instead of 12.7%. But the bottom line is, by industry standards, white eggs probably do have a lower carbon footprint. Although I'm not sure I would take any particular number as gospel.
C
The fact remains that a white laying hen will lay more eggs over its lifetime, so it's more productive. The more productive, the more output. By definition, all things being equal, the lower the carbon footprint.
A
One last thing, the eggs from caged chickens also have a lower carbon footprint than those from free range birds. And more productive chickens are also likely to produce cheaper eggs. So it might be that the carbon footprint is not the measure of all things. Our thanks to eggman Gary Ford on More or Less. We quite often look at the numbers behind political debates, but today we're going to look at the numbers behind, well, literally a political debate, and namely Prime Minister's questions. It's the highlight of the political week when MPs get the opportunity to ask the Prime Minister a question. They'll be hoping to help their constituents and perhaps snatch a bit of the spotlight too. An odd thing happened at last week's PMQs. No MP from the official opposition, other than Conservative leader Kemi Badenoch, of course, stood up to question Keir Starmer. The Labour mp, Josh Fenton Glyn tweeted that it says all you need to know about lack of confidence in the
E
Tory Party, that not a single Conservative
A
MP is on the order paper for today's PMQs. How else could you explain that none of them have asked a question? Well, that's not actually difficult. As it happens, it's the luck of the draw. But the question we want to ask is just how lucky or unlucky was it? Every backbench MP who wants to ask a question applies for a place in what's called the shuffle. And 15 MPs are chosen at random from those who apply the shuffle cares not for your party allegiance. As one Twitter respondent put, Probability has
F
been on the UK maths curriculum for many years.
A
How did you miss it? Mr. Fenton Glynn wasn't going to take that kind of comment lying down. He responded, I'm aware of probability. You seem confused. And at this point we got interested not in the argument, but in the numbers. Now, normally we don't get to find out how many MPs applied for the shuffle from each party, but Josh Fenton Glynn also posted this information on Twitter. 47 Conservative MPs applied for the shuffle out of a total of 311 shufflees. We'd like to know the probability that when 47 out of 311 people apply for 15 places, none of them are picked. Here to help is Jen Fisser Rogers, a statistician and chief scientific officer at Coronado Research.
G
So you got 311 questions in total, and we're told that 47 of them were from this conservative group, meaning that 264 of them were not in that Conservative group. So we'll call those Conservative and not Conservative. And then we know that 15 questions are chosen at random. So you could think about the first question. What is the probability that the first question drawn is not conservative? Well, there's 264 not conservatives from a total of 311 questions. So the probability is calculated by doing 264 divided by 311, and that is about 84.9%.
A
So at 85%, it is pretty likely that the first pick in the shuffle won't contain a Conservative. Next, you repeat the process, but removing the one non Conservative MP who's been selected. There are now 263 non conservatives in the shuffle out of a total of 310 MPs.
G
So the probability that the second question is not conservative is then calculated as 263 divided by 310, and that's about 84.8%. So nearly the same, but just a little bit smaller. But what we need to calculate is the probability that the first and the second question are from this not Conservative group. And to do that, we have to multiply those two probabilities together, which gives
A
us 72%, still reasonably likely to complete the calculation. Just keep going, producing a big string of probabilities to multiply together.
G
And you do that 15 times all the way to 250 divided by 297. And if you do that big sum, it ends up being a probability of around 8%.
A
There was an 8% chance that none of the 47 Conservative MPs would be selected to ask a question at PMQs.
G
So that's about one in every 12. But we might expect to see this happening. I don't think that it's an unreasonable probability. 8% 1 in 12 seems pretty likely that it would happen every so often, and I think this is sometimes as well one of the things you have to think about with probabilities. Even things with small probabilities happen if you give them the opportunity to happen often enough. And PMQS happens pretty regularly. So an 8% chance of this happening. Should we then be surprised that it happens every now and then? No, probably not.
A
The maths of probability tell you all you need to know about the absence of a Conservative backbencher at PMQS in any given week. It's a bit unlikely, but over the course of a parliamentary session, it's not remotely surprising that it would happen. Our thanks to Jen Fisser Rogers. Loyal listeners will have noticed a Welsh
D
theme to this series.
A
We've explored literacy in Welsh schools not once but twice. We've answered a loyal listener's question about how many football pitches would fit into Wales. 3 million. And I had some absolutely delicious Welsh rarebit last week, in case you were wondering. But maybe you weren't. Maybe instead you got stuck on that penultimate answer. Maybe you heard 3 million football pitches to cover Wales and you disagreed and maybe then you would not let it lie. Did you measure the area of a
D
pitch up the size of a mountain,
E
or did you measure using the footprint of it?
D
I assume it would be different for valleys.
A
I was told that if you flattened out Wales, it would be the same size as England, therefore raising the number of football pitches.
E
Discuss. When I was very little, my primary school teacher, Mrs. Richards, had a mug that said, if all of the hills and mountains in Wales were flattened, Wales would be bigger than England. I've thought this was true since about 1986. Could you please fact check it for me?
A
Hmm, that's a familiar voice there. It's an interesting question. In any case, the area we gave for Wales didn't take into account that proud country's majestic valleys and mountains. It essentially assumed Wales was two dimensional. If that seems rude in our defence, this is a standard approach when it comes to defining the area of a country and the figure came from the Office for National Statistics. Nevertheless, it is a fascinating. Would the surface area of Wales change dramatically if you ironed out all those mountains? And we here at More or Less love nothing better than taking a question that would be great fun to debate in a pub and asking a serious academic to spend their time answering it. The academic in question, Dr. Laura Graham, a computational and spatial ecologist at the University of Birmingham. We asked her to run the numbers for us.
D
Now, my understanding is that when we calculate the surface area of countries, we sort of pretend that they are like those salt flats in Utah, that there's no bumps or hills anywhere.
A
Why do we do that? Is that just laziness?
H
Maybe it's laziness, but I also wonder if it's much more to do with the fact that you can't actually stretch out land. Like, for example, we'll calculate, say, surface area of an organ or something and say, this stretches out to the distance to the moon or something. Whereas actually, with land, you couldn't physically stretch it out.
D
Wales is pointier than some places, it's pointier than England. If you wanted to take into account all of those points, or as the Welsh call them, mountains, what would you do?
H
Okay, so essentially what we need to do is we need to get a kind of understanding of what the elevation is in the area that you're measuring the surface of. Now, you could go out with a tape measure and try and make some calculations that way, but it would take a very long time, particularly if you think about all the mountains in Wales. So the easier way of doing it would be what's called a digital elevation model.
D
I mean, spoiler, you have, in fact done this. So how did you do it?
H
So I've taken this digital elevation model and what it does is it calculates for every 90 meter by 90 meter square within the whole of the UK. It tells you what the average elevation in that square is.
D
Right.
H
So then what I've done is I've used a function which essentially calculates lots of different triangles between the midpoint of each of the squares to actually get an estimate of the surface area, which uses a little bit of trigonometry to do that.
D
I'm imagining in my head a slightly old school, maybe 1990s computer game. So when they were kind of impressive enough that you would have 3D, but the computer wasn't nearly powerful enough to make it look remotely realistic. So you'd have these slightly blocky landscapes and very obvious triangles. And it sounds as though this is what Your representation is actually doing.
H
Yeah, absolutely. Exactly like that.
D
So the area of Wales flat is, I am told, 20,847 square kilometers. Nearly 21,000 square kilometers.
G
1.
D
Once you apply your method and you measure all of those triangles, what is the answer that you get?
H
So the answer that you get is 21,144.
D
Okay, so it's up a bit from just under 21,000 to just over 21,000. Interesting.
H
So that is a 1.42 percentage increase in surface area once you've accounted for topography.
D
1.42% is not that much.
H
No, it doesn't sound like absolutely loads, but if you compare that to the percentage increase you get for England, you only get a percentage change of 0.37%.
D
One of our loyal listeners, Terry Hill, specifically asked, is Wales the same size as England if you flatten it out,
H
unfortunately, I'm going to have to confirm that it would not be the same size as the England. If you flattened it out, it's actually still a lot smaller.
D
And just based on everything you've said, my understanding is that the size of whales, if you take into account all the topography, is about the size of whales.
H
Yeah, yeah, pretty much.
D
I'm glad we have expressed this in units that make sense to all more or less listeners.
A
There is a twist, however, something called spatial resolution. Laura measured the surface area of whales in 90 by 90 meter chunks, but what if instead we measured the surface area looking at the country with a Finer resolution, say 1 meter by 1 meter chunks? We would then pick up more bumps, smaller hills and valleys that were missed by the 90 by 90 average and the measured surface area would be even greater.
H
Yeah, essentially it's a fractal geometry problem. So the more precise the data you have, the bigger the estimate's going to be.
D
Yeah, it just gets crinklier and crinklier the closer you look. Absolutely.
A
Of course, this affects the surface area for England too. It is unlikely that a finer measurement
D
would tip the scales in Wales favour. So, effectively we're saying that the. The answer to this question depends on the resolution with which you're modelling the topography of whales. And that will always be true.
H
Yeah, that will always be true.
A
Our thanks to Dr. Laura Graham, Computational and spatial ecologist at the University of Birmingham. Ah, I've just remembered who it was that asked the question earlier. Comedian Ellis James, co host of the Ellis and john podcast on 5Live. But we didn't get him into the studio just to record a question. Stay tuned next week for the next in our unplanned series of Welsh blockbuster items.
E
So I went to watch Wales play Lichtenstein in the football about six months ago. I was in the sort of square in the city having a pint after the game and a man came up to me and said, you, grandfather sold my dad his first tractor in 1979. And I thought, that is Cymru connecting in a nutshell.
A
And what does that have to do with a Current affairs statistics show it all will be revealed. That's all we have time for this week, but please keep your questions and comments coming in to more or lessbc.co.uk. we will be back next time and until then, goodbye.
F
Hi there, I'm Dilly Carter and this is everything you need to know about my new podcast, Sort yout Life Out Unpacked. I interview a different celebrity every episode. They bring me in three items from their home that reveal the most about them and we unpack the stories behind those items and I give you a few tips and tricks along the way. Some of the guests that I'm going to be interviewing are TV presenters like Lorraine Kelly, reality stars like Kerry Katona, podcast royalty like Elizabeth Day, and of course our very own Stacey Solomon. Oh, and let's not forget some incredibly funny comedians like Phil Wang and Eddie Caddy. I think as with everything, sort your life out based, you are going to get so much motivation, inspiration and ideas for your own home. Sort yout Life Out Unpacked presented by me, Diddy Carter. You can watch us on iplayer and listen on BBC Sounds.
This engaging episode of More or Less explores the accuracy and meaning behind viral claims comparing heat-related deaths in Europe and the US. Host Tim Harford and expert guests dissect statistical methodologies, discuss whether air conditioning is a silver bullet, debunk social media “hot takes,” and tackle other curious listener questions—ranging from the environmental impact of egg color to whether Wales is bigger than England if you “flatten all the hills.”
(00:00 – 06:58)
Viral Claim Overview:
Comparability Issues:
How Excess Deaths Work in Europe:
Problems with US Numbers:
Conclusion:
(06:09 – 06:58)
(08:01 – 13:33)
(14:46 – 19:01)
(19:23 – 26:28)
On Data Comparisons
On Excess Deaths and 'Harvesting'
On Probability at PMQs
On Fractal Geometry and Area
| MM:SS | Segment/Topic | |--------|---------------------------------------------------------------------| | 00:00 | Introduction, overview of episode topics | | 00:55 | Viral claim: European vs US heat and gun deaths | | 02:28 | Why US and EU heat-death stats aren't comparable | | 03:17 | How the 67,000 EU figure was reached | | 04:05 | 'Harvesting' and vulnerability explanation | | 05:35 | Official US estimated heat deaths | | 06:09 | Over/undercounting and policy debate (air conditioning) | | 08:01 | White vs brown eggs carbon footprint | | 10:19 | Feed, soya, and environmental impact | | 11:34 | White hens' higher productivity reduces carbon footprint | | 13:17 | Summary: White eggs probably lower-carbon, but difficult to quantify | | 14:46 | Unusual PMQs event and statistical analysis | | 18:28 | Probability in politics: unlikely but not shocking | | 19:23 | Listener myth: Flattening Wales, area vs England | | 24:10 | How much Wales’ surface area increases when accounting for hills | | 26:01 | Fractal geometry, surface area, resolution | | 26:28 | Summary: Wales still smaller than England, even with hills |
For more or to ask your own burning statistical question, listen to More or Less from BBC Radio 4.