Do free school meals work?

Summary

Podcast Summary: "Do Free School Meals Work?"

Podcast: More or Less (BBC Radio 4)
Host: Tim Harford
Guests: Ellen Greaves (Institute for Fiscal Studies), Professor Tricia Greenhalgh (Bart’s Hospital, co-author on free school meals research), Roger Smith (Formula One author), Professor Kevin McConway (Open University), Professor David Spiegelhalter (Cambridge)
Date: September 20, 2013

Overview

This episode of "More or Less" investigates high-profile statistical claims in the news and policy world—most centrally, whether providing universal free school meals in England actually works. The program examines evidence from recent pilot schemes, debates over environmental metrics like the "1,000-year plastic bag," and the risks/benefits of cycling versus driving, all with the show's signature critical wit.

---

1. The Effectiveness of Free School Meals

Segment: [00:24]–[07:16]

Key Discussion Points

• Policy Background:
  The UK government announced universal free school meals for all Key Stage 1 (first three years of primary school) children in England. Previously, fewer than one in five children (those from families on certain benefits) received free meals. This decision was triggered by pilots in Newham and Durham (2009–2011), which reported academic benefits, while a Wolverhampton pilot that broadened eligibility didn’t see similar results.

• Pilot Results & Behavioural Changes

  • Dietary change:
    “The type of food that children ate at lunchtime changed…less crisps, actually less whole pieces of fruit, and more foods associated with hot lunches.” —Ellen Greaves [01:12]
  • Substitution effect:
    Children didn’t eat more overall; they sometimes shifted where/when they ate snacks like crisps and fruit.
    “There was some degree of substitution, so overall levels of consumption didn’t change dramatically.” —Ellen Greaves [01:38]
  • Behavioural impact:
    No clear evidence of improved behaviour at home; teachers’ subjective impressions varied.
    “Some teachers felt there was an increase in concentration…others disagreed and actually thought they were more lethargic after their school lunch.” —Ellen Greaves [01:57]

• Academic Performance

  • Results showed positive, significant improvement in academic attainment at both Key Stage 1 and 2, with the largest impact found at Key Stage 2.
    “There was an increase in attainment at Key Stage 1 and at Key Stage 2 by a significant amount compared to the control areas.” —Ellen Greaves [02:10]
  • The most improvement was seen in pupils previously eligible for free school meals, but spillover effects on classmates are possible.
    “Those that were eligible for free school meals prior to the pilot did better…there are spillovers between pupils in a classroom… it may be that increases for one group affect attainment for the other.” —Ellen Greaves [03:40]

• Unanswered Questions
  Despite clear academic gains, the precise reasons remain “slightly unclear”:
  “What’s slightly puzzling is that we didn’t see an increase in attendance... So the mechanism through which the free school meals improved attainment is slightly unclear.” —Ellen Greaves [03:15]

• External Review and the Hawthorne Effect
  Professor Tricia Greenhalgh emphasized the complexity of the intervention and noted the pilots’ multi-faceted approach (improving kitchens, training, community promotion).
  She also raised the Hawthorne Effect:
  “The Hawthorne effect is that people change their behaviour when they're being studied...They feel special so they perform better.” —Tricia Greenhalgh [05:09]

• Will This Work Nationally?
  Greenhalgh expressed doubts about pilot success repeating at scale:
  “I'd be very surprised if we see the same level of improvement as happened in the pilot sites…When things get implemented in real life…nobody’s feeling particularly special or particularly observed.” —Tricia Greenhalgh [06:45]
  The need for further research and careful monitoring was strongly emphasized.

---

2. Plastic Bag Environmental Myths

Segment: [07:16]–[13:58]

Key Discussion Points

• Origins of "1,000-Year" Plastic Bag Figure

  • A government claim stated plastic bags take “a thousand years to degrade,” but scientific evidence for this is lacking.
  • “I would never make a statement like that because you could never prove it. Like, who's going to do an experiment for a thousand years?” —Professor Ming Tian [09:34]
  • No scientific paper or author could confirm the "1,000 years" figure.

• Actual Estimates & Environmental Impact

  • “Typical estimates range from several decades to centuries.” —Government statement, paraphrased by Tim Harford [10:22]
  • The compound used in common plastic bags is different from the PET in soft drink bottles often cited in studies.

• Relative Environmental Costs

  • A cotton bag must be reused 130 times, “bags for life” 4 times, and a paper bag 3 times to match the climate change impact of a single-use plastic bag (“if you don’t reuse the plastic bag”).
    —Terry Coleman, quoting a 2005 Environment Agency report [12:07]
  • “A litre of petrol weighs as much as 50 to 100 plastic bags...That should give you a rough sense of how much fossil fuel goes into making your bags versus powering your car.” —Tim Harford [13:21]

• Takeaway:
  The 1,000-year claim is unprovable; the best evidence suggests wide variability in bag lifespan. The main environmental issue may be litter and wildlife harm, not just longevity in landfills.

---

3. Formula One and Statistical Risk

Segment: [13:58]–[19:32]

Key Discussion Points

• Challenging the 20% Per-Race Fatality Claim

  • In the 1970s, Formula One was indeed dangerous, but not to the extent dramatized in popular media (“20% chance of dying in every race”—as quoted in the film Rush).
    “If you race 10 times a year, very few drivers would make it through a season.” —Roger Smith [15:26]

• Real Fatality Rates

  • For the decade before 1976:
    • Per race: 0.33% risk of death
    • Over a season (multiple races): 4.4% risk
    • Over a 5-year career: About a 20% risk
      “If a driver raced for five years in every race, his chance of death did indeed approach 20%.” —Kevin McConway [16:13]
  • Sir Jackie Stewart’s famous “two out of three” chance was exaggerated, but over six years, the real risk was about 50% if racing in all forms of motorsport.

• Contextualizing Motorsport Dangers

  • Deaths often occurred in non-Formula One events by F1 drivers (e.g., Jim Clark).
  • Risks have drastically reduced; the last F1 driver fatality occurred in 1994.

---

4. Cycling: Does the Risk Outweigh the Benefit?

Segment: [19:32]–[26:19]

Key Discussion Points

• Health Benefits

  • Cycling is highly beneficial, especially for individuals who are not otherwise active.
    “If you're a complete slob and…the first exercise you do is to get on a bike…it has been shown to be fantastically effective.” —David Spiegelhalter [21:41]
  • On average, “An hour's cycling…will put an hour on your life, which is what you spend cycling. It's almost as if time stands still while you're cycling.” —David Spiegelhalter [22:51]

• Risks

  • 100 cycling deaths per year in the UK (about two per week).

  • “That's about one cyclist being killed every two and a half million—three million hours of cycling. About 300 years of continuous cycling before you expect a fatality.” —David Spiegelhalter [23:35]

  • Comparative risk:

    • 1 hour cycling ≈ 2 hours walking, 5 hours in a car, or 5 minutes on a motorbike [23:58]

  • Pollution represents a mild health risk, generally offset by the physical health benefits.
    “Having to breathe in some muck in an hour, cycling probably takes a few minutes off my life.” —David Spiegelhalter [24:28]

• Risk/Benefit Ratio

  • “The risk of getting squashed per hour is roughly equivalent to taking about 12 minutes off your life expectancy compared with the hour or more that…you might gain from the cycling.” —David Spiegelhalter [25:30]
  • “The benefit cost ratio looks like, I reckon it's between five and ten times the benefit rather than the cost.” —David Spiegelhalter [25:37]

---

5. Notable Quotes & Memorable Moments

• “What's slightly puzzling is that we didn’t see an increase in attendance... So the mechanism through which the free school meals improved attainment is slightly unclear.” —Ellen Greaves [03:15]
• “It could be that they were better nourished… On the other hand, it could be something called a Hawthorne effect.” —Tricia Greenhalgh [05:18]
• “I would never make a statement like that because you could never prove it…Who’s going to do an experiment for a thousand years?” —Professor Ming Tian [09:34]
• “If a driver raced for five years in every race, his chance of death did indeed approach 20%.” —Kevin McConway [16:13]
• “An hour's cycling…will put an hour on your life, which is what you spend cycling. It's almost as if time stands still while you're cycling.” —David Spiegelhalter [22:51]
• “The risk of getting squashed per hour is roughly equivalent to taking about 12 minutes off your life expectancy compared with the hour or more that…you might gain from the cycling.” —David Spiegelhalter [25:30]

---

6. Timestamps for Important Segments

• [00:24] Start of free school meals policy discussion
• [01:12] Ellen Greaves on changes in children’s lunches
• [02:10] Academic attainment results of pilot
• [03:15] Discussion of unclear mechanism for attainment gains
• [05:09] Tricia Greenhalgh introduces Hawthorne Effect
• [06:45] Greenhalgh on limits of pilot generalizability
• [07:16] Start of plastic bag degradation segment
• [09:34] Ming Tian on “1,000-year” myth
• [12:07] Environment Agency data on bag use comparisons
• [13:58] Debunking Formula One fatality statistics
• [16:13] Kevin McConway calculates realistic risks
• [19:32] Do cycling risks outweigh benefits?
• [21:41] Spiegelhalter on health benefits of cycling
• [23:35] Actual fatality risks while cycling
• [25:30] Final cost-benefit of cycling, summary

---

Conclusion

This episode of "More or Less" delivers an engaging, myth-busting discussion on school meal programs, environmental claims, and statistical risks in everyday life. The show underscores the importance of critically examining statistics—whether in policy decisions, pop culture, or personal choices—highlighting that nuance, context, and honest uncertainty are vital in understanding what numbers really mean.

Transcript

[Tim Harford]

Thank you for downloading More or Less from the BBC. This is the version of the programme first broadcast on BBC Radio 4. Here's Tim Harford.

Hello and welcome to More or Less the programme which takes the Grand Prix for statistical good sense this week. Bikes and very fast cars. I accept. Every time I get in my car,

[David Spiegelhalter]

there's 20% chance I could die.

[Tim Harford]

But first, the Deputy Prime Minister Nick Clegg proudly announced this week that free school meals will be provided in England, all children in the first three years of school. At the moment, fewer than one in five children are eligible and registered to get those meals, generally because their family is receiving benefit payments for being out of work. This follows a recent pilot conducted in Newham and Durham between 2009 and 2011, which seemed to show that giving free food to primary school children was good for their academic performance. Meanwhile, a parallel pilot in Wolverhampton, which expanded the free school meals program merely to a broader subset of pupils, seemed to show no impact. Ellen Greaves, a researcher at the Institute for Fiscal Studies, was one of the people responsible for evaluating these pilot results.

[Ellen Greaves]

We're able to tell that the type of food that children ate at lunchtime changed with, as you'd expect, less foods associated with packed lunches being eaten, less crisps, actually less whole pieces of fruit and more foods associated with hot lunches. We're not able to tell whether behaviour improved, but we know that parents reports of behaviour didn't change significantly.

[Tim Harford]

And do we know whether the children were just eating crisps and maybe fruit at home? Were they substituting?

[Ellen Greaves]

There was some degree of substitution, so overall levels of consumption didn't change dramatically.

[Tim Harford]

So the children ate the sort of things you might expect and nobody noticed a difference in their behaviour.

[Ellen Greaves]

Nobody at home noticed a difference in their behaviour. It's not clear whether their behaviour in school improved.

[Tim Harford]

Why is it not clear?

[Ellen Greaves]

There was some evidence from qualitative research that some teachers felt there was an increase in concentration their behaviour in class, whereas other teachers disagreed and actually thought that sometimes they were more lethargic after their school lunch.

[Tim Harford]

What about results?

[Ellen Greaves]

Results in terms of attainment were positive. So there was an increase in attainment at Key Stage 1 and at Key Stage 2 by a significant amount compared to the control areas that we had.

[Tim Harford]

And I recently learned the lingo as a parenting of young children. Key stage one is infant school the first three years and key stage two is the rest of school.

[Ellen Greaves]

That's right. And the current proposals want to extend free school meals up to key stage one.

[Tim Harford]

Okay, now, Nick Clegg is very proud of the fact that he's introducing preschool meals for children the first three years of primary school. Key stage one. Is that where you actually saw the biggest improvements?

[Ellen Greaves]

The biggest improvements were actually at Key Stage 2, but the improvement at Key Stage 1 was also significant.

[Tim Harford]

OK, so it's not completely perverse, it's only slightly perverse to introduce it for key stage one and not.

[Ellen Greaves]

Yes, that's right. And maybe a motivation for introducing it at Key Stage 1 is that you encourage healthy behaviours.

[Tim Harford]

Right. So it seems to make sense. You give children better food and they're going to be healthier, their attendance will improve, they'll be more productive, they'll be able to concentrate more nutrition. So it's not a surprise that the food led to an improvement in performance.

[Ellen Greaves]

Yes, that's right. There are many reasons why you think performance would improve. What's slightly puzzling is that we didn't see an increase in attendance at. And as I said, there were mixed reports on children's behaviour in school. So the mechanism through which the free school meals improved attainment is slightly unclear.

[Tim Harford]

Who were the children who actually seemed to be benefiting most? Were they children who newly had access to the free school meals, or were they children who in principle should have got them all along?

[Ellen Greaves]

Those that were eligible for free school meals prior to the pilot did better than those that weren't eligible before. So you're right, it's strange. It can't just be the free school meal, it must be something else. And that could be partly because the meals are having an emergency, but partly if we think there are spillovers between pupils in a classroom, it may be that increases for one group affect attainment

[Tim Harford]

for the other, because certain children are not acting up or the teacher isn't having to spend so much time helping them get ahead.

[Ellen Greaves]

Yes, that's right.

[Tim Harford]

But you are just guessing.

[Ellen Greaves]

That is an informed guess. Yes.

[Tim Harford]

It seems slightly unsatisfactory. It's great that we've got this pilot, we've got this evidence, but there's something slightly worrying about the fact that we don't really seem to understand what actually happened or why.

[Ellen Greaves]

I think you're quite right. Ideally, more research would have been done to see whether providing universal free school meals just in these three years of school does have the same impact. Also testing what happens in more areas across the country. These were two quite disadvantaged pilot areas, and so it's unclear whether the same benefits would happen in more advantaged areas. And it's also, as I said, it would have been beneficial to survey more directly and see what was happening in schools. What could have brought about these changes.

[Tim Harford]

Ellen Greaves observing with admirable frankness that we don't quite understand what happened in this pilot. I also spoke to Tricia Greenholsh, who's a professor of primary healthcare at Bart's Hospital and co author of a research review of free school meals programs across the world. I asked her what she made of the study in question.

[Tricia Greenholsh]

The first thing we need to really be clear about is what was the intervention? The intervention was three things. It was free school dinners offered to all infants. The second thing was quite a bit of support provided for the school. For example, they went in and refurbished the kitchens if those needed refurbishing, they trained the staff, they explained to the staff what this was all about. And the third part of the intervention was support to and through the local authority to promote the scheme to local families. And my thoughts are, as the authors of the evaluation said, we don't know why the children in the intervention schools performed better on the tests. It could be that they were better nourished and it could be that having food in their stomachs made them much more alert in the classroom and focused on their work. On the other hand, it could be something called a Hawthorne effect. The Hawthorne effect is that people change their behaviour when they're being studied and that is for a number of different reasons, partly because they feel special so they perform better. You know, they've been picked and they sort of think subconsciously, I've been picked, therefore I must be special. And also they want to meet the researchers expectations. This is a subconscious thing that the researchers come in and say, we'd like to see whether you're, whether the children in this class are going to, are going to perform better and both the teacher and the pupils subconsciously try to perform better just because someone's studying it.

[Tim Harford]

As this policy is rolled out nationally, do you think we're going to see the same effect on academic performance?

[Tricia Greenholsh]

I'd be very surprised if we see the same level of improvement as happened in the pilot sites. And I think the reason for that is that the pilot sites may have been subject to the Hawthorne effect. I think when things get implemented in real life, in inverted commerce, one sees less of that kind of effect simply because nobody's feeling particularly special or particularly observed. But in the end we don't know. We are going to have to watch very carefully and monitor the data and review the whole policy in a couple of years time.

[Tim Harford]

Professor Tricia Greenholsh and by the way, for you field experiment nerds out there, the Hawthorne Effect is named after some experiments carried out at the Hawthorne Works near Chicago in the late 1920s. The effect, which is that people change their behaviour when experimenters are fussing around them, is perfectly real. But the story of what actually happened at Hawthorn has become something of an urban myth in social science research. Remind me to tell you all about it sometime. You can always email us at more or lessbc.co.uk. And now, here's a little test for you. What's the name of the Oscar winning movie from 1999 in which this soundtrack was used? Listen for a little longer. Yes, it was American Beauty. And can you remember what appears on the screen while this music is playing? It's a plastic bag. First invented in 1957 as a sandwich bag, it had become ubiquitous by the 1970s, replacing paper bags as the method for transporting our shopping. The character Ricky Fitz found it beautiful, the most beautiful thing he had ever filmed. Not everyone is quite so entranced. We all use plastic bags. I do. Everyone does. I think everybody feels that maybe we use too many plastic bags. Earlier this week, Deputy Prime Minister Nick Clegg. Yes, it's National Nick Clegg Week on more or less announced plans to introduce a 5p charge for plastic bags in England to discourage their use. You might use them for 20 minutes walking back from the shops, but they take a thousand years to degrade. A thousand years is a long time. King Canute hadn't yet ascended to the English throne a thousand years ago. Loyal listener Tom Caldwell was suspicious. From where did this figure of a thousand years come? Has anyone seen a thousand year old plastic bag? We've turned to professor of Biochemistry and Molecular biology Ming Tian at Pennsylvania State University. As an expert on biodegradation, we hoped he'd be able to point us to the source of this claim. But no, I would never make a statement like that because you could never prove it. Like, who's going to do an experiment for a thousand years? Yeah, we haven't had plastic bags for a thousand years, I suppose, and we certainly haven't been experimenting on them.

[Tricia Greenholsh]

Right.

[Tim Harford]

But the flip side of the coin is you make a statement like that also, nobody can prove you wrong. I don't want to say it's meaningless because I think behind all this is our concern about the environment. My point is, it all comes down to the environment they end up in. If there's no sunlight, there's no oxygen, there's no water. Things last for a Long time. They found hot dogs in landfills that are over 25 years old. We asked Nick Clegg's office to provide us with a source for the figure of 1,000 years. We received this statement from the government. Scientists agree that it takes a long time for plastic bags to degrade fully. The exact time will depend on the environmental conditions. Typical estimates range from several decades to centuries. That didn't really answer our question. So we went back and asked again and this is what we were told. There is extensive peer reviewed literature on the persistence of plastic materials in the environment. Below is a link. Plastic degradation and its environmental implications with special reference to Polyethylene Terephthalate Polymers Journal, Vol. 5, Issue 1, Pages 1 to 18, available at. Gosh, lots and lots and lots of referenciness. But when we'd roused ourselves and actually looked at the main paper they pointed us to, we found no mention of the figure of a thousand years. So we tracked down one of the authors of that paper, Dr. Hayden Webb from Swinburne University of Technology, who pointed out first, there's no fixed definition of what it means for plastic to degrade. Second, that different environments will have different effects on plastics, and third, that well, the study wasn't looking at the plastic that goes into bags, it was looking at the stuff in soft drink bottles, which is a totally different compound. So zero support for Mr. Clegg's thousand year plastic bag. Although Dr. Webb did think that discouraging the use of plastic bags was probably a good idea, loyal listener Sandra Gough wondered if Mr. Clegg wasn't missing the bigger picture.

Our bag's just an easy target. I heard cloth bags need to be used several hundred times to match one use of plastic ones.

The Environment Agency wondered too. In 2005 they looked into this issue. Terry Coleman was at the agency at the time and commissioned a report. If one takes a single use bag and looks at the comparison of its impacts on climate change from its production, and that includes extracting the oil in the ground all the way through its use and its final disposal, then in comparison, the paper bag that we studied would have to be reused three times to get below those greenhouse gas impacts. The bag for life would have to be reused four times and a cotton bag would have to be reused 130 times. And that's assuming there's no reuse of the single use lightweight carrier. The Waste and Resources Action Program. They published a study which said that probably three quarters of bags or more were reused. Terry Coleman, who these days Works for Environmental Resources Management, a sustainability consultancy. Terry Coleman was talking about the impact of bags on climate change. But of course, many people are concerned about plastic bags littering the environment, killing wildlife and so on. So biodegradability is important, but it's a double edged sword. Terry Coleman pointed out that if a plastic bag is in a landfill, you're better off if it doesn't degrade, releasing pollutants into the environment. My advice to you, if you're really worried about the environmental impact of your supermarket shop, pay attention to how you get to the supermarket. A litre of petrol weighs as much as 50 to 100 plastic bags and that should give you a very rough sense of how much fossil fuel goes into making your bags versus powering your car. Now, speaking of petrol heads, here's Hannah Barnes.

Hi, Tim.

I thought I told you to take a little break from more or less this week. Hannah, what are you doing here?

Well, I tried, Tim, I really did. I took on board everything you said about work, life, balance. And I went to the cinema. I fancied that new high octane film, Rush, featuring the incredibly handsome, I mean talented, Chris Hemsworth as British Formula one driver James Hunt and Daniel Bruhl as his Austrian rival, Niki Lauda. And it's set during the 1976 season and features plenty of this. And dramatic lines like this.

There's a lie that all drivers tell themselves.

[David Spiegelhalter]

Death is something that happens to other

[Tim Harford]

people, but also some questionable statistical claims.

Each year, two of us die. I accept every time I get in my car, there's 20% chance I could die.

Daniel Bruhl playing Formula One, great Niki Lauda there. And Niki Lauda was of course, horribly injured at the Nurburgring in 1976. So the sport was dangerous. Unbelievably, seat belts weren't even mandatory until 1972. Still a 20% chance of death every time you get into a car. That really bothered me. So much so that I went to the experts first. Formula One superfan and author of all the Races and analysing Formula One, Roger Smith.

Formula One is a dangerous sport and certainly back in the 1970s, it was a far more dangerous sport in terms of fatalities than it is today. But the particular statement seems slightly over. Egged over egged indeed. If you race 10 times a year, very few drivers would make it through a season.

Yeah, and I even managed to reach Niki Lauda himself.

You spoke to Niki Lauda?

Well, I got a message to his team, yeah. And they replied. He says it was A rhetorical statement rather than a precise statistical calculation.

Niki Lauder can be as rhetorical as he likes. In my book, he's thinking louder.

Yeah, I know, Tim, but I still wanted to know what the actual numbers were.

Well, fortunately for us, at more or less, we have a number of helpful people who can assist us when we need to crunch the numbers from time to time.

Yep. And I got onto our good friend at the Open University, professor of applied statistics, Kevin McConaughey, and the results of his work are quite shocking. If we take the decade leading up to the year the film is set. In 1976, drivers had a third of a percent chance of dying each time they competed in a Grand Prix, which includes any practice and qualifying sessions. Now, that's a lot less than 20% any time you get into a car. But across a Formula One season, the chance of death rises to 4.4%. And if a driver raced for five years in every race, his chance of death did indeed approach 20%. To put that into context, each race the risk of dying was equivalent to travelling one and a half million miles by car, or 500 skydives.

And we should say that Kevin McConaughey's had to make some assumptions to produce these figures. But he's Kevin McConaughey. If Niki Lauda can be rhetorical, Kevin McConaughey can make some assumptions. Now, I've heard another statistic that's even more shocking than the 20% chance of dying permanently.

Has Nick Clegg been talking about Formula One?

No. Leg this time it was actually Sir Jackie Stewart. He said he'd calculated that during the six years that represented the peak of his career, in the late 1960s and the early 70s, anyone continuously racing had a two out of three chance of dying.

Yeah, he did say that. Or variations on that theme, some of which are extremely implausible. For instance, that the chance of dying was two thirds per race. I asked Kevin McConaughey about the more sensible claim, and although Sir Jackie was racing slightly earlier than Niki Lauda, the chances of death were very similar, just fractionally higher.

So both of these claims about risk were wrong. But Formula One author Roger Smith says that Lauda and Stewart may have been thinking more broadly when they talk about Formula One drivers dying. There is a bit of a grey area as to the fact that there was a high incidence in motorsport in general. Some of those Formula One drivers who can be listed as the death of a Grand Prix driver actually were driving other forms of motorsport. Professional drivers in Those days had to have activity in a wider range of motorsport than they do today.

Now, I think it's probably fair to say that Sir Jackie Stewart was badly affected by the death of so many of his friends. There was a particularly awful period in 1968, for example, where four drivers died in the space of just three months, three of those in non Formula One cars. That included Stewart's great friend Jim Clark, who died in a Formula. Now, the statistical wizard Kevin McConaughey also tried to calculate these odds, though we couldn't give him a complete set of data. But he reckons that if you were a Formula One driver competing for six years at the height of the Jackie Stewart era, you had a 50, 50 chance of dying, not necessarily in a Formula One car. So Jackie wasn't far off.

Wow. So I thought we were thoroughly going to debunk this statistic, but actually, being a professional racing driver of 40 years ago really was unbelievably dangerous.

It was. And thankfully, that period is behind us. And there hasn't been a death of a Formula One driver since Ayrton Senna in 1994, although some race marshals have been killed.

Well, thank you, Hannah. And now let's examine the safety record of a quieter form of transport. One of our loyal listeners in Switzerland, Adam Jeff, emailed more or less bc.co.uk to pose a question. And since we're coming up to the first anniversary of when he emailed us, we thought it was time to answer him. Getting regular exercise is well known to be good for your health and would be expected to make me live longer. Is it possible to quantify this health benefit and subtract it from the risk of being killed while cycling? And if so, does the risk of cycling come out positive or negative? That is, will cycling on average, make my life longer? Shorter? 83% of the more or less team are cyclists. Hannah Barnes, of course, will only travel by Ferrari, and so we were curious to know the answer too. I have brought my folding bike to Paddington Station and I have also brought the Winton professor for the Public Understanding of Risk, David Spiegelhalter. David, you're a cyclist.

[David Spiegelhalter]

Yeah, I cycle to work every day in Cambridge, and when I'm in London, I use the Borrow Spike scheme a lot.

[Tim Harford]

Excellent. Well, I'll try and keep you safe on the roads, otherwise the irony meter would go through the roof.

[David Spiegelhalter]

Well, I'm glad we've done the risk

[Tim Harford]

assessment, the best informed risk assessment in BBC history. I'd like to ask you about the benefits of cycling first, but we should probably make a little bit of progress towards New Broadcasting House. So why don't you hop on your bike and be careful turning left.

[David Spiegelhalter]

Exactly.

[Tim Harford]

Ah, red light doesn't apply to cyclists, does it? Joke, joke. Left turning lorry. David, you steer clear of that. You steer clear of that. Yeah. Your helmet wouldn't do you much good if that hit you. Oh, David, that was all very pleasant. I can even see a tree. I'm feeling refreshed. I've had a few minutes of very gentle exercise. So tell us about the wonderful cardiovascular benefits of cycling.

[David Spiegelhalter]

Well, there's no doubt that it does the public health good if lots of people cycle. A common statistic is that perhaps a million hours of cycling will prevent six fatalities. But, you know, what does that mean to an individual person? So the way I quite like it being explained is that a claim that, for example, cycling for an hour, for example, you know, half an hour to work and half an hour back again would on average put three hours on your life expectancy. Now, I think that's a bit high, in fact, and also it depends very strongly on who you are, frankly. If you're doing a lot of exercise, then a bit more cycling isn't going to do you much good at all. If you're a complete slob and, you know, the first exercise you do is to get on a bike and start moving around has been shown to be fantastically effective. You know, hour and a half, two hours extra on your life for each hours cycling. And that's why a recent report for the Department of Transport says that the big benefits of getting people cycling are the 61% of the population who are either inactive or moderately inactive. Now, if you're a sort of an average person, it's quite in the middle. An hour's cycling, roughly speaking, will put an hour on your life, which is what you spend cycling. It's almost as if time stands still while you're cycling. You're not getting any older.

[Tim Harford]

It's like a teleporter.

[David Spiegelhalter]

It's great, which is fine, as long as you like cycling.

[Tim Harford]

Yes, there are the odd risks. We should probably talk about those, but let's go somewhere far more dangerous to do that. Lead on, David. It's a little busier here. We just crossed the Edgware Road. I have to say, I'm failing to feel any great sense of mortal peril. But people do often think of cycling as risky. The BBC did a risk assessment for us to even do this. They insisted that we wear helmets. They Insisted we wear high visibility jackets, which they would never have insisted if we'd been walking around London. Are people right to think that cycling is dangerous?

[David Spiegelhalter]

Well, about 100 people a year do get killed in this country. That's about two cycles cyclists a week. And that works out on about one cyclist being killed every two and a half million. Three million hours of cycling. That's about 300 years of continuous cycling before you expect a fatality. So there is some risk, but actually it's quite small. An hour of cycling is about equivalent to two hours walking, about five hours in a car in terms of your

[Tim Harford]

risk of being killed.

[David Spiegelhalter]

Overall risk of being killed or about five minutes on a motorbike.

[Tim Harford]

Okay, well, we're not far from Broadcasting House. Let's see if we can make it there in one piece.

[David Spiegelhalter]

Great.

[Tim Harford]

Cyclists coming the wrong way down a one way street towards me. David, we made it across London. We probably shouldn't be entirely surprised by that. Quite a lot of pollution as we were cycling through the streets. Is that a health risk?

[David Spiegelhalter]

Yeah, I mean, that's the bit I dislike most about cycling, cycling in London. But you know, if you look at the difference between say, living in Mexico City, a very polluted city, and London overall is about half an hour of your life each day. So I reckon, you know, really having to breathe in some muck in an hour. Cycling probably takes a few minutes off my life.

[Tim Harford]

But it's not a major risk compared with the health benefits.

[David Spiegelhalter]

Exactly.

[Tim Harford]

Now, one loyal listener did write into us and complained that a BBC story had had finished on the rather gloomy note that cycling deaths were up last year and reckon that wasn't really the right spin to be putting on things. But we've got to face facts, haven't we?

[David Spiegelhalter]

Well, yeah, they were up by about 10%. That is actually within chance variability. So you can't even say the risk has gone up. But if you look at the National Transport Survey, surprise, surprise, the number of miles cycled was also up 10%. So there's no evidence of cycling's getting more risky on average.

[Tim Harford]

So let's sum it up. There are hell risks from pollution and from getting squashed by a lorry. There are health benefits from getting more exercise. How do they compare against each other?

[David Spiegelhalter]

It's quite tricky because the acute risks are getting squashed. You know, to compare that against long term life expectancy, it's a bit of an all or nothing. It's a bit difficult. And when we did the BBC risk assessment, it only asked us about the risk of getting squashed. It didn't consider the benefits of us cycling here, so. But I reckon that the sort of risk of getting squashed per hour is roughly equivalent to taking about 12 minutes off your life expectancy compared with the hour or more that on average you might gain from the cycling. So the sort of the ratio, the benefit cost ratio looks like, I reckon it's about between five and ten times the benefit rather than the cost.

[Tim Harford]

Well, that's excellent. Well, with that thought, I'm going to take off my helmet, I'm going to take off my high vis jacket, I'm going to fold away my bike and I'm going to walk into Broadcasting House and engage in the far riskier activity of sitting down in an armchair and having a cup of coffee.

[David Spiegelhalter]

Yeah, I'd watch out if I were you.

[Tim Harford]

David Spiegelhalter, thank you very much.

[David Spiegelhalter]

Thank you.

[Tim Harford]

And loyal More or Less listeners may be interested to know that David Spiegelhalter teamed up with More or less his founding producer, Michael Blasland, to publish a recent book on the everyday risks we run. It's called the Norm Chronicles. And just before we go, great moments in pedantry. Last week, while sizing up the island of Great Britain, our island expert, Professor Godfrey Baldaccino, research Chair in Island Studies at the University of Prince Edward island in Canada and executive editor of the Island Studies Journal, ventured into the realm of cultural criticism. He pointed out that Britain was perhaps a little pompous, given that it was one of very few islands to call itself great. Loyal listener Andy King was one of many who wrote into comment the reason Great Britain is known as Great is a hangover from the Middle Ages, when the term was coined simply to distinguish the larger island of Britain from the smaller island of Brittany. The pomposity came later and that's all we have time for this week. If you've seen a statistical statement by Nick Clegg or anyone else that you'd like us to investigate, you can send us an email to more or lessbc.co.uk. and we're not just au fait with electronic mail. We have a website, BBC.co.uk more or less, where you can download the program for free. This episode of More or Less was brought to you by the number 784. Until next week, goodbye.

More or Less was presented by Tim Harford, the Financial Times undercover Economist. It was produced by Ruth Alexander and made in association with the Open University. You can Download many more Radio 4 programmes for free. Find [e-mail] Radio 4.