Drinks and drugs capital of the world? (WS)

Summary

Podcast Summary: More or Less – “Drinks and Drugs Capital of the World?”
Episode date: July 6, 2012
Host: Angela Saini (BBC Radio 4)

---

Episode Overview

This episode of More or Less tackles the statistical claims that the tiny island nation of Palau is the global capital of both beer drinking and cannabis consumption. The team investigates the sources and validity of these eye-catching statistics, exploring the peculiarities and pitfalls of data from small populations. In the second half, the show briefly explains what “five sigma” means in relation to the much-publicized discovery of the Higgs boson.

---

Key Discussion Points and Insights

1. Palau’s Supposed Status as Drinks and Drugs Capital

(00:12 – 06:20)

• The Claims in Question

  • Listener Dave Hurley spotlights statistics from The Economist: Palau reportedly has the highest per capita rate of cannabis use globally (nearly a quarter of adults, aged 15–64) and leads in beer consumption.
  • “Palau has the highest rate of cannabis use on Earth, with nearly a quarter of adults from 15 to 64 lighting up.” (00:29, Dave quoting The Economist)

• Scrutinizing the Beer Statistic

  • Reporter Charlotte MacDonald clarifies that, per 2005 WHO data, Palau did top the per capita beer consumption chart—but only for that year, which looks like an exceptional outlier.
  • Palauans drink comparatively little wine or spirits, so when total alcohol consumption is considered, Palau falls to 42nd place out of nearly 200 states; still high, but not world-leading.
  • The true overall leader in alcohol consumption is the Czech Republic.

  “The WHO compared data from 2005, and for some reason Palawans appear to have gone on a massive drinking spree that year...when you look at the total amount of alcohol consumed rather than just beer, Palau slips all the way down to 42nd place.” (01:18–02:07, Charlotte MacDonald)

• Doubt over the Cannabis Statistic

  • The UN’s World Drug Report estimated that about 25% of Palau adults use cannabis—a very high figure compared to the next highest countries (Italy, New Zealand at 15%).
  • However, the UN did not have direct adult survey data for Palau. Instead, they extrapolated from a high school pupil survey.
  • This approach is highly questionable given teenagers tend to use more cannabis than adults and Palau only has one state high school, making for a tiny, non-representative sample.

  “The UN weren’t able to obtain survey data for adults in Palau. So instead they used a survey of high school pupils and they extrapolated those results to get a figure for the whole adult population.” (02:53, Charlotte MacDonald)

• Sample Size Issues

  • 60% of the surveyed teenagers in the state high school had tried cannabis at least once; almost 40% had used it in the last month.
  • US comparison: Only 23% of American high school students reported cannabis use in the last month.

  “So not just a high school, but a really high school?” (03:39, Angela Saini, joking about the high numbers)

• Local Perspective and Concerns

  • Emery Wenty, Palau’s Director for the Ministry of Education, rejects the extrapolated figure as unrealistic.
  • He notes the island’s small, close-knit community and claims such high rates would be openly apparent.

  “Palau is a very small island. If that is the case, you see it, you smell it everywhere. You don’t. … It’s just inconceivable that it’s a quarter of the population.” (04:07, Emery Wenty)

  • Wenty points out the sample is just 565 teenagers; behavior likely differs in private schools and, especially, among adults.

  “We’re talking about 565 students that are teenagers who are at the age where they would try anything and they're at the age where they're influenced by their peers.” (04:49, Emery Wenty)

• UN’s Response: Limitations of Statistics in Small Populations

  • Angela May, a UN statistician, concedes the limitations. In small countries, a handful of people change behavior and statistics swing dramatically.
  • The UN is planning to collect more robust data in the Pacific to improve reliability.

  “In small countries, when you have just a small number of people that change behavior, that of course, is reflected in bigger changes in percentages. So we have to keep in mind when we read data, particularly for small countries.” (05:23, Angela May)
  “We are going to have a meeting in the Pacific Islands where we hope to collect more information and...reduce, if you want, the margin of error.” (06:04, Angela May)

---

2. Explaining "Five Sigma" and the Higgs Boson Result

(06:20 – 09:13)

• Coverage of CERN’s Higgs Boson Announcement

  • The show transitions to the science news headline: physicists reporting the likely discovery of the Higgs boson at a “five sigma” confidence level.

  “Both experiments, I can tell you, have reported results that are at this five sigma threshold. This is the level at which you have to reach.” (06:46, news soundbite)

• What Does "Sigma" Mean?

  • Sigma refers to standard deviation in statistics; the further from the average (mean), the higher the sigma.
  • 1 sigma off the mean in human height: a bit above average. 5 sigma: extremely rare; e.g., a man over 7ft tall.
  • In experimental science, a 5 sigma result means that the probability of seeing such an extreme result, if only random chance were operating, is about 1 in 3.5 million (1 in 3,486,914).
  • For scientists, 5 sigma is the threshold for claiming a discovery, but it is not an absolute guarantee of the underlying reality.

  “For the Higgs result to be five sigma, the probability of getting the same experimental data, if nothing more than random chance were operating, would have to be 1 in 3,486,914. This is the threshold at which physicists say they have a discovery.” (08:05, Angela Saini)

---

Notable Quotes & Memorable Moments

• “Is Palau really a paradise for beer-swilling stoners, or is there something about the tiny Pacific island that is making statisticians look like they've been smoking funny cigarettes?”
  — Dave Hurley (01:00)

• “So not just a high school, but a really high school?”
  — Angela Saini, punning on the cannabis usage statistic (03:39)

• “Palau is a very small island. If that is the case, you see it, you smell it everywhere. You don't. … It’s just inconceivable that it’s a quarter of the population.”
  — Emery Wenty, Palauan Ministry of Education (04:07)

• “In small countries, when you have just a small number of people that change behavior, that of course, is reflected in bigger changes in percentages.”
  — Angela May, UN statistician (05:23)

---

Key Timestamps

• 00:12 – Introduction and the Palau statistics question
• 01:18 – Charlotte MacDonald analyzes the beer drinking statistic
• 02:21 – Explaining the cannabis statistic and its origins
• 03:19 – Issues with sample size and data extrapolation
• 04:07 – Palauan Ministry of Education responds
• 05:23 – UN statistician on challenges of small-population statistics
• 06:20 – Segment shift: the meaning of “five sigma” in scientific discoveries
• 08:05 – Clarifying why “five sigma” isn’t absolute proof

---

Tone and Style

The show maintains its characteristic blend of accessible statistical analysis and dry humor, helping demystify complex or misleading numbers in the news. The hosts and experts are thoughtful, occasionally irreverent, and always focused on evidence.

---

Takeaways

• Seemingly sensational statistics (like “world leader in beer and cannabis use”) can be the result of small samples, data quirks, or poor extrapolation—especially in tiny populations like Palau’s.
• Whenever you encounter such claims, examine how the data was collected, who was surveyed, and whether results are being meaningfully compared.
• In science, statistical thresholds (like “five sigma”) indicate unlikely results under pure chance, but absolute certainty requires more than just one measurement; context always matters.

---

For more episodes and number investigations, visit More or Less on BBC.

Transcript

[A]

Thank you for downloading from the BBC. For details of our complete range of podcasts and our terms of use, go to bbcworldserveys.com podcasts hello.

[B]

For the next 10 minutes, you're with More or Less a dollop of statistical cream on the overripe strawberry of news. I'm Angela Saini. First up, more or Less Listener Dave Hurley was reading the Economist recently when he came across this surprising fact.

[C]

Palau has the highest rate of cannabis use on Earth, with nearly a quarter of adults from 15 to 64 lighting up.

[B]

Palau is a tiny Micronesian island home to just 21,000 people. But despite its size, this wasn't the first time David heard about it.

[C]

This comes after a who report in 2011 showing that Palauans drink more beer than any other nationality too.

[B]

Now, as Dave continues, this begs an important question.

[C]

Is Palau really a paradise for beer swilling stoners, or is there something about the tiny Pacific island that is making statisticians look like they've been smoking funny cigarettes?

[B]

Time to bring in our intrepid reporter, Charlotte MacDonald.

[D]

While both cannabis and alcohol figures Dave mentions come from highly respectable sources, let's take the beer statistic first. As Dave said, a recent World Health Organization report did indeed put Palau at the top of the global per capita chart of beer drinking nations. The thing is, that's not quite the whole story. The WHO compared data from 2005, and for some reason Palawans appear to have gone on a massive drinking spree that year. They just drank a lot more beer than usual in 2005. But there's another, more important point to make here. While they seem to drink a lot of beer in Palau, it looks like they don't drink much of anything else. Which means when you look at the total amount of alcohol consumed rather than just beer, Palau slips all the way down to 42nd place, although there are

[B]

nearly 200 states in the study, which means that 42nd place is still pretty high. Now, purely because I suspect some listeners will be wondering which country is at the top of the list for all alcohol consumption.

[D]

It's actually the Czech Republic. Now, cannabis the statistic which caught listener Dave's eye is that a quarter of Palawan adults are using it, which sounds like a lot. The Economist got the figure from the UN's World Drug Report, which has just been published. The two countries with the second highest cannabis usage are Italy and New Zealand. There, the report says, about 15% of adults are using it.

[B]

So not only are more people in Palau than anywhere else smoking cannabis. But they're ahead by a long, long way.

[D]

Exactly. But digging a little deeper into the figures, it turns out the UN weren't able to obtain survey data for adults in Palau. So instead they used a survey of high school pupils and they extrapolated those results to get a figure for the whole adult population.

[B]

Now, to me that sounds like somewhere an error could quite easily creep in. I mean, you would expect teenagers to smoke more dope than say, middle aged women.

[D]

Yeah, exactly, you would. And Palau is tiny, remember. In fact, it only has one state high school. So we're talking about a pretty small sample. But for what it's worth, the surveyors found that 60% of the teenage respondents in that state high school had used cannabis at least once and almost 40% said they had used cannabis in the last month.

[B]

So not just a high school, but a really high school?

[D]

Well, quite. But just for comparison, a similar survey of school kids is carried out in the US and there only 23% of students say they use cannabis in the last month.

[B]

So what do Palauans think about that statistic? That a quarter of the adult population has used cannabis?

[D]

Yeah, they're none too pleased about it. I put the statistics to Emery Wenty. He's a director for the Ministry of Education on Palau.

[E]

Palau is a very small island. If that is the case, you see it, you smell it everywhere. You don't. You don't really see it, you don't smell it. And like I said, this is a very small island. You sort of know just about everybody. It's just inconceivable that it's a quarter of the population is kind of.

[D]

Mr. Wenty told me that although there's only one state high school in Palau, there are some private schools which together have broadly similar number of students as the state school. He thinks cannabis use might be lower in the private schools. He also pointed out, as we have done, that the sample was small and that teenage behaviour might well differ from that of older adults.

[E]

I think this is a small group. We're talking about 565 students that are teenagers who are at the age where they would try anything and they're at the age where they're influenced by their peers.

[B]

Well, I put these points to Angela May, a UN statistician who worked on the recent drugs report. She accepts some of the criticisms and she also told me that the problem with surveying tiny populations like that of Palau is that a small number of people changing their behaviour can lead to huge swings in the statistics.

[F]

We have to keep in mind that it's a small country. So as all kind of statistics for small countries, the variability is much higher than in bigger countries. Where you see in small countries, when you have just a small number of people that change behavior, that of course, is reflected in bigger changes in percentages. So we have to keep in mind when we read data, particularly for small countries.

[B]

Unfortunately for Angela and her colleagues at the UN in Vienna, where average winter temperatures hover around zero, the only way to find out what's really going on in the South Pacific is to take a closer look.

[F]

We are going to have a meeting in the Pacific Islands where we hope to collect more information and also reduce, if you want the margin of error, to really better use all the data and information that they had there.

[B]

Angela May from the UN now, from small islands to small particles, and they don't get much smaller than the Higgs boson. Scientists running the world's largest particle accelerator at CERN this week were boggling us with numbers many of us didn't understand, but which apparently meant they had made a landmark discovery that would forever change physics and explain the mass of the universe.

[C]

Both experiments, I can tell you, have reported results that are at this five sigma threshold. This is the level at which you have to reach.

[B]

It's no doubt a big week for science, but what exactly do researchers mean when they talk about a five sigma result? Well, always on the lookout for a good statistical puzzle. We at more or less have been finding out. Press reports have made it sound like a measure of certainty that the Higgs boson has been found. So 1 Sigma would be low certainty and 5 Sigma high. In a way that's right, but in another, more exact way, it's totally wrong. The reason scientists say sigma is because that's the word used by statisticians when measuring how far something lies from the center of a bell shaped normal distribution curve. In nature, lots of things follow a normal distribution. Take the heights of adult men. Average height in the US is between 5 foot 7 to 6 foot 1. For someone to stand out as 1 sigma above normal, they would have to be more than 6 foot 1. For 3 sigma, they would have to be around 6 foot 7. And for 5 sigma, they would have to be more than 7ft tall. So in physics, as in life, if something is particularly unusual, it has a higher sigma value. For the Higgs result to be five sigma, the probability of getting the same experimental data, if nothing more than random chance were operating, would have to be 1 in 3,486,914. This is the threshold at which physicists say they have a discovery. But remember, this is not necessarily the same as a probability that they found the Higgs boson, only of seeing that particular set of data. For scientists to be sure that they saw what they think they saw will take many more years. That's your lot for this week. If you'd like to sample some vintage episodes of this program, head to the [e-mail] more or less. And if you have a number you'd like us to investigate, the email address is more or lessbc code. The program was presented by me, Angela Saini. The producer was Richard Knight, and the editor, Richard Varden. Next week, regular presenter Tim Harford is back after a curiously long holiday on the island of Palau. So from me, it's goodbye.

[A]

There are dozens of different podcasts now available from the BBC, including news, documentaries, science, business, arts and sport. For details of them all, go to BBC worldservice.com podcasts.