UK election podcast 4

Summary

More or Less – UK Election Podcast 4 (BBC Radio 4)

Date: May 6, 2015
Host(s): Tim Harford, Ruth Alexander, Eddie Maher
Main Theme:
A statistical deep-dive into the claims, debates, and figures shaping the 2015 UK general election, alongside a segment fact-checking a non-election statistic in the news.

---

Polls, Margins of Error, and Electoral Uncertainty

Timestamps: 00:25–03:44
Discussion Points:

• Why Poll Results Diverge:
  Ruth Alexander responds to a listener’s question about daily polls showing different results. She explains sampling, using the metaphor of "voter soup":

  "Taking an opinion poll is really like sampling a spoonful from this voter soup to see how the whole panful tastes. But if you skim the surface, you might just get croutons or you might end up with just a spoonful of mushrooms." (Ruth Alexander, 00:54)

• Sampling Error and Margin:
  Poll results differ due to sample size and selection; for a 1,000-person poll, the margin of error is ±3 percentage points.
• Current State of the Race:
  The BBC’s poll of polls has the Conservatives at 34% and Labour at 33% – virtually tied, well within margin of error.
• Rise of Minor Parties:
  Compared to 2010, the proportion considering voting for non-traditional parties (not Conservative, Labour, Liberal Democrats) is predicted to be at least double (from 12% to >24%).
• Regional Differences:
  "Taking a spoonful of UK voter soup is only of limited use when you've got such large regional and national differences as we do." (Ruth, 03:06)
  UK-wide polls underrepresent Scottish National Party (SNP) strength; in Scotland, SNP polls at 40–50%, vs. just 4% across the UK.

---

The Debate on Rent Caps and the UK Housing Crisis

Timestamps: 03:51–06:16
Key Points:

• Labour’s Proposals:
  Labour proposes three-year tenancies with rent rises capped by the Consumer Prices Index (CPI).
  About 4.5 million UK residents are private renters, equal to the number of social renters.
• The Demographics of Renting:
  Young people disproportionately rent: "Almost half of all young households aged 25 to 34 rent privately in England and Wales, and that is twice the rate it was a decade ago." (Ruth, 04:41)
• Supply and Affordability:
  A shortage of homes causes both house prices and rents to rise. Rents grew 2% UK-wide last year, higher in London; CPI was zero at the time, so Labour's policy would have frozen rents.
• International Comparison:
  Germany’s rent cap leads landlords to increase rents between contracts, raising concerns UK caps could have unintended effects.
• Root Issue:
  "Everyone agrees the big solution is to build more homes. Everyone also agrees that we shouldn’t build those homes near to them. And that is the UK housing affordability problem in a nutshell." (Ruth, 06:03)

---

Is This the Most Boring or Most Exciting Election?

Timestamps: 06:16–09:02
Statistics & Insights from Tim Harford:

• The Case for Tedium:
  TV leaders’ debates had much lower ratings than five years ago. First debate audience:
  2010 – 9.4 million
  2015 – 7.7 million
• But Also More Interesting Than Ever:
  1. More voters considering non-traditional parties—possibly double the number from the last election.
  2. "North of the border, this election looks like it’s going to be transformational," with SNP potentially reducing Labour’s Scottish seats from 41 to four or five. (Tim Harford, 07:38)
  3. High proportion of undecided voters: 40% say they might yet change their mind, compared to 10% in 1983 and 1987.
• Election Fatigue:
  The campaign lasted six weeks instead of the traditional four.

  "Each week of campaigning, if I can be forgiven a metaphor, it's like a donut... The first donut, enjoyable. The second one a bit much. Now we are through the fifth one, most of us are feeling sick." (Tim, 08:48)

---

Conservative Claim: “760,000 More Businesses” – Fact-Check

Timestamps: 09:06–11:53
Key Points:

• Claim:
  The Conservatives said 760,000 more businesses formed since 2010.
• Reality:
  The number is technically true; the UK has 5 million private sector businesses, up 760,000 in four years.
• The Nature of ‘Businesses’:
  Only 60,000 of the new entities employ anyone; the other 700,000 are sole traders—mostly self-employed.

  "The UK has gone completely crazy with Dragon's Den style entrepreneurial spirit... that's the message the Conservatives would like to sell. For certain, it's not necessarily what's going on." (Tim, 09:27)

• Self-Employment Trends:
  • Biggest increases among management consultants, bookkeepers, photographers, accountants.
  • Over 40% of self-employed are over 50, often following redundancy or early retirement.
  • A third of the increase is foreign-born workers (225,000).
  • Median income from self-employment fell between 2008 and 2013; many become self-employed as a better alternative to unemployment.
• Entrepreneurship Quality:
  Reports (e.g., Social Market Foundation) suggest the UK is not producing many “high-value entrepreneurs”, though statistics vary.

---

Maternal Mortality – The Belarus Statistic in Context

Timestamps: 11:57–16:22
Focus:
A fact-check of sensational headlines: “Pregnancy is six times riskier in the UK than in Belarus.”

• Headline Origins:
  The claim originates from a Save the Children report, using WHO statistics:

  • Belarus: 1 maternal death/100,000 live births
  • UK: 8 deaths/100,000
    Ruth questions the basis for such a direct comparison (“since when was Belarus the key place we compare ourselves with?” – 12:27)

• Statistical Differences:
  The UK is scrupulous about counting maternal deaths (using hospital staff, coroners, birth registration), making its rates look higher than countries relying only on death certificates.

• Why Belarus Looks So Good:
  The low number might be due to incomplete reporting or under-adjusting for how deaths are recorded. Numbers can fluctuate more in small countries; Belarus has only about 9m people.

• Who Should We Compare To?:
  Oxford University recommends comparing the UK only with similarly rigorous countries—France, Ireland, Netherlands. UK, France, and Ireland have identical rates; the Netherlands' slight difference may be a statistical anomaly.

• What Counts as Maternal Death?:
  Broader than just deaths in childbirth—also includes deaths up to six weeks after birth, though two-thirds of such deaths in the UK are from unrelated or exacerbated pre-existing health/mental health issues.

  “In most of all those cases in the UK, the women were already in ill health when they became pregnant. And the UK, it has to be said, has been making real progress on reducing the numbers who die as a direct result of their pregnancy.” (Ruth, 15:39)

• The Real Story:
  The true crisis is elsewhere, with sub-Saharan Africa facing maternal mortality rates 10 to 100 times higher than the UK.

---

Most Notable Quotes

• "Taking an opinion poll is really like sampling a spoonful from this voter soup to see how the whole panful tastes."
  — Ruth Alexander (00:54)

• "The big solution is to build more homes. Everyone also agrees that we shouldn’t build those homes near to them. And that is the UK housing affordability problem in a nutshell."
  — Ruth Alexander (06:03)

• "Each week of campaigning... it's like a donut. The first donut, enjoyable. The second one a bit much. Now we are through the fifth one, most of us are feeling sick."
  — Tim Harford (08:48)

• "For a lot of people, not everybody, self-employment is simply a better alternative to unemployment."
  — Tim Harford (11:11)

---

Summary Table of Important Segments

| Segment | Timestamp | Key Speaker | Highlights | |---------------------------------------|------------|------------------|-----------------------------------------------------------------| | Polls & margins of error | 00:25-03:44| Ruth Alexander | Sampling, why polls differ, rise of minor parties, SNP surge | | Rent caps & UK housing crisis | 03:51-06:16| Ruth Alexander | Labour’s proposal, renting trends, underlying housing shortage | | “Boring” or exciting election? | 06:16-09:02| Tim Harford | Debate ratings, rise of minor parties, SNP, undecided voters | | Conservative business claim checked | 09:06-11:53| Tim Harford | 760,000 new businesses: reality is a self-employment surge | | Maternal mortality in the news | 11:57-16:22| Ruth Alexander | Belarus–UK comparison challenged, how stats are collected matters|

---

Takeaways

• Election statistics are nuanced: Small differences in polling, big regional effects, and high volatility among the electorate make the 2015 UK election challenging to predict.
• Public policy claims benefit from close inspection: High-profile numbers (rent caps, business growth) often hide the real story in their details.
• Comparing countries demands context: Especially in areas like healthcare and mortality, measurement differences can create misleading headlines.

---

This episode exemplifies the More or Less approach: challenging headlines, unpacking statistical claims, and offering listeners both clarity and caution in the world of political and public data.

Transcript

[A]

Hello and welcome to More or less Election Podcast 4, a collection of interviews that were first broadcast on BBC Radio 4's PM programme with Eddie Maher, Tim Harford and Ruth Alexander have been taking it in turns to make sense of the numbers being used by politicians. On Friday, April 24,

[B]

Ruth Alexander from Radio 4's More or Less in is here. Ruth, you've been looking at the opinion polls.

[C]

Yes, that's because we had an email to our email address. More or lessbc.co.uk, if anyone else can be tempted from a listener called Martin who asked, we thought a rather good question. He said, every day we're getting polls. How come some say one party in the lead by two and some say the other party in the lead by two? Surely they should all be the same? Well, the reason they're not is because the pollsters aren't asking everybody in the UK how how they're going to vote, they're asking a representative sample. Often it's just about one or two thousand people. And probably the best way to think about this is to picture a nice pan of voter soup bubbling away on the hob, a metaphor I borrowed from a professor of Statistics, Kevin McConway, from the Open University, who is a keen, untold, but not always competent cook. Anyway, taking an opinion poll is really like sampling a spoonful from this voter soup to see how the whole pamphle tastes. But if you skim the surface, you know, you might just croutons or you might end up with just a spoonful of mushrooms.

[B]

Weren't mushrooms at the nhs?

[C]

No, they were when they're on the public spending pizza. But this is vota soup, which is quite different.

[B]

Please, not the pizza again.

[C]

No, let's not go there. So you might just get a spoonful of mushrooms and you might miss the chickpeas which have sunk to the bottom, perhaps. So it's important to stir your soup well and get a sample that includes all the ingredients in proportion. But even when you've got yourself a carefully selected sample spoonful that's representative, there's still room for error. The pollsters generally say that 95% of the time, a poll of 1,000 people will have a margin of error of plus or minus 3%. That means that a figure in the poll could be up to 3 percentage points higher or lower than shown. And that's the reason why different polls will show slightly different results. And their spoonful of voter soup is slightly different.

[B]

The BBC's poll of polls is suggesting the Conservatives are on 34 and Labour on 33. It's very close.

[C]

Yeah. Throughout the campaign, about 85 or 90% of UK polls, the vast majority, have shown there's between 0 and 2 percentage points between the two parties in the lead. So that's where we're at, slap bang in the margin of error. And that's one of the interesting things about this election. The other is the fact that such a large number of people are saying they'll vote for a party other than the traditional big three, Conservative, Labour and the Liberal Democrats. In the last election, 12% of people voted for a party other than those three. This time, the poll suggests that figure could be at least double.

[B]

It's a big change in five years.

[C]

It is. And one, actually, you might not at first notice looking at the UK polls, if you just focus on the big two. And there's another issue as well. Taking a spoonful of UK voter soup is actually only of limited use when you've got such large regional and national differences as we do. I mean, you only taste a bit of the Scottish broth that's gone into it. For example, UK polls show only about 4% of the electorate think they'll vote for the SNP. But polls of Scotland show 40 to 50% of people, 10 times the proportion, saying they'll vote SNP, giving them a fairly consistent 24 point lead.

[B]

That would make it SNP soup.

[A]

On Monday, April 27, Ruth looked at Labour's plans to introduce rent caps.

[C]

Their idea is that landlords would have to offer private tenants three year tenancies with a right to cancel six months in, and rent could only go up with cpi, the Consumer Prices Index measure of inflation, which is used to upgrade pensions and benefits and so on. Now, this would affect an increasing number of people. About four and a half million people privately rent their homes in the uk, about the same number as social renters. The number of owner occupiers, that's still the majority of people, about 2/3. It's 17 million in the UK, but that's dropped sharply over the last decade by about 6% in England and Wales. And at the same time, the numbers of people in the private rented homes has increased.

[B]

So who's renting?

[C]

Well, people who tend to be younger renters tend to be younger than homeowners. On average, almost half of all young households aged 25 to 34 rent privately in England and Wales, and that is twice the rate it was a decade ago.

[B]

And why is housing such a popular topic now with politicians?

[C]

Well, housing affordability is a Big problem at the moment. That's why, in short, we haven't been building enough homes to meet supply. And that means there's a big demand for both homes to buy and rent. Private tenants are hit by a double whammy because the shortage of homes also means rising rent. So it's hard to buy and it's getting more expensive to rent. To give you an idea, in the last year rents across the UK went up by an average of 2%. In London, it was higher still. Now, given that inflation is at zero at the moment, if Labour's policy were in effect today, rents for the next year would be frozen.

[B]

Has a rent cap worked in other countries?

[C]

Well, it's interesting to look actually, Germany already has a rent cap and there, once you start a contract, there are limits as to how much your rent can go up by. In a similar vein to what Labour's suggesting now, what happens as a result is that landlords tend to hike the rent between contracts. So there is a concern that UK market caps might actually push rents up. Now, in the end, in the uk, the big problem will remain, and that is that demand outstrips supply, particularly in crowded areas like London and the southeast of England. You know, everyone agrees that the big solution is to build more homes. Everyone also agrees that we shouldn't build those homes near to them. And that is the UK housing affordability problem in a nutshell.

[A]

On Tuesday, April 28, the PM team asked Tim to work out just how boring the election has been, statistically speaking.

[D]

Now, it seems a bit harsh, I have to say, Eddie, to require that an election allows a democratic mandate and a check on political power and also provides a rich seam of entertainment. So the question's ludicrous. That said, I am going to give you one statistical reason why the election is the most boring ever and three other statistical reasons why it's the most interesting in living memory.

[B]

How almost balanced of you. Go on then.

[D]

Well, the statistical proof that the election is tedious, at least for some, is the TV debate ratings. They're much lower than five years ago. So the first leaders debate in 2010 had an average audience of 9.4 million. The leaders debate this time around drew just 7.7 million. That's particularly disappointing if you divide it by the number of party leaders involved.

[B]

But you've got three numerical reasons to say the election is interesting.

[D]

Yes. So, number one, last election, only 12% of people voted for a party other than the traditional three Conservative, Labour and Liberal Democrat. This time, the poll suggests it could be double that. So we've seen much more interest in Ukip and the Scottish Nationalists, more interest in the interviews given by Nicola Sturgeon and by Nigel Farage. And that brings me to a second statistical reason that this is a fascinating election. Scottish voters may already have noticed, but north of the border, this election looks like it's going to be transformational. So polls of Scotland show the SNP well ahead, and that's bad news for Labour. At the moment, they have 41 seats in Scotland. If the polls are right, they could end up with just four or five.

[B]

All very interesting. What else is interesting?

[D]

Well, reason number three, why this is a fascinating campaign. Undecided voters. So, looking back to the election in 1983, at this stage, only 10% of people said that they might change their minds. And you had a similar proportion in 1987. But the latest Ipsos Mori poll shows that almost 40% of voters are still undecided this time around. Now, that's a big number. A lot of people still weighing up their decision.

[B]

All right, you've talked around this a bit, but if I pressed you for a single number for how boring this campaign is, what would it be?

[D]

A single number?

[B]

A single number.

[D]

6.

[B]

Why?

[D]

This is the longest general election campaign we've ever had. Six weeks. They used to be four weeks long. And each week of campaigning, if I can be forgiven a metaphor, it's like a donut.

[B]

Oh, no.

[D]

So the first donut, enjoyable. The second one a bit much. Now we are through the fifth one, Most of us are feeling sick.

[A]

On Thursday, April 30, Tim Harford's already here.

[B]

What's on your mind, Tim?

[D]

A Conservative Party claim they're so proud of that, they put it on their first election poster. 760,000 more businesses. They say they have been formed since the coalition government came into office five years ago.

[B]

And is that true?

[D]

Well, yes.

[B]

Thanks for coming in.

[D]

Sort of true. Official statistics show that it is true. This is between the start of 2010 and the start of 2014. And it's the number on private sector businesses in the UK. It's increased by 760,000 and it is now up to 5 million, which is a record. Now, this all makes it sound like the UK has gone completely crazy with Dragon's Den style entrepreneurial spirit. That's the message the Conservatives would like to sell. For certain, it's not necessarily what's going on, because only about 60,000 of these businesses are the kinds of businesses you might actually go and get a job at. The other 700,000 don't have any employees, they are simply self employed workers.

[B]

What sort of work are they doing then?

[D]

Well, the rise in self employment has been biggest among management consultants, bookkeepers, photographers and chartered accountants. More than 4 in 10 of self employed people are over 50. So a lot of people have taken redundancy, they've taken early retirement and they've set themselves up in consultant roles. Many other self employed people are doing skilled work. Builders, carpenters, taxi drivers are also a large group. So it might not surprise you to hear that many self employed people were born overseas. According to Oxford University's Migration Observatory, the number of foreign born self employed workers increased by 225,000 over the past five years. So that is about a third of the total increase.

[B]

And are all these self employed people doing well?

[D]

Well, no, not on average. We've got data from 2008 to 2013, so that encompasses both the severe recession under Labour and the slow recovery under the coalition. And over that time median income from self employment fell by a and we don't have the data as to whether it has since recovered. Now this rise in self employment is predominantly because once people become self employed, they don't then leave self employment and that seems to be because they can't find jobs. So for a lot of people, not everybody, self employment is simply a better alternative to unemployment.

[B]

What about more traditional entrepreneurship? How's the country doing there?

[D]

It's hard to say. There are some reports on this. For instance, the think tank the Social Market foundation did some research into what it calls high value entrepreneurs. Not entirely clear what they mean by that, but they say that the UK isn't doing particularly well on that count. These aren't official statistics, though other think tanks may give you a different answer.

[A]

On Tuesday, May 5, something different.

[B]

The team from Radioforce more or less has spent the election campaign examining the claims on tax rises, economic growth, zero hours contracts, sure start centres, the number of immigrants with hiv. The list goes on, on and on and on. But tonight, a break from the election to look at some other statistics in the news. Ruth Alexander is here. And you're looking, Ruth, at something we've heard about today. This report on maternal mortality.

[C]

Yes, that's right, it's spawned loads of headlines. Pregnancy is six times riskier in the UK than in Belarus. So says several alarmed headlines. And our fact checking instincts were immediately aroused at the mention of Belarus because, you know, since when was Belarus the key place we compare ourselves with? You know, it's got a population seven times smaller, smaller than ours, some would say it has a dictatorial precedent and it's certainly got a Soviet Union past. It's just not immediately obvious why we should be benchmarking the UK against Belarus. So I delved into the Save the Children report that the statistics are from and it was soon apparent why this comparison is being made. It's because Belarus has the best maternal mortality records in the world, at least according to the statistics anyway. They have one death per 100,000 live births, the UK has eight. So comparing us with the best is an easy way to make the UK look bad and create headlines. But there's reason to believe that the statistics are making the UK look worse than it is and Belarus look better,

[B]

so the stats might be the factor. Why does the UK look worse than it is?

[C]

Well, because we take counting deaths in pregnancy and very seriously as a nation. We don't just look at death certificates to find out how many mothers are dying in pregnancy. As many countries do. We ask hospital staff, coroners, and we use birth registration data to follow up on mothers in their first year after birth. And because we're so thorough, it makes our numbers look high in comparison to other countries.

[B]

And why, on the other side of this, do the stats make Belarus look better than it really is?

[C]

Well, its record looks so good that it seems suspicious. It's not just the uk, they're doing better than it's rich countries with really top notch healthcare like Switzerland and Sweden. Now, the World Health Organization, whose figures these are, they told me that it's actually really hard to be sure of the numbers coming out of a country as small as Belarus. You know, it's only got a population of about 9 million and small numbers tend to bounce around a lot. The who's relied on the deaths registration data for Belarus and then adjusted them to allow for the fact that death certificates in any country, not just Belarus, death certificates don't always mention that someone was pregnant. But it's possible. The WHO says it's possible they've not adjusted by enough.

[B]

So how does the UK compare with other countries?

[C]

Well, Oxford University, which carries out all the investigations into maternal deaths in the uk, says there are only a very small number of countries which are as thorough as we are, and that is France, Ireland and the Netherlands. So they say, you know, it's only really worth comparing us to them if you want to get an accurate idea. Now, our death rates are identical to France and to Ireland's, and they're a bit higher than the Netherlands. But even that might be a statistical anomaly, apparently, because the Netherlands may not be as rigorous, including death from as many causes as we are. And it's also worth thinking about what maternal mortality means. I mean, it sounds like it must mean women dying in childbirth, but actually there's a wider focus than that. It encompasses women who die during pregnancy or within six weeks afterwards. Now, in the uk, the pregnancy is a direct cause of death in a third of cases. In 2/3 of cases. So the vast majority, the death was caused by unrelated medical or mental health problems that got worse during the pregnancy. And in most of all those cases in the uk, the women were already in ill health when they became pregnant. And the uk, it has to be said, has been making real progress on reducing the numbers who die as a direct result of their their pregnancy. So we really are. We're doing okay. In fact, more than okay. And in fact, when you look at the statistics, the real story there, it isn't about the differences I've just been talking about. It's about countries like those in sub Saharan Africa, where maternal mortality there can be 10 to 100 times higher than it is here in the UK.

[A]

Thanks for listening to the final More or Less Election podcast. Don't forget, it's the general election on May 7. If you have any story ideas for us, please email moreorlessbc.co.uk or tweet bcmoreorless.