← More or Less
More or Less
Wedding Stats
00:27:53
In More or Less this week: a cornucopia of wedding-related numbers. And AV explained.
Loading summary
Transcript98 lines
- [00:00]
A
Thank you for downloading the More or Less podcast from the BBC. You can find out more about our programme on our website, BBC.co.uk radio4. But before you do that, here's Tim Harford.
- [00:15]
B
Hello and welcome to More or Less A Happy Marriage of Numbers and News. We may be fans of Cold Logic and Hard Data on this program, but let's rim in our hearts for romance. So we're devoting much of this edition of More or Less to wedding related statistics. Think of it as our way of throwing a street party. And although we can't offer bunting or cake, we can offer startling insights like this.
- [00:39]
C
The 750 million audience that has been quoted widely for the Charles and Diana wedding, intriguingly, actually has no known source.
- [00:49]
B
But let's start with the fact that despite Friday's great event, the numbers suggest marriage just isn't all that popular anymore. Wesley Stevenson's here. Why is it can't be true, can it?
- [00:59]
D
Well, I'm afraid it is. It is true. Marriage has been on the slide, in fact, since the 1970s. And that's not just in terms of absolute numbers, but also in terms of marriages per head of the marriageable population too. So that's the marriage rate. The only blip was in the early 2000s when we started to see an upturn. It looked a bit odd given the prevailing slide, and in fact it was something that puzzled the newspapers at the time too, until they found the explanation. The upturn was of course attributed to the influence of the marriage of David and Victoria Beckham, now Posh and Becks,
- [01:31]
B
the saviours of marriage. Are you sure?
- [01:33]
D
Well, I was a little suspicious about this, so I went to visit Mark Rimmer. He's the head of registration at Brent and Barnet Council and also on the executive committee of the Local Registration Services Association.
- [01:46]
E
It certainly wasn't down to Posh and Becks here in Brent, that's for certain. There aren't many Man United supporters in Wembley, I don't suppose. I think here it was purely to do with people trying to get in and get married before the government introduced legislation to clamp down on sham marriages. That's why we had the huge increase. There was certainly one couple which does stick in my mind. She was an incredibly beautiful, leggy six foot Eastern European female and the guy with her looked as though he'd just planned out of a skip in Wembley High Road with his cans of strong lager sticking out of his pocket. That one was one of those that you think this is an absolute blatant. Abuse of the system. So. Yeah, but that was an everyday occurrence then.
- [02:25]
D
Can't you just turn them away?
- [02:27]
E
Unfortunately not. We don't have the powers to turn them away. The only powers we do have is to report suspicions of sham marriages to the Home Office where then they can decide whether enforcement action is appropriate.
- [02:37]
B
You said the marriage rate is lower than it's ever been, so what changed?
- [02:41]
D
Well, yes, that's right, it is. The law changed. That's what changed. Around 2004, new regulations were introduced that made these sham marriages much more difficult.
- [02:51]
E
The effect was immediate and dramatic. In 2005, we went down in Brent alone from having 1,600 weddings, weddings a year down to 800 that in London equated to over 15,000 marriages that disappeared in year one. And over the country, 25,000 marriages disappeared as a result of that legislation being introduced.
- [03:13]
D
And that's all because of David Beckham and Rebecca Lewes, is that right?
- [03:17]
E
That's obviously what it was about. Yes, undoubtedly.
- [03:21]
B
But Chan marriages haven't stopped altogether, have they?
- [03:24]
D
No. On the day I was shown around the facilities in Brent, it became apparent that it was. Isn't that unusual at all.
- [03:30]
E
We've had a whole variety of events today in terms of marriage and civil partnership. This morning we had what strongly appeared to be a bogus marriage, a sham marriage. The first one of the day, they couldn't even communicate with each other in the same language, which gave us the clue that one will be reported to the Home Office. We had a civil partnership, two guys getting married to each other, which again is not an everyday occurrence. And then we had a nice Asian wedding. This afternoon, however, we've got another foreign national being married, an interesting coupling between Algerian and Hungarian and we have a Sri Lankan couple getting married. So again, a typical day in the life of Brent Council. Really.
- [04:08]
D
Is that typical in terms of the
- [04:11]
E
ones you're suspicious about, the ones about foreign nationals and European Union nationals are very typical. The classic is a foreign national marrying another European apart from a British national. It's a lot easier to get a stay if you marry a European other than a Brit.
- [04:27]
B
So even though marriages are at an all time low, some of these marriages are still likely to be bogus.
- [04:32]
D
Yes. And the law is set to change again next month, which Mark Rimmer told me means that all the work to keep the numbers down is going to be overturned.
- [04:40]
E
I think, unfortunately, we're going to be Back to pre 2005 levels of marriage abuse. The government introduced that legislation back in 2004. 5 Effectively that's been challenged in the High Court and much of it has been deemed to be incompatible with human rights legislation. So what's going through Parliament right now is a remedial order under the Human Rights act, which will abolish most of the provisions that were introduced to stop sham marriages. That will go on the 9th of May. I'm very worried, as my colleagues are worried right across the country, that that will be the start of a flow of sham marriages once more, and we are concerned the UKBA will won't have the resources in place to stop it. There will be a huge increase in the number of marriages, I suspect, this year, and it will be put down to Kate and William getting married, I'm sure, but that'll be nothing to do with it.
- [05:28]
D
Interestingly, he also said that the divorce rate has tracked that of the mari. What?
- [05:32]
B
I thought we agreed not to mention divorce in this programme, but Australian marriages
- [05:36]
D
have gone down, so is the rate of divorce, but.
- [05:38]
F
Shh.
- [05:38]
D
But there are actually some really interesting statistics related to divorce.
- [05:43]
B
Wesley Stevenson, thank you very much. Weddings, royal or otherwise, can be very expensive. But how much does an average British wedding cost? It's a subject I take a keen interest in. I have two daughters. We turn for advice to the online journalism expert Paul Bradshaw, who, among other pursuits, runs a collaborative investigative journalism blog. The blog Help Me Investigate recently took on the subject of how much a British wedding really costs. Here's what Paul discovered.
- [06:17]
G
The figure of £20,000 being the average cost of a wedding comes from a magazine called you'd and you'd Wedding. And that kind of figure seems to have taken on a life of its own ever since.
- [06:28]
B
Love and marriage. Love and marriage go together like a horse and cow.
- [06:36]
G
They came to this £20,000 figure by inviting readers to complete a survey in their magazine. Part of the investigation that we conducted found that probably about 0.6% of that year's brides or grooms will have completed the survey. So quite a low proportion, about 1,500 readers in total. And they were asked a variety of questions about how much they spent on different parts of the wedding or how much they intended to spend. There's all sorts of factors in how that figure was calculated which I think make it particularly interesting.
- [07:15]
B
It's an institute you can't disparage, ask the local gentry.
- [07:22]
G
One of those factors is obviously that the average that they use. There are various different types of averages. You can take the total sum of everything and divide it by the numbers of people, or you can take the Middle value across the range of people, or you can take the most popular value. It's very strongly suspected that they used the kind of most common average, which is a mean. This isn't a very useful average to use with any financial information. With spending information, it's really only useful for places where you get natural distributions, things like heights or weights. If you're looking at wedding costs, for example, if one person in that sample says they're going to spend a million pounds on that wedding, then that's obviously going to skew the overall average significantly. Likewise, at the lower end, you're getting people who are self reporting they are cheap to complete this survey, they're not being sought out. And people are going to be unlikely to fill out a survey if they're not spending much on their wedding. So it won't necessarily reflect the true reality of things. The final issue, of course, is that the whole sample is selected from readers of a particular magazine, an aspirational magazine. One of the things we found was that the social breakdown of the readership of that magazine, for example, was obviously not identical to the social breakdown of people who were getting married. For example, according to the magazine's own statistics about the breakdown of its readership, 30% of their readers came from the C2 social classification, which is a semi skilled class of workers or class of professions. By contrast, 18% of people who got married that year were from that class. So that particular type of profession was over represented in the readership of the magazine. By contrast, people from the lower social classifications, the D and D classifications, they only made up 13% of the readers of the magazine, but 27% of people who got married that year. So in those particular categories, you can see a clear difference between how representative the magazine readership was of people getting married in general. And that's before you take into account what proportion of vorse readers actually filled in the survey. Love and Marriage. Love and the Wall Street Journal have a blogger called the Numbers guy who writes a lot about statistics and he actually looked at the same subject in America and the average cost of weddings in America. They came up with a $20,000 figure and they managed to get access to the original data and looked at a different type of average, the median, which is the mid value. And I think this is a much better way of looking at marriage costs. And in America the median came in at about half the mean, so the median was about $10,000. And you could guess that based on the American results, it may be a similar figure in the UK. That actually the middle the median average UK wedding cost would be around £10,000.
- [10:35]
B
Paul Bradshaw, you and your wedding stressed to us that their survey has always been transparently, honestly and consistently conducted, which Paul Bradshaw doesn't dispute. You're listening to More or Less with me, Tim Harford, in association with the Open University. Our website is BBC.co.uk more or less and our email address is more or lessbc.co.uk and you can follow me on Twitter as Tim Harford. Now a long standing curiosity in economics is something called the marriage premium, a tendency for married men to earn more than single men. Dr. Francesca Cornaglia is an economist at Queen Mary University of London and LSE's Centre for Economic Performance. And she's been researching a rather unusual angle on the marriage premium.
- [11:21]
H
Yeah, when we started working on marriage premium, actually people were surprised. Oh, not again. Another paper. Marriage Premium, because it has been extensively debated in the economic literature. So this evidence that there is an advantage in terms of wages for married men has been there in the literature for a long time. Roughly, it goes depending on the sample, people have been using from 10 to 40% higher wages than single counterparts.
- [11:45]
B
So this is a 30 year old married man versus a 30 year old single man and they've got, I don't know, a quarter or a third more income.
- [11:53]
H
Yes, yes, that's exactly the case. Yeah.
- [11:56]
B
And what are the theories as to why this might be?
- [11:59]
H
It could be for many different reasons. Like basically selection might be based on unobservable characteristics that are correlated with both marital status and productivity.
- [12:08]
B
So women like to marry rich men?
- [12:11]
H
Yes, or the other justification could be that in actual fact married men are perceived as more stable and therefore it's employer discrimination.
- [12:19]
B
Okay, so that's the marriage premium in general. But you've been looking at a bit very specific and unusual data set.
- [12:25]
H
Yes, we investigate the relationship between wages, marriage and productivity using data on professional baseball players. And baseball is a unique case study because it has a long history of statistical collection and numerous direct measures of productivity.
- [12:39]
B
How far back does the baseball archive go? I mean, is this sort of 20, 30, 40 years?
- [12:43]
H
It's something like 140 years. Starts in 1871 and we have it down to 2007.
- [12:49]
B
This is an amazing data set. How is the productivity of baseball players measured?
- [12:53]
H
We use batting average, which is defined as the number of hits divided by the number of opportunities to bat in a season. But we also use more popular measures like the ops, which is on base plus slugging.
- [13:05]
B
On base plus slugging. What is that?
- [13:08]
H
It combines the obp, which is the on base percentage, which takes into consideration a number of ways a battery can get to a base. So combines these OPS statistics with a measurement of the player's ability to hit for power. So it's a weighted average of the number of bases reached per at bat.
- [13:25]
B
Okay, I, I can feel myself getting stupider as you were talking. I'm just going, I'm just going to have to do more research on baseball.
- [13:33]
H
I had to do that at the beginning. I knew nothing about baseball and I started with Wikipedia and I made my way through.
- [13:38]
B
Okay, so we've got all these different statistics that will mean something to people who know something about baseball statistics. And you've got all these men, and they are all men, I guess, and some of them are married, some of them are single, and you've got 140 years of data. So what do you find out?
- [13:54]
H
The whole idea was just to test whether the marriage premium that we do observe and is of the same size as other people have observed in the past is around 20% was due to an increase in productivity of these married men. Meaning getting married makes the men more productive or it was to do something else. Because until now they were saying, right, we only measure wages, but that's a good proxy for productivity. Just implying that getting married had to make the man more productive. What we actually find in the data is that wages are going up, but that's not because of marriage makes men more productive, because productivity stays the same before and after marriage.
- [14:29]
B
But the married men are still paid 20% more than the single men they are.
- [14:33]
H
So the interesting triggering question now is why what happens? It all boils down to image. A wife can have a direct effect on the wages, which is just boosting his image, making the man appear as more reliable, more stable, more trustworthy to the employer.
- [14:51]
B
You've clearly got this amazing data set. But I guess the question is, haven't you just told us a fact about very select bunch of very talented, very highly paid men and not really told us anything about the wider labour market?
- [15:06]
H
Well, that's what we at the beginning were concerned about. But then really looking into the characteristics of the Basel players, I mean, what their life is about and the characteristics of their life. Like they have to travel a lot, the long work hours, they're very often out away from home. It's in common with many other jobs like CEOs or politicians or very highly paid workers. So sure, we can't make an inference from our results to the general Labour markets. But to some labour markets which are not baseball players, I think we feel confident to do it.
- [15:41]
B
Dr. Francesca Cornaglia. Now, the royal wedding itself was of course followed not just by happy Britons, but by people all around the world. Figures from 1 to 2 billion have been quoted for the global television audience. But that was before Friday's big day. How sure can we be that any figures for global audiences are correct?
- [16:02]
C
My name is Kevin Alevee, I am director of Initiative Futures Sport and Entertainment. We measure the TV audiences for major events from sport and indeed from non sporting events from right around the world, in order to understand their true popularity. The 2 billion figure which has been bandied around as the audience for the royal wedding is rather on the high side. No sports event, or indeed non sports event, has ever had an audience figure of that scale. The single most watched event of all time was in fact the opening ceremony of the Beijing Olympics in 2008, and that was watched by very slightly under 1 billion people. And there's no credible reason for thinking that today's royal wedding would have been watched by double the number of people who would have watched the opening ceremony in Beijing.
- [17:11]
G
A royal engagement.
- [17:12]
B
Prince William finally pops the question.
- [17:23]
C
In many different countries around the world, there are panels of households whose viewing habits are monitored. And then on the basis of those households, their viewing habits for the population as a whole is extrapolated. In other countries where such data don't exist, we make estimates based upon our understanding of the media landscape in those countries and the demand and interest in the events in question. There have been many very intriguing, to use a polite term, techniques in terms of how TV audiences have been estimated in the past. One of the most common rules of thumb which have been used by many agencies is that they've considered what's the population of a country and then just assume that, for example, a blanket 10% of the population would actually have watched. Which means that invariably in such estimates, countries like Nigeria, Pakistan, Bangladesh, Philippines and so on, countries with very large populations would therefore have very large implied TV audiences. The next thing to think about is, well, what methodology have they used? Have they, for example, included the audiences for news clips in amongst the audience figures?
- [18:35]
A
In the early morning, the route was already lined with thousands of people waiting especially to check cheer at the first glimpse of the bride.
- [18:42]
C
Because, for example, for the royal wedding, it's very likely that a very high proportion of those people who watch the royal wedding would have seen only 30 seconds on a news clip, as opposed to hours and hours of the actual ceremony. Things are much more sophisticated these days. We're able to gather a lot more relevant data than merely the size of the population of a country. And what we can then do is use econometric modelling, which is a mathematical technique, in order to understand the relevant contribution of each of the many factors that drive TV audience demand. So we can be much more precise in our estimates nowadays. Essentially, there will be around a 3% margin of error in the estimates nowadays compared to what would easily have been a 50% margin of error historically.
- [19:44]
B
Charles Philip Arthur George, wilt thou have this woman to thy wedded wife to live together after God's ordinance in the holy estate of matrimony? Wilt thou.
- [19:58]
C
The 750 million audience that has been quoted widely for the last 30 years for the Charles and Diana wedding, intriguingly, actually has no known source. Most likely at the time, a journalist or a royal correspondent wanted to be very bullish about the importance of the wedding and came up with 750 million figure, really as a finger in the air estimate. And in the absence of any alternative or better data, that figure simply stuck.
- [20:33]
B
Kevin Alavy, you're listening to More or Less Now. The referendum on av, the alternative vote system is imminent. And all manner of campaigners and journalists have tried to explain how the AV system works. But if our inbox is any guide, they haven't done too well. We've had lots of emails from listeners asking us to explain it. So we will, or rather, Simon Terrington will. Simon studies the maths of voting systems and we invited him to tell us all we need to know about AV in seven minutes. Simon was delighted to accept our challenge. He had only one Can I use limericks? We said no, but he obviously didn't get the memo.
- [21:16]
F
Good evening, ladies, gentlemen and poets. As you know, because of financial pressure, we've recently had to merge two esteemed societies. So let me bid you a warm welcome to the inaugural dinner of the Society for Sappho GY and the Promotion of Conversation in the form of Limericks. We hope you've enjoyed the main course. And now, what shall we have for dessert? Ah, yes, Rick, the former president of the Society for the Promotion of Conversation in the form of Limericks. Is there something you'd like to add?
- [21:49]
I
The main course has gone down our throats. Before we all rush for our coats, we must eat our dessert. Let's agree it won't hurt to decide what we eat next with votes.
- [21:58]
F
Alright then, hands up for banana cake and jam tart. Banoffee piece ECLAIRS Waffles. Thank you. And death by chocolate.
- [22:15]
A
That's great.
- [22:16]
F
Well, the vote was very close, but of the six options, Jamtart won with 25%. So Jam Tart it is.
- [22:22]
A
Um, hang on.
- [22:24]
F
Yes. The chair recognizes Sophie the sophologist. Oh, Sophie, I hope you're not going to waffle on about voting analysis. I just want the waffles.
- [22:32]
A
You've used the first past the post system. Here everybody expresses a single preference and the one with the most votes wins. Well, in this case, that has a limitation. A quarter of us might have put jam tart first, but the rest of us hate it. This system has produced a bad result overall. I suspect that the jam lovers are just using it to preserve their majority.
- [22:56]
I
We have noticed a certain fragility in first Past the Post's ability to spread general contentment and limit resentment. It can't maximise total utility.
- [23:04]
F
Total utility? What does that mean? I was just after a pudding.
- [23:08]
A
Utility is sometimes expressed as the sum of human happiness. We know that we can please some of the people all of the time or all of the people some of the time, but it's really hard to please all of the people all of the time. Maximizing utility just means pleasing as many people as possible as often as possible.
- [23:29]
G
Possible.
- [23:30]
A
In the first past the post system, and in particular when there are many choices, it is possible to produce an outcome that keeps a small group happy and everybody else very unhappy indeed. And not just that. Banana cake got 19% and banofipie got 18%. Together, they have a natural majority. But the vote for banana split.
- [23:53]
I
There's a defect that sometimes can strike when the choices you have are alike. What do you say? You could go either way and end up with something you don't like.
- [24:01]
F
I know what I don't like, and that's people who rhyme alike with like. If you keep behaving like that, no pudding for you.
- [24:08]
I
There was a flaw there, but I'm hurt. Let's forget about that and revert. The mistake wasn't treasonable. We just want what's reasonable. We all just want our just dessert.
- [24:17]
A
Well put, Rick. Perhaps the alternative vote method would be more likely to give us our just desserts. First, we ask if any one dessert got a majority of first choices. If not, like in the case of the jam tarts, then we start by crossing off the choice that fewest people put first. After that, we ask those people what their second choice is and redistribute their votes accordingly. That will stop this problem with the vote splitting between similar options. For example, if banofee pie is eliminated at some point, then those guys will move over to the banana cake. If banana is their ruling passion and we keep going until we are left with just two, then the one with the most votes wins.
- [25:00]
F
It all sounds a bit complicated. Do people really feel strongly about their third and fourth choices?
- [25:06]
A
Well, people can just give a first choice if they want to, although this could give them less influence over the final outcome.
- [25:13]
F
Doesn't EV tend to produce middle of the road compromise puddings?
- [25:17]
A
To an extent it does, but is that really so bad if that dessert is closest to what most people want? Also, one could argue that it might work the other way. Under av, people may vote for extreme desserts knowing that they wouldn't have to live with the consequences because they will be backing a minority pudding which will be eliminated early anyway. For example, people might vote Death by Chocolate, the extreme chocolate option, knowing that they could vote for an eclair as a second choice under first past the post. They would pay a price because their vote would be lost at this point. But in the alternative vote system, their vote transfers. The problem with this is that if Death by Chocolate gets a lot of votes, it will start to claim democratic legitimacy. It'll want to appear on Question Time. If enough people vote for it, we might even end up having to eat it.
- [26:10]
F
We could, but that would still require half of us to put it reasonably high up the list of preferences.
- [26:15]
A
True.
- [26:16]
I
So AV is not what it seems. The realisation of dreams that voters would get us to some grand consensus. Instead, they might go to extremes.
- [26:24]
A
So to sum up, first past the post offers simplicity and is less likely to lead to extreme voting behaviour. It also doesn't put as much pressure on people to rank puddings, which they might not have strong feelings for. Either way, the alternative vote system is more likely to lead to a dessert that is more acceptable to more people and avoids problems with the votes for similar pud splitting.
- [26:47]
F
Can we please just make a decision? I'm starving. Oh, Rick, I can see that glint in your eye. I hope you're not going to go out on a lim. Rick.
- [26:55]
I
To all this talk. There's an addendum, two methods and you can defend them. But let's end this debate before it gets late resolving it by referendum.
- [27:03]
F
Why don't you chaps go off and have your vote? I'll stay here and guard the puddings.
- [27:10]
B
That was Simon Terrington as Rick with Caroline Wood and Katie Burtill. Thanks to all three. You might also be interested in a more mathematical take on voting systems, which we broadcast in May last year. There's a link to that item on the webpage. For today's program, go to BBC.co.uk more or less that's all we've got time for this week, but we'll be back next week with more data derived derring do. Until then, goodbye.
- [27:36]
A
More or Less is presented by Tim Harford, the Financial Times' undercover economist. The More or Less team is Richard Varden, Richard Knight, Wesley Stevenson and me, Hannah Barnes. More or Less is made in association with the Open University.