Is prostitution really worth £5.7 billion a year?

As of the 23rd May 2022 this website is archived and will receive no further updates. was produced by the Winton programme for the public understanding of risk based in the Statistical Laboratory in the University of Cambridge. The aim was to help improve the way that uncertainty and risk are discussed in society, and show how probability and statistics can be both useful and entertaining.

Many of the animations were produced using Flash and will no longer work.

The EU has demanded rapid payment of £1.7 billion from the UK because our economy has done better than predicted, and some of this is due to the prostitution market now being considered as part of our National Accounts and contributing an extra £5.3 billion to GDP at 2009 prices, which is 0.35% of GDP, half that of agriculture. But is this a reasonable estimate?

This £5.3 billion figure was assessed by the Office of National Statistics in May 2014 based on the following assumptions, derived from this analysis. To quote the ONS:

  • Number of prostitutes in UK: 61,000
  • Average cost per visit: £67
  • Clients per prostitute per week: 25
  • Number of weeks worked per year: 52

Multiply these up and you get £5.3 billion at 2009 prices, around £5.7 billion now.

This assessment has been severely questioned. Dr Brooke Magnanti, aka Belle de Jour, reckoned it might be ten times too high. In contrast others have said it should be £9 billion as it ignores male prostitution. Jolyon on Tax Relief 4 Escorts, who claims a maths degree from Cambridge, has done a detailed critique. He points out the flaws in the survey on which the 61,000 is based, and claims the assumed workload is too high and that the cost per visit (which the ONS based on PunterNet) seems too low: it is somewhat ironic that the ONS use an information source that a previous minister, Harriet Harman, tried to shut down.

My feeling is that the assumption that has the most problems is the workload. ONS are suggesting that the average person who works in prostitution has around 1,250 clients a year. This is based on Dutch experience, whereas the pattern of working in the UK is likely to be very different, with a complex industry comprising street-walkers, escorts, the informal market, those who work from fixed premises and 'independents' who advertise, for example, on AdultWork. Many are part-time.

As always, it's best to do a simple reality check. The ONS assumptions come to around 75,000,000 visits a year. Let's say 60,000,000 are from locals rather than foreign visitors, which is more than a million a week. There are around 20,000,000 men between 18 and 65 in the UK (taking an arbitrary upper limit), so this would mean that on average each of them buys sex three times a year. In fact the latest Natsal survey found that around 4% of men between 18 and 65 reported paying for sex in the last 5 years, that's about 800,000 men. If there were really more than a million visits a week, then the average man who paid for sex at any time in the last 5 years, did so considerably more often than once a week. In fact the proportion who pay for sex each year will probably be less than 2%, which means that less than 400,000 men are taking up over a million visits each week - that's around once every 3 days for each of the 400,000. I am no expert on the behaviour of this subgroup, but this does seem rather high, to say the least: a study of men who pay for sex in Scotland found a mean of only 5 partners in a year.

The assumptions also mean that the average person working in prostitution is turning over nearly £100,000 a year, which Jolyon from Tax Relief 4 Escorts says is completely implausible, and he should know.

Although this is a big statistical challenge, such an important contribution to the economy deserves a more robust analysis. When better figures come out I predict the UK will be due a substantial rebate. But that won't help David Cameron now.

27th October: Some figures have been revised since first posting, but the gist stays the same.


Hi David It was good to read your analysis, and can I thank you for linking through to my piece. I think you're right that the ONS assumption with the biggest problems is workload. When the figures first came out in June this year and real sex-workers started discussing them on forums and twitter, idea of seeing 25 clients week in week out was pretty universally dismissed as laughable. I know as an accountant that sitting down and talking to more than 3 tax clients a day is mentally exhausting enough and not something I'd want to keep up over a sustained period. So I can quite see why anyone selling a more physical experience would find this workload impossible to maintain. There certainly are a small proportion of very successful and commercially minded women who turnover more than £100k per year. But they don't do it by seeing 1300 clients a year and charging them £67 a time. A much more likely scenario is 10 regulars paying an average of £500 per month, supplemented by some one-off clients and a few longer over-night bookings. If you haven't seen it there is some fascinating research on this in the 2nd Freakonomics book. In practice many sex-workers decide how much they want to earn each week or month, and once they've achieved that they shut up shop until next month. At the bottom end of the market it's about earning enough to pay the rent, or in some cases feed a drug habit; while at the other end it's more about enjoying the freedom that comes with self-employment to work when and how you choose. I doubt very much that there are more than a very small percentage of sex-workings seeing 25 clients per week throughout the year. The other big difficulty, which makes the whole exercise very difficult to estimate, is the unrealistic ONS assumption of a homogeneous market. This leads to averages that are very difficult to assess for reasonableness. For example their average price per of £67 per encounter probably doesn't represent what anybody actually charges. Typically escorts will have a starting price above this level, at maybe £100 or £150 for an hour. While street work prices are probably much lower. The last time (actually the only time, about 10 years ago) that I was given a guided tour of an East London brothel the menu of services showed there wasn't much on offer below £50. When the ONS published their data at the very end of May they invited responses, so as a good public spirited citizen I wrote to them. They didn't reply. Maybe now that their calculations seem to have partly resulted In a large bill from the EU they will have another look at some of the criticism of their results? If anyone wants to do some robust research I would be more than happy to help them understand what is a very complex industry. Thanks again for your kind comments about my analysis. I did indeed spend 3 years grappling with the Maths Tripos in the 1970's so it's good to know it wasn't entirely wasted!

Dear Jolyon

thanks for your insights, very valuable. As you say, such an important topic deserves a more careful analysis. d

The argument on prostitution seems persuasive. But this is not why we have received the new bill. The revised estimates are based on the ONS massively revising its estimates of some activities such as charity within the framework of ESA1995. One of the great problems with navigating the economy is that the ONS keeps changing its mind not just on where we are but on where we were 5 or more years ago. One of the dangers in having innumerate decision takers is that they leave issues like this to technicians who do not see the wider picture.

Dear David, Since I work at Eurostat in the National Accounts department, I cannot leave it uncommented :) I am sure that you are aware and I also want to mention it to the reader. The majority of GDP upwards revisions are due to capitalisation of Research & Development, only a small portion is due to inclusion of illegal activities. Legal prostitution was already part of GDP before, so we have to see which part of the numbers is due to illegal activities, since only that part entered just now. So while I agree that the numbers ONS used to estimate prostitution seem questionable, a downward revision of those numbers would only result in a tiny little GDP downward revision. Thus I just doubt the conclusion that "When better figures come out I predict the UK will be due a substantial rebate.". There might be a rebate for the UK, who knows. But it would be more likely a political one than a statistical one. And frankly speaking I hope that there is no rebate anymore, because how can UK government explain to EU citizens that they pay less to the budget just because they are British? :) Cheers, Daniel