How many prostitutes are there in the UK? According to the Office of National Statistics the answer is 60,879. In figures due to be released next week this number is being used to add £5.314bn to the official size of the UK economy. But the ONS only attempted to measure the number of female prostitutes. If male prostitutes are included in the count then the contribution that prostitution makes to the UK economy rises to £8.856bn.
From the 30th of September the UK’s national accounts will attempt to measure illegal transactions to which all parties consent, including the sale of illegal drugs and prostitution. Illegal transactions are difficult to measure because the participants, while willing, are anxious that their business goes unnoticed. As a result there are very few obvious ways to directly measure illegal transactions and the ONS have been forced to rely on 10-year-old survey data in order to try and estimate the level of prostitution activity in the UK.
But there is a better way of measuring the number of prostitutes than using survey data. While many of the activities associated with prostitution are illegal in the UK, paying for sex is actually legal and as a result, prostitution services are widely marketed on the web. We can use Import.io to directly count the number of prostitutes who are marketing on the web and attempt a better estimate of the number of prostitutes in the UK.
The ONS used a study by the charity Eaves Housing For Women published in 2004 as the basis of their estimation of the number of prostitutes in the UK. The Eaves study tried to count all the female prostitutes in London over a 6 month period by telephoning every brothel and escort agency that they could find advertised. There are two main problems with the ONS using the Eaves study in this way: Eaves only surveyed the sex industry in London and Eaves only surveyed women working in the sex industry. We looked at each of these problems in turn.
But first, our choice of website
In order to do the work we built an extractor to the website AdultWork. AdultWork is an online marketplace for adult service providers. We chose to perform our analysis on AdultWork because AdultWork is much more popular relative to other similar websites.
Problem 1: London only (not a problem)
The ONS used population figures for London and the UK to extrapolate out from the Eaves estimate of the number of prostitutes in London to an estimate of the number of prostitutes in the UK. The first thing that we did was to test the validity of this extrapolation. We used our extractor to pull down the AdultWork profiles of all the adult service providers in the UK who offer escorting. We then divided the profiles into two groups: those from London and those from the rest of the UK and compared the ratio against current population statistics for London and the UK. The prostitution ratio matched the population ratio. Our data support the validity of the extrapolation from the number of prostitutes in London to the number of prostitutes in the UK. The fact that the Eaves study only surveyed the sex industry in London is not a problem.
Problem 2: Women only (big problem)
The Eaves study is quite clear in stating that “no information is included regarding men working in the sex industry”. This means that any estimate that the ONS makes of the number of prostitutes in the UK is only an estimate of the number of female prostitutes. We used the same set of AdultWork profiles to see how much of a difference including men would make to the total number of prostitutes. When we counted the number of male and female service providers on AdultWork it turns out that 42% of them are male. This is not a small proportion. The fact that the Eaves study only surveyed women working in the sex industry is a major problem.
What’s the impact?
If 42% of prostitutes are men and there are 60,879 female prostitutes then we would estimate that there are 44,085 male prostitutes and that the total number of prostitutes in the UK is 104,964. If you use this number in the ONS calculation for the amount that prostitution contributes to the UK economy then you get £8.856bn (0.6% GDP) rather than £5.314bn (0.4% GDP). A difference of £3.542bn (0.2% GDP).
Using web data for this kind of economic analysis has two main benefits. It means that our measures can be up to date, even close to real time and that our measures can be more complete and less reliant on scaling assumptions.
Want to know more?
For more information you can watch the 15 minute video of me presenting this story at our last Data Summit and access my slides below.
Be sure to join us for our Data Summit in San Francisco for more great data stories.