Craigslist and Prostitution

The Washington Post reports that Craigslist has stopped offering links to "adult services"ads. The change has only applied to U.S. users. Critics of course contend that Craigslist ads are a tool for prostitution and human trafficking. I've never seen one of these ads before - I assume they're nominally for dancers or escorts. Prostitution is always an interesting question for economists. Many conclude that there is no good economic reason for restricting consensual transactions for sex - any restrictions on prostitution have to find their basis elsewhere (which is sort of a useless point to make, I suppose - lawmakers don't generally turn to economists to comment on vice laws in the first place).

Anyway - this story reminded me of a paper that was presented at the First Annual IZA Conference on the Economics of Risky Behaviors, which I presented at last Spring. The paper, by Scott Cunningham (Baylor) and Todd Kendall (Compass Lexecon), was one of the best of the conference I thought. They looked at the relationship between internet prostitution and street-walking prostitution, and of course a discussion of these Craigslist ads featured prominently. Here is the abstract of the draft from last spring:

The increased proliferation of Internet and other technologies has profoundly changed the market for prostitution in the United States. We argue that the Internet lowers client search costs through the expediency of the search technology, allowing clients to learn about prostitutes in an area (including making comparisons), and thus mirroring the function of the street market. The Internet also reduces the cost structure of the prostitution supply function by both reducing the fixed costs and the variable costs of producing commercial sex through its negative effect on the probability of detection and the probability of arrest. On the one hand, such reductions in the fixed costs of production suggest that the marginal escort may be shifting towards younger women for whom the opportunity cost of detection is higher or who are more risk averse. On the other hand, thin market externalities in niche markets are overcome with Internet technology, which predicts a wider variety of worker characteristics as thin market externalities are solved via the Internet. We explore this conjecture empirically using a variety of data sources. First, we present evidence from a panel fixed effects model relating state-level prostitution arrests to the proportion of families in the state who have Internet access at home to show that Internet penetration predicts a decline in street prostitution. This negative correlation is robust to controls, but does not predict declines in other non-sex-related crimes. Second, we analyze 1998-2005 National Incident-Based Reporting System data (NIBRS). This data has to date not been used to learn about prostitution. We focus on both the cross-ORI and within-ORI changes in streetwalking offenses to learn more about the changes in jurisdictions happening to the kinds of women working in different locations. We present evidence showing that a majority of women caught by law enforcement are streetwalkers. Thirdly, we analyze data from an online clearinghouse called TheEroticReview to learn more about the change in the kinds of women working as prostitutes over time. And finally, we present data from a new dataset being fielded by the authors called the Survey for Adult Service Providers to analyze the characteristics of women currently concentrating their activities in the Internet-based, underground prostitution market. Policy implications are discussed.

I familiarized myself with the NIBRS this summer for our Summer Academy at the Urban Institute - a dozen minority undergraduates come for two months to be mentored and work on a research project, and my mentee (this is the second year I've participated) used the NIBRS for her work on changing Latino arrest rates resulting from new immigration enforcement policies. It's a fascinating dataset - detailed, individual level arrest records for the entire country going back years, and it's all public-use. If anyone is interested in any issues involving crime you should check it out.

Anyway - I digress. I haven't really thought about this paper for the conference until I saw this article in the post today, but it was a memorable one - great, thorough work with some innovative datasets. I didn't really ask any questions of this presenter because he went shortly before I did (and I was furiously looking over my notes), but now that I have a little time to look at this again, the obvious counter-argument to his conclusion is of course internet pornography. It may not be the case at all that internet prostitution is a well functioning market - it may simply be that with the availability of pornography potential Johns don't feel the need to seek out prostitutes. It's not an airtight alternative - it's not like pornographic videos weren't available before the internet - but it's plausible that videos weren't as readily accessible as, say, a magazine subscription, and what is being substituted for street-walkers is actually internet porn rather than internet hookers.

The other alternative is simply an interaction of internet porn and sex addiction. Let's say videos were readily and easily available before the internet - juts as accessible as magazines, for example. I am not a customer, so I wouldn't know. I still imagine if you're a sex addict watching the same videos (or reading the same magazines) over and over again gets boring. So pre-internet porn users who have sex addictions, for that reason, might engage the services of prostitutes. Then comes the internet - suddenly you have substantially more video available such that even a sex addict wouldn't necessarily get bored. Again, it's not necessarily internet prostitution that is being substituted for street-walking, but internet porn. I know internet prostitution is a major phenomenon - I'm not saying it isn't substituting for some of the street walking. But it's not necessarily the primary substitute. Since the authors only look at internet penetration into a community, it doesn't seem entirely clear what the primary substitute is. If it is internet porn, rather than internet prostitution, this could actually be a good thing for the safety of the women involved in the industry. They could get paid for performing in pornos with professional co-stars, rather than working the streets. Granted, I would imagine the pay isn't as good for a porn-star as for a prostitute, but in the Youtube age where you don't need a producer or publisher for that sort of thing (i.e. - lower entry barriers), the trade-off between safety and pay for these women might actually encourage many to cross over. That can't be a bad thing.