How Legalization Can Expand a Black Market

New research shows that legalization can make the black market larger rather than smaller

Human trafficking is one of the most revolting crimes on the planet, and it is therefore understandable that some governments have taken radical action to attempt to eliminate it, including legalizing prostitution. The argument, which seems sound on its face, is that expanding the supply of legal prostitution will draw demand away from the illegal market that relies on human trafficking. It is thus both surprising and important that new research from the London School of Economics shows that legalizing prostitution increases, rather than decreases, human trafficking.

This finding is stunning until one recognizes that demand for prostitution, gambling, drugs and the like is highly elastic. When the demand-suppressing effect of illegality is removed, demand can increase, sometimes dramatically. All of this new demand does not necessarily go to the legal market. Some new legal market entrants engage in the parallel black market some of the time (e.g., play legal slots at the casino by day and participate in illegal mob-run poker games at night). If the size of the newly legalized market is rapidly growing, these “fractional customers” can make the black market larger than it was in the pre-legalization period when all market participants were 100% reliant on the illicit market to meet their demand.

Other people may develop demands in a newly legalized market that over time lead them to move exclusively to the illegal market. For example, a man who begins going to prostitutes when sex work is legalized may realize over time that he wants to have sexual encounters in which he is physically abusive to prostitutes. He would then migrate to the black market where he can more easily indulge his violent propensities. Similarly, people who becomes addicted to a newly legalized drug may come to want to consume it so often and in such quantities that only the black market will meet their demands.

It has long been speculated that the phenomenon of legalization of a market growing a black market has already happened with gambling. There has clearly been an explosion of legal casinos, poker competitions, state-run lotteries etc., but no definitive research indicates whether the prevalence of illegal gambling has increased, reduced or stayed constant as a result. The new LSE research is I believe the first formal test of the possibility that legalization can grow a black market, and for prostitution it appears that it can (full paper here).

There is clearly more empirical work to be done in this area. But in the meantime, wise heads in the policy world will not take it as a given that legalizing something will necessarily shrink the black market.

h/t: Eric Voeten, who blogs at our sister site, the Ten Miles Square section of Washington Monthly.

Author: Keith Humphreys

Keith Humphreys is the Esther Ting Memorial Professor of Psychiatry at Stanford University and an Honorary Professor of Psychiatry at Kings College London. His research, teaching and writing have focused on addictive disorders, self-help organizations (e.g., breast cancer support groups, Alcoholics Anonymous), evaluation research methods, and public policy related to health care, mental illness, veterans, drugs, crime and correctional systems. Professor Humphreys' over 300 scholarly articles, monographs and books have been cited over thirteen thousand times by scientific colleagues. He is a regular contributor to Washington Post and has also written for the New York Times, Wall Street Journal, Washington Monthly, San Francisco Chronicle, The Guardian (UK), The Telegraph (UK), Times Higher Education (UK), Crossbow (UK) and other media outlets.

38 thoughts on “How Legalization Can Expand a Black Market”

  1. Wise heads in the policy world would examine the nature of the product/service being offered, constraints on supply & demand, limits imposed by regulation (including pricing, product diversity & availability) in order to determine or estimate the effect of legalization on market dynamics. There is not a general functional mapping of legalization onto markets that the LSE study purports to impute via the instance of prostitution.

  2. This finding would seem to hinge on the definition of ‘trafficking’. Does the UNODC make a distinction between migrant sex work and sexual slavery or does it assume all people migrating for sex work from poorer nations are being forced, coerced and/or exploited?

    1. Good point.

      I think it’s worth pointing out that there are other areas of legal work with a major trafficked migrant element, but we don’t ban those. Legal farm work led to a massive inward flow into the US of trafficked migrants, but we aren’t about to ban farm work in the US just to stop that – even when it occasionally does lead to some repulsive slave-like conditions (they broke up a farm a few years back that had been keeping migrants like slaves).

  3. I thought that this had been known in the Netherlands for years. It goes to show, I suppose, that policy analysis has limited predictive value.

  4. This study appears to be comparing apples and cough syrup. Unless legalization comes with a strong inspection regime and seriously expanded labor rights and outreach, there’s no necessary connection between the fact that the product is now legal and the question of whether the people producing the product got into their jobs in a legal manner. If you just reduce enforcement in general, the black market will grow because it’s cheaper to traffic sex workers than to pay them an official, legitimate wage.

    If marijuana were legalized would anyone be surprised if there were a substantial increase in the number of undocumented workers out in the fields?

    1. Paul, both Germany and the Netherlands do have strong labor laws for prostitutes (and exit programs, etc.). In fact, that was what legalization largely was about in both countries. Selling sex as such had never been illegal, only pimping and brothels: legalization of prostitution as employment was necessary to provide prostitutes with employee benefits and rights, but had the side effect of facilitating employment in the sex industry.

      The problem is that when, as a result of legalization, the sex industry grows, it becomes increasingly difficult to police it effectively. For example, the German police annually solves human trafficking cases related to sexual exploitation involving 600-700 victims, out of an estimated population of 200,000 prostitutes. Needle, meet haystack.

      However, the bigger problem is likely legal or quasi-legal exploitation. The vast majority of prostitutes in these countries come from abroad [1], especially poor Eastern European countries. While they technically have strong rights under the respective labor laws of each country, asserting them may mean that they have to go back to Romania etc. While there is a “right to reside” within the EU, this is limited to three months, unless you are employed, self-employed, studying, self-sufficient, or retired. This makes the right to reside de facto contingent on having a job, which makes holding on to your job far more important than for a citizen. Also, legalization made this kind of migration possible in the first place.

      [1] With an average monthly income that’s little above the welfare level, prostitution only makes sense if you’re from a poor country, work as an expensive high-end escort, or do it as a second job to supplement your income.

      1. That’s what I’ve heard as well. There tends to be a “gray market” element to the supply side of prostitution even in many countries where it’s legal. Personally, I’d just ban non-citizens from the prostitution trade, while putting in laws giving trafficked prostitutes immunity from deportation/prosecution if they agree to testify against their pimps.

      2. The improved conditions for the 200,000 sex workers should have some weight in the calculus. The status quo in a place like the US involves arresting many of them, and the Swedish solution still drives them underground (as the service that they’re offering is illegal to solicit.) In the “respect for the law” category, I’d also add the idea that unenforceable laws have consequences too, in terms of encouraging people to view the legal structure as illegitimate.

        1. You are assuming that the market did not expand as a result of legalization. For example, in Munich the number of known prostitutes increased from 1226 in 2000 to 2881 in 2011. Sweden probably has fewer than that. The Netherlands are currently actively trying to reduce the size of their red-light sector because it has become too big.

          Also, prostitutes often actively avoid actual employment. While this has benefits, it also means paying taxes and some also fear a loss of autonomy: in theory, brothel owners have only limited managerial authority (compared to normal employers) and cannot require a prostitute to have sex with a specific person or to provide specific services against her (or his) wishes. In practice, it apparently doesn’t always work out this way …

          1. That doesn’t change the improved conditions for the sex workers who were in the field before, does it?

          2. Really insightful comments as usual Katja, and it’s good to have some numbers to compare over time. Do we know how accurate the Munich numbers are and whether their accuracy changed as the result of an intervening policy change?

          3. Keith, I wish I knew more about that. Those numbers were provided by the Munich police, as quoted by a couple of Munich newspapers [1]. According to the city’s website, Munich operates a voluntary registration scheme for prostitutes. This is presumably beneficial for them, and prostitution itself was not illegal in Germany before (only pimping/brothels), so there wasn’t much of a disincentive to register that doesn’t also exist now (e.g., unwillingness to pay taxes). I expect that there is some room for error, but it doesn’t seem to be out of line with other observations about the growth in prostitution in Germany [2].

            [1] See also this more detailed infographic, provided by the “Welt”, but only covering the years 2000-2009 (with estimates for 2010).
            [2] One of the most worrying tendencies is that the growth seems to be primarily due to the growth of “cheap” prostitution, with all that entails about “working conditions” in the sex industry.

          4. Marc, it is very arguable whether the conditions have actually improved. Among other things, increased competition seems to have depressed prices (see the TIME article).

        2. The improved conditions for the 200,000 sex workers should have some weight in the calculus.

          As seems to often be the case in these sorts of discussions, people are objecting to claims that were never made. Keith never said that the improved conditions shouldn’t or didn’t matter. All he said was that legalization might increase the size of the illicit market. He never said that that meant that legalization was a bad idea or anything about the relative merits at all.

          1. Quite right J. Michael Neal. The post says (a) Here are new scientific data and (b) Public policy in this area is complicated. It is sad how many people are terrified by and angry at both (a) and (b), not just on this issue but as a general principle.

            These tea-party-esque attitudes are why I stopped reading comments on RBC legalization threads, other than those of people I know make good points (In this case that was you, Katja, Ed Whitney and Pat Oglesby). I am sure as a result I miss some good points by new arrivals with no track record, as well as by established members whose names I just miss as I rapidly scan the names of the commenters on these threads, and I really do regret that.

          2. If you state the negative and ignore the positive, which is what he did, you’re making a value judgment. Keith listed a parade of horribles on prostitution without a single mention of any benefit that could accompany legalization. If you’re pretending to weigh pros and cons then you present both.

            Words mean things. Omitting certain words implies things too. This piece was clearly meant to lead people to think that legalization of prostitution, and by extension marijuana, is a terribly dangerous thing. And there are, in fact, dangerous aspects. But the status quo has dangers too, and failing to explicitly acknowledge that is advocacy, just as ignoring any possible costs would be. I expect better from here.

          3. If you’re pretending to weigh pros and cons then you present both.

            Keith never pretended to weigh the pros and cons. He made a specific comment about one aspect of the issue. You are confusing a post about policy with a post about advocacy.

            And frankly there wouldn’t be any point in Keith (or Mark, who often gets accused of the same things) making an advocacy post on this blog. As far as I can tell, no one who needs convincing about the benefits of legalization reads this site. (Granted, private correspondence to the authors may have a different tilt than the comment threads, but that’s beyond my purview.) A post talking about the benefits of legalization would be a waste of pixels. Sometimes people have better things to do than to preach to the converted.

            Further, all of the front pagers here have published longer works that cover the issue far more broadly than this blog does. And you look really damned silly when you accuse them of believing things that are directly and explicitly contradicted by their public commentary. That you haven’t chosen to educate yourself about their beliefs before you run your mouth on the blog isn’t their fault. They certainly gave you the opportunity to do so, what with posting links to their publications on the front page.

            The idiot here is you.

  5. I’m missing something. In the cited LSE article,, I see a clear answer to only the question “Does legalized prostitution increase human trafficking?” The answer is Yes. That makes sense. An analogy would be that after alcohol Prohibition was repealed, imports of Scotch whisky into the United States increased. That makes sense, too.

    But I’m not finding a clear statement that the black market, as opposed to trafficking, will “expand.” An analogy would be that after alcohol Prohibition was repealed, the black market in alcohol grew in size – but it didn’t.

    Similarly, I can’t see from the Ekberg article,, cited in note 8 of the LSE article, whether the black market expanded or shrank after prohibition of prostitution. Both articles concern themselves primarily with trafficking, not the black market. These are separate evils for prostitution.

    Back to trafficking: The LSE article me leads to only this analogy for marijuana legalization: If possession were legalized in a state, “imports” from other states might increase – at least if all sales of marijuana remain illegal (as was the case in alcohol Prohibition, when possession was legal but sales were not). But I don’t see the analogy for the expansion of the black market. (Maybe I’m skimming too fast.)

    Pat Oglesby
    Center for New Revenue.

    1. I think they were saying that overall trafficking flows increased into countries with legalized prostitution, even when the percentage of the prostitution trade being done by trafficked people decreased due to the availability of legal commercial sex.

  6. I agree that this is an important phenomenon that belies the assumption of legalization automatically diminishing black markets. However, the current discussion of legalization we are having in America is not about prostitution or about drugs in general but about marijuana.

    There is an obvious difference in the human misery of the black markets for prostitution and pot, but I believe the most important differences in the context of this article are those of production of supply.

    Marijuana can be grown cheaply and abundantly, at home, and with little to no expertise. This challenges the idea that addicts might want to consume so much, or in such a way, that they have to rely on the black market.

    The possibility that the overall market may grow substantially after legalization of marijuana is disquieting. Despite this, I suspect there are only two reasons the black market would gain customers on net: a failure to fully legalize the drug or onerous regulation that makes it too expensive.

    I predict that marijuana legalization will indeed diminish its black market due to its unique characteristics of production. (So long as black market refers to violent drug cartels and not simply any production/sales outside of regulation, such as home grow operations.)

    I think the author’s argument is much stronger as regards to prostitution, gambling, or harder drugs. I hope those opposed to marijuana legalization do not stretch the argument beyond its merits.

  7. This can serve as a reminder that it is an oversimplification to treat elasticity as a single variable characteristic of a market. For addicts, demand may be inelastic, but for some markets, there may be a large number of participants who are not addicts, and the more of them there are, the more elastic the market demand may become. A guy who can live comfortably in God’s Wrath, South Carolina may fly to Vegas for a week and demand professional services not available at home, and then fly back and return to normal functioning. But if local laws changed and the services in question became available at prices comparable to those in Vegas, his demand would respond to the new market conditions. There would be few addicts in the local market and the demand would be elastic indeed. In Busted Hymen, Nevada, the demand is likely to be less elastic. A single measure of elasticity for the whole market could be misleading. It would be sort of like a mean treatment effect in a random-effects meta-analysis in which there is large heterogeneity of effect. The number would mislead and bad policy decisions could follow.

    1. Ed Whitney: You make good points. Economists tell me that with the term “elasticity” are really two types, market participation elasticity and then demand elasticity of established market participants. Legalization of a market could affect one, both or neither of these, which makes the net effect hard to forecast.

    1. Nope – that protection of a small coerced fraction of the total number of sex workers is being used as cover for moral disapproval of what the large majority of them choose freely to do.

    2. So you think that legalizing what was once prohibited can expand the black market?

  8. My local pot club has monthly purchase limits greater than the amount I could consume (were I still doing such things, which I’m not) in a lifetime. There are also countless other pot clubs within an hour’s drive, each presumably with its own, similar purchase limits. I’m not sure there’s any need for the illegal market to fill in any gaps.

    1. Andrew,
      Well, count your lucky stars. Were it that way everywhere, then war is over. As you’ve noted some places, it already is. For everyone else, it comes down to three things, I suppose: Location. Location. Location.

  9. And, presumably, the same applies to people who smoke, use drugs that you don’t, or who have abortions – right? I’m sure that we can come up with an even longer list of popular ideas for removing privacy from others if we put some effort into it…

  10. I don’t necessarily agree with Keith on these issues, but I think your line of thinking is overly simplistic here.

    I think paying for sex is generally immoral, yes. Not because paying money to “have sorts of fun somebody disapproves of” is immoral, but because sex that you pay for is usually the result of somebody exploiting an economic or personal dependency, and paying for sex allows that exploitation to be perpetuated.

    For the most part, I think that taking or not taking drugs is a personal decision, assuming it’s voluntary. What makes drugs tricky is that this choice may not actually be voluntary in the case of addiction [1]; also, while taking drugs may not cause personal harm, I have no fundamental objection to society taking proportionate measures [2] to minimize risks to other people that arise from the consumption of drugs (such as laws against driving under the influence, or against the facilitation of addiction [3] in minors).

    Looking at adverse effects on the rest of society is not in any reasonable meaning of the word “bias” regarding the morality of the underlying basic activity. Looking for such effects may be motivated by such bias, but the effect of an activity on oneself and the effect of the same activity on others are distinct concerns.

    [1] Even so, I’m reluctant to approve of interference absent significant actual physical harm, though I have no problem creating economic disincentives such as taxes, prohibiting advertising, etc. to proactively limit the effects of drugs with an unknown risk factor on the rest of society.
    [2] To be clear, I consider the War on Drugs to be hugely disproportionate and actively harmful.
    [3] In the specific context of minors, I mean addiction to include psychological and not just physical addiction.

  11. The study compares Sweden with Germany and Denmark, two countries having very different cultures compared to their Nordic neighbors.

    Sweden is a matriarchal culture. Its government is very right wing. Swedish culture sees prostitution as strictly a form of female exploitation. Swedes are therefore more likely to detect or pay close attention to anything they think harms women, as Julian Assange discovered to his misfortune.

    Even so, when prostitution was outlawed in Sweden in 1999, according to the study, “the number of prostitutes in Sweden decreased rather substantially from 2,500 in 1999 to 1,500 in 2002, with street prostitution in particular decreasing by between 30-50%.” That may seem like a big decrease, but it didn’t end prostitution, and may have driven much of it underground, as the researchers admit. Also, I wouldn’t call the numbers substantial, especially when a major sporting event like a World Cup soccer match in Germany can draw 40,000 prostitutes from all over the world.

    Thus I’m not surprised Sweden has fewer exploited, foreign sex workers per capita than does Denmark or Germany, prohibition of prostitution notwithstanding. When we’re discussing a culture, government, and climate such as Sweden’s, it doesn’t take much urging to get people to go somewhere else for their fun.

    1. Servetus wrote:
      “Swedish culture sees prostitution as strictly a form of female exploitation. Swedes are therefore more likely to detect or pay close attention to anything they think harms women, as Julian Assange discovered to his misfortune.”

      I don’t think you meant to imply that the L’affaire Assagne has anything to do with prostitution, which miscasts what he’s accused of, as well as his accuser. I will agree that Sweden’s rather staid and at the same time extreme legal definition of something akin to what we’d call domestic sexual battery colors the whole affair, beyond the obviously suspicious circumstances of it coming to light. I’m certain there will be some resolution in some form, but I’m equally certain it will not involve prostitution.

      1. Yes, you’re correct. Assange probably didn’t know he was committing a crime by not wearing a condom without the permission of his partner. Or that his transgression would be used against him by two close allies, the U.S. and Sweden. He may have been set up. And no, it doesn’t have anything to do with prostitution.

        It’s an example of a cultural differnce being used to entrap Assange, and an example of what happens when bare statistics ignore unique cultures and their effects on international commerce of any kind, whether it’s prostitution, drug trafficking, or selling Bentleys to Arab sheiks.

        1. Thanks for the reassuring clarification.

          I think you have it pegged just about right regards the issue between Assange and Sweden. One would hope that “intent” has some function in Swedish law, should it come to that, given “ignorance of the law” is so frequently used against the poor in the good ol’ USA — unlike all those banksters and financiers who should know better, but needn’t worry about it.

          1. Its a pity that no one has asked sex workers to have any input into this debate and research. Swedish sex workers (Rose Alliance) will tell you that in the years quoted as seeing a 30-50% drop in street based work the internet became widely available in Sweden as a means of advertising.In this period many Swedish street based workers left the street to work via the internet. Think about it, its very cold in Sweden in the winter. The same thing has been happening in New South Wales where I work. As technology such as the internet and mobile phones became widely available street based work has decreased from being a major part to a very minor part of sex work with now their being only a handful of street based workers in an estimated population of 10,000 sex workers.

            As to 40,000 sex workers flying in to a major sporting event. Get real, this myth that the press trots out for very major sporting event has been comprehensively debunked elsewhere. That’s about 140 wide bodied jets just to get the sex workers to a major sporting event. If you ask sex workers themselves about sporting events they will tell you that if anything major sporting events usually lead to a reduction in business, not an increase.

  12. Once again, great insights to a difficult topic, albeit the cringe-worthy, boilerplate of pro/con of legalization of marijuana
    that almost always infiltrate any topic on this forum.

  13. This article and most of the comments above are flawed in that they do recognize the difference between legalization and decriminalization.
    I will speak only to sex work which is an area in which I have experience. Legalized frameworks means that an activity is specifically regulated whilst decriminalization means no criminal laws or regulations specifically control the activity.
    Any legalized framework leads to a two tier system with one tier accepting the regulation and another tier going underground and avoiding the laws regulations of the legalized framework.
    I suggest that you look at the decriminalized legal frameworks that exist for sex work in New Zealand and New South Wales and you will find that demand is not highly elastic in sex work as stated.

  14. It should not be a surprise that sex workers move to live and work in countries that offer more attractive environments, where our work is legalised or decriminalised, in preference to illegal options. What most of the commentators here fail to recognise is the agency we apply in choosing where and how we work.
    And, to state the obvious, people within economies can’t be compared with products. Unlike an inanimate shipment of marijuana that is farmed for a market, exported, distributed, sold, rolled and enjoyed, without having exerted any control over it’s destiny or asociated conditions, we sex workers are real people with real lives and part of a reality-based community, something the commentators here would have you ignore.

  15. There is really no logical reason to legalize prostitution. The legalization of prostitution will increase the availability of prostitutes and lower the cost of their services. This would then increase demand, drawing even more under aged girls and boys into prostitution. As a society we have an obligation to try and limit the amount of people involved in a job that is basically sexual assault for money.

Comments are closed.