Arrest the drugs

Reason’s Radley Balko writes a good essay on some of the unintended consequences of the ‘war on drugs’.

The most common arguments against the drug war is that it leads to more crime and more deaths and yet doesn’t achieve much change in behaviour. Balko looks at some other issues, including the militarisation of the police force, the undermining of medical treatments, the costs of having 350,000 non-violent drug criminals in jail, the watering down of the rule of law and the negative impact on foreign policy.

But in a piece of good news, Massachusetts has recently changed their laws so that possession of up to an ounce (28 grams) of marijuana is no longer a criminal offense. Despite the predictions of doom from the fear-mongers and bureaucrats, 65% of voters decided to ease up on the evil weed. As they should.

12 thoughts on “Arrest the drugs

  1. Radley is normally one of my daily reads over on his website; “The Agitator,” where he deals with this subject and others. If you go back through it, you will understand that what is in this one is just the tip of the iceberg.

    Those druggies are pretty scary, but I think the authorities need cleaning out first.

  2. It would good if you could come up with some economic model of the effect of different drugs. I sometimes ask HC for this when he argues the opposite, but I’m yet to see anything concrete, especially in an Australian context. All most people give is “drugs are bad” or “the war on drugs costs lots of money”, both of which are no doubt true, but they obviously trade-off against each other.

    I think it would great to have a little computer model for each drug (or at least the main ones), where you could have a parameterized model of the main factors that cause drug prices/usage to go up and down and their costs. That way people could stick in what parameters they believe are correct for things like supply and demand and how much various aspects of the war on drugs costs, and then see if more users equals less costs. For some drugs, I imagine some of those parameters could be pretty well estimated based on difference between states (e.g., compare states where decriminilization has occurred versus ones where it has not).

    I think this would be really interesting — there must be some drugs that have a decent economic cost (say, ecstasy, where you get brain damage from too much, and I assume is moderately easy to police), versus others where there is probably fairly little cost (say, GHB, where you get little damage, and must be also fairly easy to police). Others would probably be somewhere in between (e.g., pot, where you get lung cancer, but mainly later in life, but it is hard to police). These of course all depend on how much you think it costs to police these things and how much the drugs effect the individual — but if you allowed people to stick in their own parameters they could see the costs/benefits for themselves.

  3. How is ecstasy easy to police?

    Friedman suggested that smaller drugs are relatively easier to smuggle, and so the war on drugs changed the competitive situation for drugs in favour of smaller drugs (eg ecstacy, cocaine etc) and against larger drugs (eg marijuana).

    There is also the point that in a free society, people must be free to make mistakes. Even in North Korea, people are free to do exactly as they are told. That’s not real freedoom.

    As for modelling drugs — that could be done for a few of the basic points. There are estimates of elasticities of demand. And it is possible to measure the impact that policing has on the retail price of drugs by comparing different countries with different levels of enforcement (eg price of cocaine in Columbia v America). But most of the issues would be hard to model, such as the impact on organised crime, police corruption, undermining the rule of law, relative changes in drug quality and consistency, diversion of police resources, impacts on people who were jailed for drug offenses etc.

    Instead of building a model, it probably makes more sense to do a benefit-cost analysis. This has been done before. Though there are so few clear benefits from drug prohibition that it hard to see how it could even get close.

  4. JH: My guess was that policing ecstasy is comparatively easy fo ecstasy — as you can use strategies like harassing kids going to dance parties (although I’m aware that this strategy itself changes behavior over the medium term, as happened in the UK). Alternatively, this won’t work for pot, as you can grow your own pot too easily and the usage is far more widespread. I could be completely wrong on that — perhaps it’s harder to police, it was just an example.
    I don’t doubt drug usage would be hard to model — but I imagine that having a page where you can stick in your own parameters (like those economy or climate change simulators) would be far convincing to many people that a simple cost-benefit analysis. If you made it open source, people could start adding their own chunks. So if you were, say, an expert on policing, you could add your button for the cost of that, or if you were an expert on lung cancer, you could add the cost of that.

  5. Costs are never static. A model of the economy that assumes that prices are an input to the economy rather than a product of the economy is a deeply flawed model.

    Are people living in Massachesetts not still subject to federal laws?

  6. As I understand things in Australia having small quantities of dope is not criminal. It is however still illegal in the way that parking in a no parking zone is illegal. In my book it should be fully legalised with the same status as tobacco. Even though the regulation that such legality entails would probably end up making the product more expensive.

  7. TerjeP, the people who make such assertions about the nature of prices are most likely Keynsians or worse. If that is not bad enough, Keynsianism is considered mainstream economics.

  8. Terje — drug laws are done State by State in Australia. The ACT and SA decriminalised marijuana a while ago. SA has since re-criminalised marijuana. In the ACT you get a $100 fine for possessing small quantities of weed.

    conrad — the more interesting thing I took from your comment is the idea of an open source “wiki-model”. I think models can add value to a discussion, but there is always the fear that people will take them too seriously. I’m not sure why you call a benefit-cost analysis “simple”.

  9. In the US Obama has backed away from his implied tolerance, and Californian medical marijuana suppliers have been raided by the feds and charged although the medical use is legal under state law. There are also reports coming in from other states where medical marijuana is legal of federal police raids. This has been going on in the past as well.

    I have my doubts about economic modeling, certainly there could be benefits from knowledge being out there as long as it was done by an unbiassed source. It probably should not be ‘open source’ as that would entail constant rectification after idiots put their own ideas in. Any ‘economic modeling’ should not be done by the governments as the model would be Keynesian, and the recommendations Dicensian.

  10. JH: I mean comparatively simple compared to a full blow model which tries to look at effects across time — I’m sure even a very simple cost/benefit analysis for something like drugs is a pretty decent undertaking.

    Also, I don’t doubt that people take models too seriously in many cases — but I think such a model would generate a lot of discussion, since people would be more likely to consider multiple effects at once, rather than just “drugs are bad” or “effects of the drug war are bad”. You can imagine starting off with a few factors, but gradually adding to these and refining the model over time. In addition, if people don’t like, say, smokers-die-earlier style or brain-damage-leads-to-social-problems style arguments, they could simply take these out of the model and compare models with and without them.

    I think this would be really handy because, for all I know (and no doubt many other people have the some thought), for some drugs, the increase in usage may already trade off favorably with policing costs over time, and we needn’t even worry about more complex factors if we are simply after cost minimization. At present, I think there is a distinct lack of public information about these types of things.

    Despite this, there are certainly information sites like,

    for example, but they don’t allow people to really understand how things are contributing. For example, what is the cost of an extra HIV infection? Probably pretty decent over a life-span (that could be a parameter).

    On this note, I know, for example, a needle-exchange program in Canada did some analysis of that, and it ends up huge. With a nice little model, you could simply plug this in as a cost, and see how it trades off with creating more users (not that I think needle-exchanges cause that — I think only wacky conservatives believe that — but they could just stick in a big parameters for this, and I could use 0. Perhaps even with a big parameter, the trade-off is still positive in favor of the needle exchange). This sort of thing would allow people to get some idea of how big the extra demand created would really need to be to trade-off with the health benefits.

    Thus, you would some decent answer to what the cost is. So, for example, if a needle exchange causes usage to increase 1%, but HIV (and whatever else) to drop, it might still be far cheaper to have the exchange. You could then ratchet up the parameter to see what how much demand would need to increase to cause a break-even with the cost saved on disease prevention.

    This info would be very hand, because if some conservative comes along and says needle-exchanges create users, you could simply counter that by saying, well, they don’t create the 20% more that you would need (or whatever the real number is) to pay for the medical costs of not having the exchange. They then have to argue that there is something else intrinsically bad about using that makes paying the extra cost worthwhile.

  11. You can never predict all the consequences.
    The US drug war has had such consequences as conflicts between their military programs and their drug enforcement. Who could have predicted stuff like that 30 years ago when Nixon announced the war on drugs?

    This issue should be opposed on an ethical basis if you actually want to understand fundamental principles and change people’s minds. Because this understanding is what’s lacking in our culture. It’s easy to find negative consequences to enforcement out there, yet I still have people telling me that it would be unethical to legalise drugs because it “effects others”. In reality it is unethical to ban drugs as this initiates force. And of course goverment restrictions on freedom (necessary for human survival and prosperity) sure as hell “effect others”.

    Government initiated coercion and freedom restriction is destructive to human life. Simple.

Comments are closed.