Well, for free under very controlled circumstances. From a blog post by “M.S.” on The Economist:
But we do have a good idea of how to stop more people from destroying themselves specifically with heroin injection, which has a higher fatality rate than most controlled substances. As with most drug problems, the solutions involve decriminalisation and universal access to treatment programmes, including alternative blocking drugs like methadone and buprenorphine. In the case of heroin, there is also another proven way to reduce harm: setting up safe injection rooms monitored by healthcare staff, and—for registered addicts who cannot or will not comply with treatment regimes—providing heroin itself for free.
Switzerland and the Netherlands pioneered this “Heroin Assisted Treatment” (HAT) approach in the 1990s, and both countries adopted it as national policy in the 2000s. Heroin use has steadily declined since; by the 2000s the Dutch incidence of new heroin users had fallen to essentially zero, and the aging population of addicts from the ’70s and ’80s continues to shrink. The average age of Dutch heroin users rose from 34 in 1997 to 45 in 2009.
Decriminalisation of marijuana use has also played a role in shrinking Dutch heroin use, since it separates the use of cannabis from the use of harder, more restricted drugs. HAT trials have since been run in Spain, Britain, Germany and Canada. The evidence consistently shows that HAT drastically reduces heroin-related crime, since addicts don’t need to steal to get money for their fix, and it slashes heroin-related deaths and HIV infection, since addicts are shooting up under medical supervision.
More interestingly, HAT is also correlated with lower overall heroin use. This is in part because free government heroin tends to drive out private-sector providers. Most addicts will end up shooting up in safe rooms monitored by public-health staff, where they will be encouraged to enroll in a treatment programme or, if they fail or refuse treatment, simply receive free heroin. This gradually erodes the market for dealing heroin for profit; as they say in the tech world, you can’t compete with free. The result is what you see in the Netherlands: the slow disappearance of heroin use.
As I understand it, the idea is not to make heroin legal in the sense that anyone can walk into their friendly neighborhood heroin shop and buy some, but a legal drug that addicts can acquire from designated places with a prescription.
In comments, M.S. added:
Heroin use rose across Europe while it fell in the Netherlands and Switzerland. There’s essentially no policy disagreement among drug experts that harm reduction saves many lives and vast amounts of money without increasing usage. The initial application of HAT is this: once you’ve provided universal free treatment and gotten all the addicts pulled into the system through safe injection sites, what do you do with those last recalcitrant users who can’t or won’t quit and are underwriting the private heroin market? You give them free heroin. It doesn’t increase usage because people have to go through the wringer before they start getting the free stuff, and it likely reduces it by putting dealers out of business.
Meanwhile, you clearly have your own pre-existing, unjustified prejudices if you demand that HAT reduces usage before you’ll consider it. Many people, *clearly including Philip Seymour Hoffman*, are able to live reasonably happy, productive lives while regularly using heroin. The harm they do by using is 1. epidemiological, 2. financing illegal drug trade, and 3. the risk they will unintentionally die, hurting themselves and their loved ones. Giving these people free heroin at controlled sites fixes their problems even if it doesn’t reduce usage at all; unless you can prove it *increases* usage, which runs against all the available data in every trial run so far, I can’t see what the argument against it is. The amount of money HAT saves society is generally estimated in the range of $5k-$10k per patient per year.