We’ve come a long way since Reefer Madness. Over the past two decades, 16 states have de-criminalized possession of small amounts of marijuana, and 22 have legalized it for medical purposes. In November 2012, Colorado and Washington went further, legalizing marijuana under state law for recreational purposes. Public attitudes toward marijuana have also changed; in a November 2013 Gallup Poll, 58 percent of Americans supported marijuana legalization.

Yet amidst these cultural and political shifts, American attitudes and U.S. policy toward other drugs have remained static. No state has decriminalized, medicalized, or legalized cocaine, heroin, or methamphetamine. And a recent poll suggests only about 10 percent of Americans favor legalization of cocaine or heroin. Many who advocate marijuana legalization draw a sharp distinction between marijuana and “hard drugs.”

That’s understandable: Different drugs do carry different risks, and the potential for serious harm from marijuana is less than for cocaine, heroin, or methamphetamine. Marijuana, for example, appears incapable of causing a lethal overdose, but cocaine, heroin, and methamphetamine can kill if taken in excess or under the wrong circumstances.

But if the goal is to minimize harm — to people here and abroad— the right policy is to legalize all drugs, not just marijuana.

In fact, many legal goods cause serious harm, including death. In recent years, about 40 people per year have died from skiing or snowboarding accidents; almost 800 from bicycle accidents; several thousand from drowning in swimming pools; more than 20,000 per year from pharmaceuticals; more than 30,000 annually from auto accidents; and at least 38,000 from excessive alcohol use.

Few people want to ban these goods, mainly because while harmful when misused, they provide substantial benefit to most people in most circumstances.

The same condition holds for hard drugs. Media accounts focus on users who experience bad outcomes, since these are dramatic or newsworthy. Yet millions risk arrest, elevated prices, impurities, and the vagaries of black markets to purchase these goods, suggesting people do derive benefits from use.

That means even if prohibition could eliminate drug use, at no cost, it would probably do more harm than good. Numerous moderate and responsible drug users would be worse off, while only a few abusive users would be better off.

And prohibition does, in fact, have huge costs, regardless of how harmful drugs might be.

First, a few Economics 101 basics: Prohibiting a good does not eliminate the market for that good. Prohibition may shrink the market, by raising costs and therefore price, but even under strongly enforced prohibitions, a substantial black market emerges in which production and use continue. And black markets generate numerous unwanted side effects.

Black markets increase violence because buyers and sellers can’t resolve disputes with courts, lawyers, or arbitration, so they turn to guns instead. Black markets generate corruption, too, since participants have a greater incentive to bribe police, prosecutors, judges, and prison guards. They also inhibit quality control, which causes more accidental poisonings and overdoses.

The bottom line: Even if hard drugs carry greater health risks than marijuana, rationally, we can’t ban them without comparing the harm from prohibition against the harms from drugs themselves. What’s more, prohibition creates health risks that wouldn’t exist in a legal market. Because prohibition raises heroin prices, users have a greater incentive to inject because this offers a bigger bang for the buck. Plus, prohibition generates restrictions on the sale of clean needles (because this might “send the wrong message”). Many users therefore share contaminated needles, which transmit HIV, Hepatitis C, and other blood-borne diseases. In 2010, 8 percent of new HIV cases in the United States were attributed to IV drug use.

Prohibition enforcement also encourages infringements on civil liberties, such as no-knock warrants (which have killed dozens of innocent bystanders) and racial profiling (which generates much higher arrest rates for blacks than whites despite similar drug use rates). It also costs a lot to enforce prohibition, and it means we can’t collect taxes on drugs; my estimates suggest U.S. governments could improve their budgets by at least $85 billion annually by legalizing — and taxing — all drugs. U.S. insistence that source countries outlaw drugs means increased violence and corruption there as well (think Columbia, Mexico, or Afghanistan).

It’s also critical to analyze whether prohibition actually reduces drug use; if the effects are small, then prohibition is virtually all cost and no benefit.

On that question, available evidence is far from ideal, but none of it suggests that prohibition has a substantial impact on drug use. States and countries that decriminalize or medicalize see little or no increase in drug use. And differences in enforcement across time or place bear little correlation with uses. This evidence does not bear directly on what would occur under full legalization, since that might allow advertising and more efficient, large-scale production. But data on cirrhosis from repeal of U.S. Alcohol Prohibition suggest only a modest increase in alcohol consumption.

To the extent prohibition does reduce drug use, the effect is likely smaller for hard drugs than for marijuana. That’s because the demands for cocaine and heroin appear less responsive to price. From this perspective, the case is even stronger for legalizing cocaine or heroin than marijuana; for hard drugs, prohibition mainly raises the price, which increases the resources devoted to the black market while having minimal impact on use.

But perhaps the best reason to legalize hard drugs is that people who wish to consume them have the same liberty to determine their own well-being as those who consume alcohol, or marijuana, or anything else. In a free society, the presumption must always be that individuals, not government, get to decide what is in their own best interest.