I agree with Eugene Volokh that for some problems there is no solution that will do more good than harm. While I think he understates the limited but real prospects for reducing violence through controlling access to weapons by dangerous people, I also think that others overstate those prospects.
But it seems to me that Eugene is plainly wrong when it comes to alcohol, which he uses as a comparison case in arguing against gun controls. We could easily make substantial reductions in alcohol-related violence – and accidents, and health damage to drinkers – with a few straightforward and administratively feasible policy changes without major unwanted side-effects.
First, we could raise taxes. Doubling the federal alcohol tax from the current ten cents per drink to twenty cents would reduce homicide and automobile fatalities about about 7% each, saving about 3000 lives per year. It would cost a two-drinks-per-day drinker (at about the 80th percentile of all drinkers about $6 per month. (Fully internalizing the external costs of drinking would involve taxes nearer a dollar a drink.)
Second, we could make it harder for people who break the law when they get drunk to continue drinking, either by subjecting them to alcohol testing in programs such as South Dakota’s 24/7 Sobriety (twice-daily alcohol testing on probation) or by creating a “do-not-serve” list of convicted drunken drivers and drunken assailants and requiring alcohol sellers to check customers against that list (or putting a “do-not-serve” marking on the convicted person’s driver’s license).
Neither of these approaches would eliminate the drinking problem, or even the narrower drunken-violence problem, but the combination would substantially reduce them. And yet so far it has proven virtually impossible even to get such obviously rational policies on the political agenda; “drug warriors” and “drug policy reformers” alike remain stubbornly indifferent to means of reducing the damage done by the one intoxicant we’re not currently making war on.