Risk, Liberty, and Drugs: A Response to Theodore Dalrymple

Theodore Dalrymple’s provocative four posts outlining a case against the legalization of drugs provide an interesting contrast to the contemporary momentum in the Western world today toward relaxing prohibitions on substances long deemed too dangerous for public consumption.  This is particularly true in the United States, where three states and the District of Columbia are currently experimenting with heavily regulated, legal cannabis consumption among adults.  Dalrymple says that his arguments fall into two broad categories, philosophical and practical.  Here I’d like to address his philosophical arguments directly.

Dalrymple claims that John Stuart Mill’s individual sovereignty argument serves as the basis for “libertarian” arguments in favor of legalization. Mill consistently argued in favor of individual sovereignty, albeit with many limitations, throughout his work, and Dalrymple fashions this position as a sort of foundation for the advocates who support legalization.

In truth, much of his first post does not focus on philosophy, but rather explores the policy implications of what he claims Mill’s position seems to imply. Dalrymple argues that advocates of drug legalization have to address the consequences of drug use for those in society who choose to continue to abstain. For example, what limits should be placed on the age of consumers? What are the potential social costs, to borrow a term of Ronald Coase; and who is to bear those costs? How might society deal with the possible increases in healthcare costs through, presumably, more usage and more abuse in a world with pooled risk health insurance (a strange argument since we already bear those costs through illegal use)?

He concludes by arguing that what he calls Mill’s “very simple principle” actually does not provide simple criteria for either allowing legalization or drawing lines about what kinds of limits should be placed on usage.

To begin, I find this to be a fairly shallow argument when it comes to claims about social costs. Perhaps Dalrymple is unaware of it, but libertarians have long acknowledged that liberty involves not the individual but the individual in society. Take for example Ludwig von Mises’ classic book Socialism (1922), in which he says explicitly that, “in so far as each individual acts as a member of society he is obliged to adapt himself to the will of his fellows. In this way no one depends more upon others than others depend upon him.” Because of the social nature of liberty, adds Mises, we are all dependent on the approbation of others. Mises claims that even the “oriental despot” must consider the consequences of his actions because he, too, can be overthrown.

In short, none of us lives in a vacuum, and the classical liberal tradition has understood this for a long, long time. Ronald Coase’s revolutionary work on social costs addresses many of the issues Dalrymple raises. Economists have worked hard to focus on this set of problems, and they have a pretty good toolbox to deal with them.

Additionally, while Mill’s principle is important to the liberal tradition, and serves as the basis for many of our political discussions about the centrality of the individual in the Western world, I would argue that Dalrymple is not considering a much wider range of philosophical arguments consistent with legalization that classical liberals and liberals in the European tradition support. I will discuss three other principles—paternalism, polycentricity, and the precautionary principle—that should inform public debate about whether or not to criminalize drugs.

At its core, drug criminalization is a paternalistic policy. Paternalism is the notion that someone knows what is better for another person than that person knows for his/herself. This concept has been very sexy in public policy circles since the publication of Richard H. Thaler and Cass R. Sunstein’s well-known book, Nudge: Improving Decisions about Health, Wealth, and Happiness (2008). Thaler and Sunstein defend what they call “libertarian paternalism,” or a sort of allegedly benign form of incentivizing individuals, through small institutional changes, to make “better” choices for themselves. Perhaps the most famous example of this would be a default “opt-in” for voluntary retirement contributions to employer-provided pension plans.

Historically, libertarians and classical liberals, as well as liberals in the European tradition, have looked skeptically at any claim that those in positions of power know what’s best for others. F.A Hayek made the case clearly that not only does liberty depend on a wide range of individual autonomy, but also that planners simply cannot ever know the wide range of local knowledge and preferences possessed by individuals in a society. Furthermore, aggregating that information into a coherent public policy is often impossible because of what Mises called the “socialist calculation problem.” Even if we could know what individuals would truly prefer (which we cannot), we cannot ever aggregate those preferences into a policy to mandate outcomes.

Of course one could push this argument, as Dalrymple does, to say that no one reasonably believes that four-year-olds should be allowed to take cocaine or that parents should be allowed to be “incapably drunk in the charge of their children.” In that same vein, we do not allow individuals to lock their children in heated cars with the windows rolled up or mothers to take care of their children while incapacitated. But we obviously do allow parents to leave their children unattended in, say, other parts of the home, or to consume alcohol with their children in a manner that still enables them to perform their functions responsibly. Banning alcohol consumption among parents would likely drive the human reproduction rate close to zero outside of Utah and the Muslim world.

One cannot make the heroic leap from moderate consumption to complete incapacitation unless one is arguing that all drug use immediately leads to complete loss of capacity or chemical addiction. Liberalism is based on the concept that individuals are both inherently imperfect and capable. The Western tradition has long grappled with how to make society “better” according to many criteria, starting of course with Plato. However, there is a distinct difference between arguing that people will make poor choices and therefore must be constrained completely from a choice, and acknowledging the risks associated with a choice while punishing abuse or excess.  Drug criminalization errs on the side of paternalism in contrast to comparable application in other policy areas.

The second philosophical idea that should be relevant to legalization discussions is polycentricity, a term coined by Michael Polyani and later used by Vincent Ostrom to describe multiple, overlapping political and social orders. Ostrom and his wife Elinor were students of federalism, a key facet of the American governing structure.

The United States has multiple layers to its government, starting at the local level moving up to counties, states, and finally the national government. The underlying principle behind federalism is that local governments and individuals are more in tune with the conditions in their communities, the preferences of their fellow citizens, and the available solutions to local problems than are central planners or national bureaucrats. Systems with polycentric orders compare favorably with the blunt instrument of centralized power because they recognize that citizens group together based on shared values. Towns and cities are inherently different.  Just as my church going in laws living the rural Midwest would be unlikely to support the legalization of drugs in their town, Americans in the Haight Ashbury neighborhood of San Francisco would be likely indifferent or support legalization.

Finally, the precautionary principle. This is a sort of cousin of paternalism, but on the aggregate rather than individual level. We frequently see the precautionary principle applied when it comes to the review processes of the Food and Drug Administration and in contemporary discussions of environmental regulation. Because of the perceived risks of allowing for certain behaviors, government has an obligation to curb social activities and prevent the possible consequences.  Advocates for green energy, carbon taxes, regulation of greenhouse gases by the Environmental  Protection Agency, and higher gas taxes cannot say absolutely that all of these positions will prevent environmental catastrophe. They argue instead that the possible risks are much higher than the cost of such limits on individual liberty.

Much as we see in the global climate change debate, discussions of drug policy that are dominated by the precautionary principle are fraught with uncertainty and veer sharply away from individual liberty. In both instances we simply have limited knowledge about aggregate risk. In the case of climate change, we do know certain human behaviors emit carbon, but we must rely on computer simulations and estimates to predict outcomes. Policy is derived not from concrete experience but from expert analysis.

Drug policy is much the same. We do not know what a world with legal drugs would look like.  We have examples of limited decriminalization, such as Portugal, but large-scale efforts have never occurred. Moreover, social costs, the potential increase in usage, and changes in the social order are based on limited data, and on expert opinion that sometimes isn’t very expert. While Dalrymple rightly notes that Portugal’s experiment has not been without costs, he fails to note that Colorado has seen a decline in its crime rate since legalizing cannabis.

The philosophical arguments in favor of legalization, or at a minimum, allowing for decentralized experiments to increase the autonomy of individuals and get a better handle on risk, are far broader and more well-founded than Dalrymple claims. And while there is no doubt that legalization is not a panacea, it is not a flimsy position based solely on an expansive and naïve claim of individual autonomy.