In stating an argument for heroin prohibition, and in considering the justifiability of other forms of drug control, I have made a number of empirical assumptions that may fairly be questioned. Indeed all my key empirical assumptions might be false, and even if they are true, it is natural to wonder whether we are actually justified in believing them. This is ultimately a matter for social science to determine. But some may wonder, how much evidence must we have in general, and what kind of evidence, to be justified in supporting coercive government policies of this kind?

Coercive government policies can be justified in general only by good reasons, and good reasons are constituted only by true propositions. A person is justified in believing a proposition true when the information available to her provides her with good epistemic reasons to believe it and these reasons outweigh the good epistemic reasons to doubt it that are also provided by the information available to her. Being justified, however, in believing a proposition that supports a policy does not suffice to justify one in supporting it. One must also be justified in believing that the reasons in support of this policy outweigh the reasons against it. In my opinion, one is justified in believing a judgment of this kind when it withstands one's critical scrutiny: when this judgment about the relative weight of reasons continues to seem correct over time and to cohere with the other judgments one makes about the relative weight of reasons.