A couple days ago Vice Squad mentioned self-exclusion as one component of a policy regime that can aid problematic gamblers, and noted the possibility that forms of self-exclusion (perhaps instituted as part of broader consumer licensing policies) could be applied to other vices. Even if self-exclusion does not work all that well, it has two appealing features, namely, it is voluntary and it does not impose on gamblers who are uninterested in being excluded.
I have tried to learn a little bit more about gambling self-exclusion, and have been aided by two articles: "Casino Self-Exclusion Programmes: A Review of the Issues," by Nadine R. Nowatzki and Robert J. Williams (18-page pdf here), published in July 2002 in International Gambling Studies; and, "Self-exclusion: A Proposed Gateway to Treatment Model," by Alex Blaszczynski, Robert Ladouceur, and Lia Nower, in the April 2007 issue of the same journal. Some things I learned from these articles:
(1) Gambling self-exclusion is quite recent, with the first formal program adopted in Canada in 1989. [Update! -- Looks like Austria got there, oh, more than a half century earlier.]
(2) Lots of folks who self-exclude from a gambling location breach their agreement by returning (sometimes disguised) to the location and gambling. Compliance with self-exclusion agreements is higher in the Netherlands, because entrants to casinos in the Netherlands must present identification, which can be checked against the self-excluded list. It also appears that the self-exclusion program in the Netherlands is accessed by a higher proportion of problem gamblers than are similar (but not identical -- there is lots of variation among self-exclusion rules) programs elsewhere.
(3) Despite the enforcement problems, self-exclusion programs are associated with some beneficial outcomes. Blaszczynski, Ladouceur, and Nower (page 62) describe a recent evaluation of one self-exclusion regime: "Participants reported that, at follow-up, the urge to gamble was significantly reduced while the perception of control increased significantly for all participants. The intensity of negative consequences for gambling was significantly reduced for daily activities, social life, work and mood."
(4) Though exclusion programs are designed to help those who have self-control problems with gambling, the decision to enter an exclusion agreement can itself be made in an o'er-hasty fashion, perhaps after suffering a significant gambling loss. Blaszczynski, Ladouceur, and Nower (page 69) suggest that gamblers who insist on immediate exclusion be provided with a 24-hour exclusion while their case is under consideration, thus giving them time to cool-off before a decision is taken regarding a longer-lasting exclusion.