Standard economic explanations of addiction (such as the rational addiction model of Gary Becker and Kevin Murphy) do formally allow for drug users to invest in rehabilitation. They can spend money to undo the effects of an addiction – e.g. Mel Gibson can devote time, effort and dollars to dealing with his alcohol problems in an outpatient program attach to a rehabilitation clinic. Unfortunately Becker and Murphy do not analyse the implications of such treatment technologies.
For example, what happens if the cost of treatment becomes very low and the probability of successful treatment very high? Suppose, in the limit, that you can undo an addiction at zero cost. Two situations arise (i) when the addiction can be undone as well as the harmful medical/psychological effects of having used the addictive substance for a sustained period and (ii) when the addiction can be undone but not the harmful effects of past cumulative use.
In case (i) optimal addictive consumption involves heavier use of the drug followed by immediate detoxification, with this coupled process being repeated endlessly. Higher steady state levels of use of a drug will be generated by the existence of a zero cost cure. Bulemia nervosa might almost fit into this pattern of consumption with binge eating followed by the 'cure' of purging. I say 'almost' because cumulatively this pattern of consumption is presumably damaging – high mortality is associated with binging. But individuals might themselves see purging as a way of undoing the effects of binge eating at low cost.
In case (ii) damage is generated as you cumulatively consume some product. For example as you continue to consume alcohol you suffer liver damage. Here the damage is cumulative but the extent of additional damage is not increased significantly by a single additional usage. This is a problem in the economics of procrastination and self control. An additional single usage does no substantial damage in itself and yields much pleasure so the incentives are to procrastinate. The appropriate response by an agent depends on how sophisticated they are with respect to their self-knowledge of their incentives to procrastinate.
If they realise they will always procrastinate they might invest in some precommitment technology that will make future use impossible or costly – for example they will go bushwalking in a forest where the drug is unavailable. Alternatively they might be able to formulate an internal personal rule that they will never again use. Their incentives to do these things are based on their knowledge that they will otherwise procrastinate. On the other hand, if they know now that they will always be tempted and are unable to commit to a technology that prevents them from using or an internal personal rule they might immediately abandon all future attempts to cease using and just continue using for ever. Why bother they will tthink - we are weak-willed and it is pointless to try to limit use. This is a very standard story: See for example Ted O’Donoghue and Matthew Rabin.
If they are myopic (and just maximise current benefits) they will use and then detoxify repeatedly and, in the longer-term, do themselves damage.
The general finding is that regardless of how sophisticated they are with respect to the self-control problems they face, if there is no pre-commitment technology or internal commitment device that forces them to stop using, they will continue using at a high level indefinitely. Developing treatments that deal with addiction problems may therefore lead to sustained higher levels of consumption indefinitely and thereby make an individual worse-off longer-term.
These types of stories have an empirical correlate in the advertising literature. David Prentice points out to me a recent paper by Toshiaki Lizuka and Ginger Zhe Jin which shows that advertising the availability of tobacco cessation products increases the tendency of people with a college education to smoke. The idea is coincident with that expressed above - if you know you have access to cheap means of abandoning a harmful addiction you may have incentives to use more.