One of the most unusual and criticized theories in Economics is that of rational addiction, which states that smokers and druggies choose their addiction under full rationality and information. This can be rather surprising to a non-economist, but one can find data that supports (well, does not reject) such theory.
Timothy Hinks and Andreas Katsaros provide some evidence that could further validate the rational addiction theory by looking at public smoking bans in England, Wales and Northern Ireland. Using the British Household Panel Survey, which includes questions about happiness and smoking, they find that those who reduced their smoking show no change in happiness compared to those who did not change their smoking. But once public smoking bans were imposed, those who reduce smoking, in particular heavy smokers, exhibit decreases in happiness. In other words, they were forced to reduce smoking by being chased away from public places, and feel worse for it. That is consistent with the rational addiction theory in that imposing an unanticipated constraint makes people worse off. But one could also imagine irrational addiction theories that would be consistent with that result. Indeed, the critical aspect here is that you are locked into a state (addicted smoker). Whether you got there rationally or irrationally does not matter.