OK, first things first: taking up and then quitting smoking won't make you earn more. I think (well, I really hope) that the Daily Telegraph article is tongue in cheek, but even so, it's a stupid and frankly dangerous thing to say. Everyone knows smoking is a deadly habit, and that nicotine, when consumed by smoking a cigarette, is incredibly addictive.

As Homer Simpson says to Bart when he announces he will take up and then quit smoking, it's "one of the hardest things you'll ever have to do". If people were to attempt to get a pay rise by starting smoking and then stopping, they may find themselves unable to achieve the second part of their plan. But, this is a good opportunity to discuss why there might be an association between being an ex-smoker and earning more money, and what that does and doesn't mean.

Also, to quickly get something else out of the way, the paper that the Telegraph article references is a Federal Reserve Bank of Atlanta working paper. As such, as far as I can tell it's not been peer reviewed, and although peer review is by no means perfect (a conversation for another time), not having been peer reviewed at all is potentially worse. But let's put that aside and look at what the working paper actually found.

The authors of the paper do not reach the same conclusion as the Telegraph article. Quite the opposite: they themselves say that it's the other differences between non-smokers, ex-smokers and current smokers that account for the difference in wages, rather than their smoking status. In studies such as these, where people are observed, those differences are called "confounders".

People who decide to smoke are different from people who decide not to smoke in lots of ways (and, indeed, from people who start smoking, but then successfully quit), other than merely smoking status. If you wanted to design an experiment to test whether smoking causes a difference in wages, you would want to take a group of people, randomly assign them to either smoke, smoke for a while and then give up, or never smoke, and then look at them again in a few years and see what their wage was. This is both impractical and unethical, so we can only observe people, meaning we have to account for the other differences in our statistical analyses.

In the working paper, the authors found a difference in education levels between smokers and non smokers. This immediately suggests a reason for the difference in wages between the groups, other than smoking causing lower wages. Higher education could translate to more skilled employment which might have a higher wage. Education levels are also associated with social class differences, which are again linked to the amount of money earned.

With regards to why ex smokers might earn more, if you've taken up and then quit smoking, you're likely to be older: current smokers are the ex smokers of the future. Also, quitting smoking is hard, and can take years to achieve successfully. Older people are further along their chosen career path and as such are probably earning more money. So in terms of the Telegraph's advice, perhaps it should be: "Want to earn more? Progress through your career, you'll be earning more towards the end of it than you were at the beginning." Helpful.

If smoking really caused a decrease in wages, you would expect there to be a dose response relationship, whereby the more a person smoked, the larger the difference in wage would be, compared to a non smoker. The authors of the working paper looked at this, and found no difference in wage gap when they looked at different intensity smokers. This is more evidence against smoking causing the difference in wages.

While the positive determinism suggested by the Telegraph headline might be nice to believe, unfortunately it's not the case that we can control our own destiny in such a way. Although, since taking up smoking is a dangerous thing to do, this is probably a relief.

There are plausible suggestions as to why smoking might cause you to be undesirable to an employer (and thus potentially affect your salary): cigarette breaks, time off for coughs and illnesses and the like. But the paper that informs the Telegraph article actually finds evidence suggesting this is unlikely to be the main cause of the wage difference seen. As Homer would say, "Doh!"