If you read a headline claiming chocolate can help you lose weight, what would be your first reaction?
 
Based on their own experience, some people may be sceptical. Some may even be sceptical enough to check out the research themselves. But it’s likely most won’t.
 
Not even those whose job it is to inform the general public.
 
In a great example of media trolling, journalist John Bohannon purposely designed a shoddy science experiment. We’re talking classic methodological warning signs, like having a small sample size, looking at too many variables, and featuring in a pay-to-publish journal.
 
The study concluded that – as you may have guessed – chocolate may help you lose weight. You shouldn’t need a PhD in science to question whether this study is legitimate.
 
Nevertheless, the media unquestioningly published the results of the study – as Bohannon knew they would.
 
Bohannon’s point was that in today’s media environment, it is too easy for bad science to be reported and believed. With pressing deadlines and an incentive for sensationalism, sometimes reading the press release is all a journalist has time to muster.
 
This is a problem. The only thing worse would be if bad science became the basis for equally bad policy. And there is a real risk of that.
 
This is the very subject of The Health of the State, The New Zealand Initiative’s latest report on public health and lifestyle regulations.
 
The report looks at current regulations like those around e-cigarettes, as well as policies commonly advocated for, such as sugar taxes, or more restrictions on alcohol advertising.
 
It finds that a lot of the studies which form the evidence base of these regulations are methodologically flawed. Further, there are some studies which are robust, but don’t actually prove what policy advocates say they prove.
 
For example, proving a soda tax reduces a household spending on soda is not the same as proving a soda tax will reduce obesity.
 
The Health of the State showcases the things that can go wrong when interpreting health studies, making the limitations on individual choice even more morally contestable.
 
These regulations need to be challenged. A good start would be for the government to consistently apply quality cost-benefit analysis. Although that alone won’t count the costs to individual liberty.
 
Until then, I guess few people would object too loudly if the government was to subsidise chocolate.