How much impact can taxes and bans have on the conduct of people?

Quite a lot. A 500% tax would clearly reduce demand while outright bans of desired products keep law-abiding people from using a product - and make the others rich.

In defiance of well-established history of resentment about government steering drink choices for freedom-loving westerners - the UK lost a whole colony when they decided to force their subjects to buy only the brand of tea elites wanted them to buy(1) - a subset of progressive social authoritarians maintain that higher taxes on soda drinks, juices, etc. would fix obese people, despite data showing it has never once worked.

A tax on soda in Ireland and France did reduce soda consumption (yayyyy) but did nothing at all for obesity (oh), which was the whole reason they put the higher taxes on soda. Basically, it just made soda more available for the 1% and caused poor people drink more tap water, which wouldn't ordinarily be the purview of government.

Regardless of fact and reason related to higher taxes on certain foods, a paper in BMJ (BMJ 2013;347:f6189) claims government should do raise taxes, and insists those higher taxes will cure the fat - but even in their speculative estimate a 20% tax will only help 180,000 obese people - out of a population of 63 million.  BMJ is prone to hyperbole(2) - in 2008 they declared that 33% of the UK would be obese by 2012 though the UK obesity rate is 23.1% in 2013 and so newer estimates now say 2020 will be the magic 33% mark (the future is always bleak when it comes to epidemiologists making health estimates). Even in the fuzzy world of epidemiology, raising prices for 63 million people in the hope that it might help .003 of them doesn't pass a common sense test, yet that was still enough to lead Jason Block, epidemiologist at Harvard, to call on another country "to implement high taxes and measure the results" in their press release - it's therefore no surprise that he is one of their references.

Using prospective cohort studies of household survey data to make conclusions about obesity and taxes on soda is suspect but they rationalize it with 'other studies have done this' - which is epidemiology invoking its own argumentum ad verecundiam. Creating estimates and then doing a Bayes analysis of those estimates and stating they have confidence in the result is baffling. They declare a 95% credible interval of a drop in obesity in adults of between 0.8% to 1.7%. 12,000 iterations of a Markov Chain Monte Carlo analysis would be fine if the data was real - writing here, Bruce Bukiet predicted a Red Sox victory in the World Series that way and nailed it. But he had actual performance statistics of baseball players to feed in. Using surveys and then matching them to obesity statistics and prices for soda is not the same thing - that Markov chain can random walk us to just about any result.

Uh oh, this is how they made their model for the causal pathway between sugar sweetened drink taxation and obesity. Credit and link: BMJ 2013;347:f6189

At the end of it all, they endorse a proposal by the Academy of Medical Royal Colleges that even the UK government - which loves taxes - thought made no sense, saying 'hey, let's just try a one year 20% tax.' Have you ever known a government to get rid of a tax? I haven't either. There is a reason UK billionaire Richard Branson moved to an island and it wasn't "privacy" or lifestyle, it was UK taxes. But even the claim that government would get an extra $400 million plus save poor people from themselves has not worked yet.

However, because this was in BMJ, look for a New York City politician to use this as evidence they should be able to ban Big Gulps. After all, they banned trans fats to save fat people from themselves and we saw how that worked...diabetes went up. A lot.

Citation: Adam D M Briggs, Oliver T Mytton, Ariane Kehlbacher, lecturer, Richard Tiffin, Mike Rayner, Peter Scarborough, 'Overall and income specific effect on prevalence of overweight and obesity of 20% sugar sweetened drink tax in UK: econometric and comparative risk assessment modelling study', BMJ 2013;347:f6189 31 October 2013

NOTES:

(1) And we got a constitutional prohibition on legal sales of alcohol in the US, which made one bootlegger so much money he was able to get his son into the presidency.

(2)  I can't figure BMJ out. It used to be they only wrote gag articles in December (examples: Head banging is bad for your head or Why Rudolph's nose is red) but they seem to push them out all during the year now. A paper looking for truth behind the saying "every time Wales win the rugby grand slam, a Pope dies" was in April of 2012 (BMJ 2012;344:e2742), but not April 1st so they may have been serious. But at least they are fair - articles often contradict each other (which happens here also) so they may have an article next month claiming free soda would reduce obesity, like opposing claims that public satisfaction with the NHS is at a record high but GPs are being forced to ration access to hospital care or scrap the 'unwinnable' drugs war and divert funds into curbing global antibiotic misuse but fears about complications shouldn't lower antibiotic use. A drug war is "unwinnable" but soda taxes fix obesity? I am starting to think that epidemiology articles this decade should come with a warning label.