OK the results are preliminary but a big study (pdf) (8000 people) into Mexico's 'soda tax' has shown that it hasn't affected overall calorie intake one jot:
Obesity and its costs are high and rising and we know little about the effectiveness of different policy tools. We measure the short term impact of one such tool: taxing high caloric density foods. The results are still preliminary, but the evidence shows that the effects of the Mexican taxes on calories consumed in-home are very small. Results also show that lower SES may pay a higher percentage of their income from these taxes.
And that last sentence tells us that the poor ("lower SES") are the ones paying the tax. So Jamie Oliver, nannying MPs and the legions of public health fussbuckets are proposing a tax that won't solve the problem (an overstated problem but that's by the by) but will disproportionately fall on the least well off.