Cutting purchases of sugary drinks through vending machines

Obesity is one of the key issues the world faces. While a tax on sugar-sweetened drinks has been proposed by many, there is not much experimental evidence to show that it works.

Any sales tax is likely to be resisted by retailers fearing a drop in sales. A randomised controlled trial could show the true impact of any specific tax on overall sales and reveal whether people would gravitate towards healthier substitutes.

We worked with Alfred Health to run a randomised controlled trial in which we randomly allocated certain vending machines in a hospital to have a 20 per cent increase in the cost of unhealthy drinks.

The 20 per cent price increase meant that fewer people bought sugary drinks. However, the number of purchases overall did not go down, suggesting that there would be no negative impact on retailers.