Fat city should not become tax city.

• Author: Snowden, Christopher
• Position: National Affairs

EXTENDING "sin taxes'" to certain foods and soft drinks increasingly is seen as a practical response to obesity and obesity-related diseases. Several countries and most U.S. states have some form of tax on soda, candy, and/or fatty foods, but they so far have had little or no effect on population health. For a tax to be justified, it should be efficient, fair, and effective. Taxes on food and drink meet none of these criteria.

On the question of whether obesity taxes are effective, it generally is accepted that higher prices lead to lower levels of consumption. If consumption of a given product directly leads to obesity, then it is reasonable to expect that higher prices could lead to lower rates of obesity. The obvious comparison, which often is made by those who campaign for food and drink taxes, is between cigarette taxes and lung cancer, but the comparison does not hold.

Cigarette smoking is the main cause of lung cancer, and most cases of lung cancer are caused by cigarette smoking. By contrast, obesity is the result of consuming more calories than an individual burns off. Drinking soda is neither a sufficient nor a necessary cause of obesity, and consumers easily can switch to other high-calorie products if the price of one product rises. A cigarette smoker faced with unaffordable prices might switch to alternative nicotine products, such as e-cigarettes, smokeless tobacco, or patches--any of which would improve that person's health--but a soda drinker can switch to beer, milkshakes, or fruit juice without reducing his or her calorie intake or lowering his or her weight.

Evidence of the efficacy of food and drink taxes is divided sharply between theoretical models (which tend to show a modest, but significant, effect on body weight), and studies of taxes that actually have been implemented (which do not). The economics literature suggests that soft drinks have a price elasticity of 79 cents, albeit with a very broad confidence interval, meaning that a price rise of 10% should reduce consumption by 7.9%. A conventional economic model therefore can estimate the drop in consumption that would result from a price rise, and a more speculative public health model can estimate the impact on body weight and health outcomes. However, even the more optimistic models, based on relatively high price hikes, claim quite modest changes in overall calorie consumption. One study found that a penny-per-ounce soda tax would reduce average energy consumption by nine calories a day, for example. Another model predicted that a 10% tax on milk and fizzy drinks would have even less effect

Advocates of such taxes argue that small changes in calorie consumption can have a meaningful effect on health over time. Real world evidence has found only a modest effect...