Yesterday, the Government published a "call to action" which lays out its strategy for tackling obesity in England. The document sets out plans for government to work with the food and drinks industry to achieve a reduction in obesity. One measure for tackling diet related health problems is to introduce some form of 'fat tax'. Although not mentioned in the report, following the introduction of such a policy in Denmark last Monday, the Prime Minister David Cameron stated that introducing a similar tax in the UK "is something we should look at". The Danish tax places a surcharge on foods that contain more that 2.3% saturated fat, including products such as butter, milk, cheese, pizza, meat and processed food. Mr Cameron suggests that introducing a similar measure could be one way of tackling rising health costs and falling life expectancy among some groups of people. Many other countries are considering policies aimed at improving diet through increasing the prices of certain products (for example, Hungary recently imposed a tax on foods with high fat, salt and sugar content, various US States have adopted or contemplated soft drink taxes and the Scottish Government intends to introduce a minimum price for alcohol).

The rationale behind levying a tax on foods that are rich in saturated fat is that, by increasing their price, the tax will encourage people to eat less of these foods. The hope is that this will help curb consumption of saturated fat which, according to the Food Standards Agency, is on average 20% above the recommended level. Excessive saturated fat consumption is a risk factor for heart disease, so lowering consumption should lead to a reduction in the incidence of heart disease, all else equal.

IFS has recently received long-term funding from the European Research Council and the Economic and Social Research Council to investigate the impact of these types of policies.

It is important to be clear about what, precisely, the objective of such a policy is - what role is there for government in influencing what individuals decide to eat? Are people not in a better position than government to trade off their likes and dislikes and to choose to eat the products they favour most? The two main reasons why government intervention might be justifiable are:

1. Some people may not be fully informed about the amount of saturated fat they can eat as part of a healthy diet and which products contain large quantities of fat. Providing information may seem like a natural response, but it may be that some people are difficult to reach or are not capable of processing this information accurately. Young children are the most obvious example. Raising the price of foods rich in saturated fat may help reduce consumption levels in a similar way to conveying information.
2. While most of the costs of ill health arising from a poor diet are borne by the individual themselves, some costs are incurred by other members of society, for example, through increased health care costs and increased sick leave. When individuals decide what to eat they have no incentive to take these external costs into account; this results in a socially excessive level of consumption. By increasing the price of consuming fatty foods, a tax is one way of tackling this over-consumption.

Whether a 'fat tax' is the best policy response to high levels of diet related illness will depend on what of these failures in the market the government is targeting. Griffith and O'Connell (2010) discuss these failures and how well suited various policy options are to addressing them (see Griffith and O'Connell Public Policy towards Food Consumption, Fiscal Studies, vol. 31 no. 4 pp. 481-507) .

Before deciding that introducing a 'fat tax' is a good way to tackle poor diet, the government must also be clear about what element of diet it seeks to influence: is the aim to reduce consumption of all sources of saturated fat, only certain sources (for example, confectionary but not milk), or a set of nutrients often excessively consumed (like salt and sugar, in addition to saturated fat). This will influence the choice over the appropriate set of products to tax.

A second question is what impact any particular policy reform would have. This is not a simple question to answer. It depends on the specific structure of the tax, the market environment, and most importantly on how sensitive consumers are to changes in price. Both the structure of the tax (for instance, whether it is specific or ad valorem) and the nature of competition in the market will determine how taxes are fed through to prices. Griffith, Nesheim and O'Connell (2010) explore these issues by estimating the impact of a fat tax levied on butter and margarine and show that these considerations have a very substantial effect on how a tax would impact consumers, firms and government tax revenue (see Griffith, Nesheim and O'Connell Sin taxes in differentiated product oligopoly: an application to the butter and margarine market, CEMMAP Working Paper CWP37/10). Although the relationship between tax and price changes is a complicated and partially understood one, there are some things the government can do to maximise the potential rise in prices, given the magnitude of the tax; for instance, Griffith, Nesheim and O'Connell (2010) provide empirical evidence that a specific tax is likely to result in higher prices than a comparable ad valorem tax designed to raise the same amount of revenue.

The response of consumers to any price changes will be crucial. The more sensitive a consumer is to price the more effective the tax will be at reducing consumption of the taxed goods. But whether the tax is successful in improving diet will also depend on what individuals would choose to consume instead. A tax levied on butter, but not margarine, may be effective in curbing saturated fat consumption. But since margarines tend to contain more salt than butter, the policy may have the perverse consequence of exacerbating other health problems.

A final important point concerns whether a 'fat tax' would be regressive or progressive. As lower income households tend to spend a higher fraction of their total budgets on food, a disproportionate proportion of any such tax is likely to fall on their shoulders. But the point of such a policy is to encourage people to change their behaviour. Often lower income households are the most price sensitive, which would mean they are likely to change their behaviour most as a consequence of a price rise. To assess the overall effect of the tax and whether low or high income consumers would be affected more, we would need to set the costs imposed by the tax on consumers through higher prices against the potential health benefits arising through their changed behaviour.

Governments around the world are considering how policy can address the increase in diet related disease. One policy response that is increasingly being proposed and adopted is to introduce a 'fat tax'. For this to be an effective response to high and growing levels of diet related health disease, a number of circumstances must be met. Firstly, the government must be clear about why intervention is justified and how exactly it wants to improve diet. Secondly, to be effective, any tax must feed through into higher prices. Although the relationship between prices and tax is complicated, there is some evidence that by structuring the tax carefully, the government would be able to influence how much prices increase by. Thirdly, the price increases must bring about the desired changes in consumer behaviour. There is substantial debate about the role for government in fighting diet related health disease, and how government can best help. There is less evidence about what the impact of introducing policies like fat taxes would be: how exactly would the tax feed through to prices and how would consumers respond? Researchers at IFS are working on improving the evidence base to help policymakers understand the likely impact of various specific reforms.