£21 Billion, Costs, Benefits and Externalities

Externalities are a fundamental concept in welfare economics, because they are costs or benefits which are excluded from the agent’s preferences when they are making a decision. What may be beneficial for the individual agent may not be beneficial for society as a whole. For example, someone may choose to listen to loud music late at night because it makes them happy, but they are (perhaps) not considering their neighbours’ feelings when they decide to do this. The annoyance this causes the neighbour is an externality.

Economics can help us understand externalities and how to deal with them, and there are many possible ways. The focus of this blog is on taxation, because this is often used in my chosen field of alcohol economics.

The basic premise is set out below, at a national aggregate level. Any external benefit is ignored for simplicity – but it could instead be the case that the externality is the net externality (negative externality + positive externality). 

The private cost is the aggregate cost faced by all the individual drinkers making decisions on how much alcohol to consume. This not only includes the price of the drinks, but also their associated private costs – such as the hangover in the morning and the effect on their health. The social cost includes other things, including the (external) cost of ill health and the cost of policing. The total externality is the shaded grey area. £21 billion has been cited as the total externality of alcohol at its current level of consumption, although this has been challenged, most recently by the Institute of Economic Affairs.


To improve social welfare, the aggregate social cost should be equal to the aggregate social benefit. An economist would ideally like to set what is known as a “Pigouvian” tax, where the marginal tax is exactly equal to the marginal external cost, but this is not feasible because the marginal external cost is non-linear (the gap between aggregate social cost and aggregate private cost is not the same as units increases). 

Suppose instead that a flat tax was imposed to increase the aggregate private cost so that it intersected the social benefit curve at the same point as the aggregate social cost curve (essentially pricing-in the externality). This leads to a total decrease in consumption from Q1 to Q2. Q2 is seen as the optimal level of alcohol consumption in the population.

But there are obviously a few flaws with dealing with the externality in this way. Firstly, to do this one must be able to exactly calculate the three lines – social benefit, private cost and social cost (which means calculating the external cost). What to include further complicates things, and distinguishing between a private and an external cost can be relatively tricky. A further problem with dealing with alcohol in the aggregate – as done by the £21 billion figure – is that it doesn’t tell us much about the distribution and the individual effects that an “optimal” tax would have.

Consider two different people – one a light drinker and one a heavy drinker. For simplicity, assume that their cost functions are exactly the same and all that determines whether they are a heavy or light drinker is the benefits they receive from drinking.

In both diagrams, ‘a’ represents their pre-tax level of consumption, ‘b’ their “optimal” level of consumption, and ‘c’ their post-tax level of consumption. It should be clear that after this tax, the light drinker is actually drinking below the “optimal” level whilst the heavy drinker is drinking above the “optimal” level.


So even being able to accurately define all three aggregate curves in figure 2 does not answer the question of what is “fair”. Whether the externality is £21 billion, £2.1 billion or even negative does not help.