What’s the price on your head?

How much would you spend to save a life? How do you make that decision? It’s much more complicated than you think… Read on to discover the difficult implications of putting a monetary value on saving lives.

When Nevado del Ruiz erupted in Colombia in 1985, about 23,000 people died as a result. If you were asked to estimate how many pounds each of those lost lives was worth, what would your response be?


Your first reaction might be that it is impossible to put a value on a life, that life is infinitely valuable or fundamentally priceless. But that’s not very helpful advice to policy makers who have to decide how much to spend on building improvements or relocation to reduce the risk of loss of life. For example, suppose that Mount St John has a 1% chance of a major eruption in the next 20 years. There is a population of 100,000 living nearby, and if there was an eruption, then each person would face a 2% chance of losing their life. Thinking only about the risk to life, how much would it be worth spending to reduce this 2% chance to 1%? £10 million? £100 million? More? Without some monetary value placed on those lives it is impossible to say.  Policy makers need to make many important decisions about investing in road safety, regulating nuclear power stations, or deciding what cancer treatments the NHS should pay for, all of which depend critically on these types of valuations.

In fact, we all make decisions about life and death every day.  We travel in cars, and sometimes we drive them faster than necessary; we cross busy roads and we don’t always walk to the pedestrian crossing. If we really considered the loss of a life to be infinitely costly, then we would never leave the safety of our homes, as it would never be worth the increased risk. Similarly, we are willing to pay for smoke-alarms and improved car safety features, but only up to a point.

It is important to understand that we are not asking how much we would pay to save one known person’s life for sure, but how much we would pay to reduce the average level of risk in the population. The UK refers to this as the ‘Value of a Prevented Fatality‘ (VPF) and the US as the ‘Value of a Statistical Life‘ (VSL). As well as using different terms, the two countries also use very different monetary values. In 2016, the UK used a value of £1.8m to assess risks across a range of bodies, whereas the comparable US figure was about £7m in the Department of Transport, and even higher in other government agencies.

What explains this difference? It seems highly unlikely that people or circumstances are sufficiently different across the US and UK to justify a value of life that is four times as high, so the problem is likely to lie in difficulties determining the correct value. How are these values determined?


Early attempts tended to take a ‘human capital‘ approach. This values a life by the average earnings lost over a lifetime. So if the average person would earn £25,000 per year for 40 years, then this implies that the required value is £1 million. The obvious problem with this is that implies that the only part of someone’s life that has a value is their paid work and neglects everything else. Clearly there is more to life than just work!

Instead, the values used by policy makers today are based on getting people to reveal how much they would pay to reduce the chance of a fatal accident. In the US, the VSL is often estimated by looking at the difference in wages between high risk and low risk jobs. For example, suppose that miners face an additional 1 in 5,000 chance of death each year compared to similar jobs, but receive an extra £2,000 a year in pay. This would give a value of one statistical life of 5,000 multiplied by £2,000, giving £10 million.

This ‘revealed preference‘ approach has the advantage compared to the human capital approach that it estimates how someone values their life as a whole, not just wages. However, if people do not really know what risks they are taking on when they make their choices, then it may not give a good estimate.


An alternative is to conduct surveys about hypothetical decisions. The value of the UK’s widely used VPF of £1.8 million is based on a survey of 167 people commissioned by the government in 1997. Respondents are asked questions such as how much they would be willing to pay to reduce their annual risk of a fatal road accident from (say) 5 in 100,000, to 4 in 100,000 [the actual set of questions asked was a bit more complicated than this].  The 1997 study has been criticized on a number of grounds, but one obvious concern is that the average wealth of the people surveyed was much lower than average. What difference do you think that made to the results? More fundamentally, do people really have a sufficiently good understanding of probabilities to value a decrease in risk of this sort?

The question becomes even tougher when we need to take an international view. The graph below shows the VSL for different countries, calculated from wage-risk studies. The highest values are about ten times the lowest values, but do you think it would be appropriate for an international organization to use this information when making decisions about where to allocate spending? And what about the value of future lives? If the value of a (statistical) life today is £5 million, how much should we be willing to spend on climate change measures that will save a life in 500 years’ time? These questions are just as much philosophical and political problems as economic ones and it is important to look at the whole picture.

KD Graph

The question of how to put a monetary value on saving a life is an incredibly challenging one and it is tempting to conclude that there is no right answer. Yet every time policy makers make a decision on how much to spend to reduce the risk of natural disaster, road accidents, or climate change they are making an assumption about the answer, at least implicitly. With better answers they can make better decisions, even if they are not perfect ones.

Your task

  1. Returning to Mount St John, how much would the value of a (statistical) life have to be in this case to justify spending £10 million pounds or £100 million pounds?
  2. How much would you be willing to pay to reduce the risk of a fatal road accident from 5 in 100,000 to 4 in 100,000? What does this imply for your estimated VSL? Try asking other people and compare answers.
  3. How easy do people find it to answer questions like Q2? Are there different ways of asking the question that make it easier? For example, does it help if you use population figures to work out the absolute number of road deaths that the probabilities imply?
  4. How do the answers you get from Q2 compare to the answers you would get if you took a human capital approach?

Click here to check your work.

Kate Doornik

Dr Kate Doornik is responsible for organising economics teaching at St John’s, and leads courses in microeconomics and macroeconomics.