The sunk cost fallacy


Kate Hudson, Chair of CND, has a letter in The Guardian.

...there is no lack of resources for investment in the UK's energy sector, only a lack of forward thinking. The government is set to heavily subsidise a new generation of nuclear power stations, despite recent reports that decommissioning the existing generation could cost £83bn of taxpayers' money. Investment in renewable and clean energy sources would be a far more productive use of public funds in the long term.

There are several errors to unravel there: of course there is a lack of resources for investment. We cannot invest in everything, we face limits upon all our resources all the time. The question we need the answer to is what is the best method of using our available resources?

It could be that renewables and "clean energy sources" (of which number I would count nuclear as a possibility but still...) are the best use of public funds. It might be that nuclear is, it could even be a network of minions using stationary bicycles to power dynamos. What we need therefore is a system, a method, for working out which of the various available alternatives is indeed that best use of our limited resources.

That method is called a cost benefit analysis: how much will it cost to do something and what is the benefit of our doing so? But the important point is that we start from now: what has happened in the past might be used to guide our estimates of either costs or benefits, certainly, but money that we already have to spend, come what may, should not factor in our calculations.

For money that we already have to spend is a sunk cost. Whether we use windmills, new reactors or those bicycles, we still have to clean up the previous generation of nuclear sites. Whatever we do going forward, those costs are there: thus they should not be used in calculating which path we should follow.

It still might be true that the bicycles are the best option (I certainly don't think so) or the renewables (not with the current generation of technology they're not) but the waving around of that £83 billion figure is simply a distraction, a piece of political shroud waving.

Oh, and one other point. Given that the plan to cover the seas and the hills with windmills is costed at £100 billion, that sunk cost actually looks rather cheap.