Sunday, October 15, 2006

Species Survival

I was send this question my a friend studying Social Policy, who in turn had been asked it by someone at LSE.

I'm trying out an idea, and I wondered if you could possibly answer the following hypothetical question for me.

Imagine that you are a policy maker in charge of environmental protection. You are considering two policies, policy A and policy B, that can be summarised as follows:
Policy A has a 84% chance of saving 1 species of animals and a 16% chance of saving 40 species of animals.
Policy B has a 55% chance of saving 5 species of animals and a 45% chance of saving 10 species of animals.

You are not sure which species policies A and B will save - therefore you can assume that you would value all species that can be saved equally. Policies A and B are assumed to have no negative 'side effects' (i.e. the policies can only 'do good').

You have a budget of $1million to promote these policies to your legislature. Policies have a better chance of being enacted if their 'promotion' budget is large. How much of the $1 million would you allocate towards promoting policy A, and how much would you allocate towards promoting policy B, and why?

There's plenty more I could say - an awful lot of it quibbling with things under-defined by the question (when it says policies have a better chance of being enacted with more money, does that mean allocating $250,000 might in fact go to a policy that isn't even enacted and achieve nothing? What are these chances? Are the probabilities we're given the chances of success given enactment?)

My response was as follows:

As an actuarial matter, there's hardly any difference. The expected outcomesare 7.24 and 7.25 respectively. Not that numbers mean much.

I don't think you can make an informed choice from such abstract data. Much would depend on which species were involved - what the knock on effects would be for others in the ecosystem and such.

It's also possible framing effects may be involved. People respond differently to losses rather than benefits, tending to favour risks (which I presume meanspolicy A)

I'm not sure what the background assumption is. Since policy A can save up to 40 species, presumably at least that many are endangered. Let's say in fact 50 species are going to die out if we do nothing. Now we can rephrase these policies as:
A - 84% chance 49 species extinct, 16% chance only 10 extinct
B - 55% chance 45 species extinct, 45% chance only 40 extinct

I wonder how many people's intuitions would change put that way.

Personally, in light of the actuarial similarity and lack of any other relevant information I'd be pretty indifferent. On a policy level, I'd be happy to adopt whichever was democratically favoured. (I think it's fruitless to search for 'best policies' - better establish a list of acceptable options, and then leave matters to public deliberation/vote)

I'm interested to know what others think - both answers to the question and what it's actually getting at.



Post a Comment

<< Home