over 10 years on 2008-06-01


i currently live with my college roommate, bcm, who recently returned from a year in paris to start work at a notable private equity firm.  ever since college bcm and i have greatly enjoyed applying our economics classes directly into our private living situation, so we regularly construct simple but effective economic incentive systems to generate mutually beneficial outcomes.

one recent example of how this successfully works has to do with exercise.  bcm and i both decided that we were out of shape and wanted to get back to working out more regularly.  so, we constructed a simple and effective system to encourage each other to work out.  we agreed that we would both like to work out at least 4 times a week and that for every missed workout we would pay a $20 fine.  so, the maximum weekly exposure we each have to eachother is $80, if we each work out the same number of times (up to 4) then no money is transacted, and it only gets expensive if one of us fails to keep up with the other.  this generates some interesting negotiations now and again, but generally works quite well and we have been working out much more regularly.  regulation actually increased the general welfare.

while in paris bcm picked up smoking, an activity which generates incredible negative externalities in the environs of our common apartment, and which has serious long-term costs for bcm, and his decedents.  we have been trying to figure out some sort of mutually agreeable system that will encourage him not to smoke.  what we have realized is that our difficulty figuring out a system for this problem is completely synonymous with the world's issues creating an economic incentive system to save the environment. 

bcm is very much like a small polluting nation, and our dinnertime discussions about smoking are 100% synonymous with failed international negotiations.

we have talked through several models for convincing bcm to quit smoking, but they are ultimately based on the two same options proposed for the environment - we need to use either a pigovian tax or a cap and trade model

the first, and most obvious, was a straight up tax/transfer from brian to me -- $1 per cigarette.  this would work if i could get the proposal passed, but the problem is that brian has zero incentive to agree to this.  in the slightly more complex model of the same thing, we would model the cost of offsetting my own unhappiness, the cost of extra cleaning services, and the long term cost of treating brian for his cigarette use, etc. and charge that as the correct pigovian tax.

the second option, a cap and trade system, also fails.  it fails because there are only two of us in the equation, meaning that there is little difference between this and a straight pigovian tax.  if we auctioned off the credits to each other at the start it would be no different than a straight up tax on bcm.  if we did a modified cap and trade and issued credits initially the problem is who gets the credits to start.  theoretically we could deal in a few other players to expand the market to the full list of effected parties (like bcm's future unborn children, the cleaning service people who benefit from bcm's smoking, etc.) but you end up with more complications and value judgements on who gets what credits and what their incentives are (clean service people, who benefit when bcm smokes, will sell bcm credits at a lower price than i would, driving down the total value for me).

upshot, i don't think that we can design a system that bcm will agree to to get him to stop smoking until he wants to stop smoking - because on the micro level anything we do is a straight tax on him.  this does not bode well for our long term prospects of environmental policy reform.

original swl blogposts and letters 2007-2010