During last night’s debate, Robin repeated an argument many economists have made: In the long-run, maximizing efficiency is actually better for everyone.  If we consistently adopt any policy with benefits greater than costs, then the times that you win will outweigh those that you lose, so you’ll be better off on net.  Optional escape hatch: Even if you don’t happen to win on net, you can still ex ante expect to win on net; efficiency raises everyone’s expected well-being.

When I give my lecture on efficiency, I always point out how crazy this claim is.  “Everyone” means everyone.  The world has over six billion people.  Some of them have really weird preferences; a subset of these people with weird preferences are already extremely satisfied.  Is it really likely – or even credible – that every single one of these people would be better off if we junked all inefficient policies? 

Take for example Kim Jong-Il, dictator of North Korea.  Right now he owns 20 million slaves.  He’s the master of life and death for a nation.  How exactly would Kim be better off if, for starters, communism were abolished?  Even if you offered him billions in compensation, he’s got a lifestyle that money just can’t buy.

Or to take a more mundane example: How about an elderly, childless misanthrope?  It seems quite likely that he won’t live long enough for Robin’s “long-run” defense to kick in. 

Of course, you could always modify the efficiency standard to say, e.g., that “The welfare of 99% of people must go up.”  I don’t think Robin’s willing to do so.  But suppose he was.  In the real world, what percentage of the population would lose if we consistently followed the efficiency maxim?

Personally, I think it’s got to be at least 10%.  Human preferences are incredibly diverse – as you can see if you read nationally representative survey results instead of just talking to the people like yourself.