Saturday, 20 September 2014

Scotland decides: Are referendums a good idea?

The Scottish independence referendum has finally taken place with a comfortable majority voting no to independence. The result is, however, somewhat bizarrely being interpreted as a 'clear call' for greater devolution. Why bizarre? We have gone from four million Scots being asked a yes, no question 'should Scotland be an independent country' to the conclusion that a majority of the sixty million people in the UK want greater devolution to the regions, and they want it before March 2015! Surely there is not a better way to judge the 'will of the people'?
       The answer, unfortunately, is 'probably not'. And the main reason why is the difficulty of making inter-personal utility comparisons. What we have here is a problem of collective choice. A single decision has to be made, e.g. Scotland becomes independent or it does not, but that decision will affect many. Some people stand to benefit a lot from independence and some stand to lose a lot from independence. In a utopian world those that stand to benefit could compensate those who will lose out - transferable utility in the parlance of cooperative game theory. In practice, however, transfers are clearly a non-starter for something as complex as Scottish independence. 
       Without transfers there is simply never going to be unanimous agreement. This lack of unanimity inevitable means the policy maker has to trade-off the desires of one person with the desires of another. And that means making inter-personal utility comparisons. But, how can the happiness of one person be compared with the disappointment of another? Economists and philosophers have long sort a solution to that problem. But what we have basically learnt is that there is no solution. There are various approaches to tackling the problem such as utilitarianism - maximize total happiness - or Rawlsianism - maximize minimum happiness. These approaches, however, only serve to kick the problem down to another level. We are never going to get unanimity on whether to use utilitarianism or Rawlsianism.
        The difficulty of inter-personal comparison means that there is no such thing as an optimal policy. There is no best or most efficient policy. I think this logical fact is a bit disconcerting to most people, including policy-makers. It seems a bit worrying that there is not an optimal way of doing things. A certain freedom, however, comes from knowing optimal is not an option. The Scottish referendum, for instance, could seemingly have gone either way. But that's fine because either was way would have been ok. While it might have been optimal for Jack to have independence and optimal for Jock to not have independence there was no optimal overall outcome.
        Policy making on a whim need not, therefore, be a bad thing. The West Lothian question, for example, has been knocking around in British politics for way too long. Progress is typically stymied by the claim there is no optimal solution. Well there is no optimal solution, but that is no reason to do nothing! To say, however, that there is no optimal policy does not mean that there are not good and bad policies. And there are plenty of examples of bad policies around the world.