Michael Gallagher (6th September, 2010)
Concerns are raised from time to time, not just in this country, that geographical representation in parliament creates behaviour among legislators that is not consistent with the national interest. Legislators elected, under whatever electoral system, from geographical constituencies have an obvious incentive to put the interests of their constituents first even when a “rational” decision-making process might require acknowledging that not every local interest should be defended to the death. They might indeed argue that this is not merely a matter of responding to electoral incentives but that protecting and advancing the interests of their constituents is part of their job.
For example, if a country has too many hospitals, in the sense that resources are spread too thinly among them and that the overall health of the nation would be better protected by having fewer but bigger hospitals – as some say is the case in this country – a national policy-maker might conclude that smaller hospitals should be closed and their resources transferred to larger regional hospitals, but the MPs for the constituencies where hospital closures are sought are likely to resist such closures as strongly as they can.
In chapter 9 of his book “The Concept of Constituency: political representation, democratic legitimacy, and institutional design” (Cambridge University Press, 2005), Andrew Rehfeld argues in favour of randomly-constructed rather than geographically-constructed constituencies. Upon reaching voting age, everyone would be randomly allocated to a constituency, to which they would belong for the rest of their life. In the Irish case, for example, leaving aside any other possible changes in Dáil size, electoral system or constituency side, there might be, say, 33 constituencies, each with 5 members elected under PR-STV as at present, each containing approximately 100,000 people. (Though in fact Rehfeld is against any form of PR and believes that each constituency should return only 1 MP.) Each electorate of 100,000 people would be, in effect, a random sample of the population of the entire country, and every small region of the country would be likely to contain members of each of the 33 constituencies. Among other things, this implies that voting behaviour in each constituency would be likely to be very similar. Non-geographical constituencies may seem an outlandish concept but they are commonly used for upper houses, such as Seanad Éireann for example.
Rehfeld suggests that this might not be fully feasible until internet technology and access have advanced to the point where geographically dispersed groups of people, which is what constituencies would now be, can meaningfully discuss matters of common interest – though he expects this to happen sooner than we think.
The argument in favour is that each MP would now have an incentive to pursue policies that are good for the nation as a whole. MPs would, in effect, be operating as if behind the famous Rawlsian “veil of ignorance”, so when it came to upgrading schools, for example, they would have no incentive to try to get particular schools moved up the qualification list, because this would probably cost them more votes than it gained them. Instead, their incentive would be to put in place a decision-making method that sees resources allocated to the school that at any moment is the most deserving in terms of a set of objective criteria. Likewise, no MP would be lobbying to have public or private investment directed at a particular part of the country; instead, it would go wherever makes best sense for the nation as a whole.
There are also difficulties, some of which Rehfeld addresses, though he acknowledges that his scheme is not fully fleshed out. One obvious question is: if education / health / transport services in an area seem to be failing, whom do people contact? Maybe no MP would be contacted on the matter, as none would have much responsibility for the affected area, or maybe every MP in parliament would be contacted as all would have a few constituents in the area. Would this scheme mean that all decisions are made in “the national interest”, or is that concept too contestable to be useful? Might the outcome be not that policy outputs are in everyone’s (ie the broad national) interest but that geographical peripheries and perhaps other minorities, now left without anyone with a responsibility to defend their interests specifically, would lose out systematically to majorities and to the centre?
Is this an idea that could be made to work? If not, does it mean that the benefits of geographical constituencies, with MPs who see their representational role as consisting at least partly of defending and promoting the interests of their own small patch of the country, outweigh the disadvantages?