Constitutional Change

This post is about the proposed change to the British electoral system, from "First Past The Post" (FPTP) to "Alternative Vote". British electors will have the opportunity to make a choice between these two (and no other) options in next week's referendum. Most of the public arguments both for and against AV have been pretty fatous and feeble, with plenty of ad hominem.

John Kay is a welcome exception. In his latest article A voting system fit to bar Le Pen from power (FT April 27 2011), Kay offers a lucid argument in favor of AV, and makes the following observation.

Britain has an informal system of alternative voting already, whose operation depends on voters making good guesses as to the likely result. This strengthens the case for the formal adoption of AV, but also explains why it would not make very much difference in practice. ... Even if the alternative vote is not the official system, voters will tend to behave as if it were.

John Kay's analysis therefore distinguishes the formal system of voting in the UK (currently first past the post) from the defacto system in use. He suggests that UK voting behaviour already partially reflects an informal conceptual model based on AV, and the proposed change would merely help the formal system to capture the emergent voting behaviour more accurately.

This is one of the reasons why it is difficult to predict the likely consequences of the change, because nobody knows exactly how British electors will adapt their (emergent) voting behaviour to the new formal voting system. Kay is probably correct in predicting that the first order effects of the change will be much less significant than either the pro-AV or anti-AV campaigners have claimed.

One of the arguments put by the anti-AV campaign is that AV is more complicated than FPTP. But FPTP already provokes some people to adopt complicated voting behaviours, and it is not evident that the behaviours associated with AV would be any more complicated than the behaviours associated with FPTP. The point here is that we should look at the total sociopolitical complexity of a given voting scheme, not merely the counting procedure.

There is also a second-order question - whether any given outcome from this referendum makes further constitutional change (e.g. full proportional representation) more or less likely. I have seen some divided opinion about this, but it seems pretty speculative. Some of the campaigners acknowledge that the first-order effects will be pretty small, and they see the current campaign as merely a preliminary battle in a longer-term and more fundamental reform.