There are a number of errors in Oborne’s article about the Parliamentary Boundary Commission. For a start, population and census data are not (unlike in the US) the basis for constituencies – the basis is registered electorate, which is different. Population would include people not on the electoral register for reason of foreign citizenship, age or under-registration (a serious problem particularly in the large cities, that was aggravated by the poll tax when the current boundaries were drawn up).
Neither is the requirement for the boundary commission to use out of date figures as big an issue as Oborne claims. If the boundary commission were to use the latest available figures, rather than the 2000 electoral register, three more counties would gain seats and three more would lose seats – not dramatic stuff, and worth only 6 to Labour’s majority rather than up to 20 as Oborne suggests.
The examples quoted by Oborne and most other critics of the boundary system – the small electorate in the Western Isles and the large electorate in the Isle of Wight – are not a consequence of the slow pace of the review. They are rare, long standing anomalies permitted by the ability of the boundary commission to take ‘special geographical circumstances’ into account. A few islands having rather too few or too many electors is not a threat to democracy. There is something to be said for making constituencies more rigorously equal-sized than they are at the moment, but advocates of this course have to accept that it would undermine one of those pillars of first-past-the-post (FPTP), namely that strong and stable link between MP and constituency. Frequent boundary changes would destabilise the link, as would the fact that many constituencies would cease to bear any relation to natural communities. Would people really prefer a seat such as ‘Southampton Central and Cowes’ to an oversized constituency covering the whole Isle of Wight?
Oborne says that ‘it would be an easy enough matter to change the basis of calculation to reflect votes cast rather than population.’ This, to put it bluntly, is bonkers. The number of votes cast in a constituency, and its relationship to the turnout in other constituencies, is not fixed. It will vary with each election and instantly throw the calculations out each time. If this bizarre suggestion were to be enacted would give rise to anomalies even greater than those under the current system. It is also dubious in principle, as it implicitly regards the non-voter as undeserving of representation. It amounts to a collective punishment of electors for low turnout (often the fault of the political system rather than the electors). It is an example, like the creation of constituencies that are not communities, of a suggestion made for the convenience of one group of politicians at the expense of what voters want from their local representatives.
There is no way of ensuring that FPTP produces equal treatment between two major parties. There are all sorts of reasons, including political geography, tactical voting (very important in the contrast between 1992 and 1997), the parties’ strategies, differential turnout, the distribution of each party’s vote, and – in a small way – boundary determination, which can affect the way FPTP works. Many of these factors work unpredictably. The only way of ensuring that there is a proportional relationship between votes and seats is to introduce a system of proportional representation – it really is that simple. Ferdinand Mount, and Keith Best of Conservative Action for Electoral Reform (CAER) are quite right to see PR, rather than tinkering with boundaries, as the solution.