Skip to main content
Coronavirus (COVID-19): Latest updates.


The worst electoral system in the world?

23 January 2014

The publication of the Williams Report has prompted lots of discussion about the future of local government in Wales. Williams has recommended mergers of many current councils, and a significant overall reduction in the number of local authorities from the current 22. The report makes strong arguments for the potential benefits of council mergers.

Like just about everyone, I agree that if you were starting with a blank sheet of paper, you almost certainly wouldn’t design a local government system for Wales that included 22 local authorities. Nonetheless, I am also a little cautious about the Williams proposals. Plentiful adjustments to local government boundaries and organisation in the UK in the last half-century have not obviously produced enormous benefits. While the latest blue-print for change does seem more solidly evidence-based than some previous ones, we may also want to remember the strong and consistent finding in business studies research: mergers tend to substantially under-deliver on anticipated benefits.

Nonetheless, some change is probably on the way. And one topic that has already arisen in discussions about implementing Williams is how any new local authorities would be elected. There have been suggestions that support for implementation of Williams from both Plaid Cymru and the Welsh Liberal Democrats might be conditional on reform to the electoral system used for Welsh local elections. Both Plaid and the Lib-Dems support the use of STV in multi-member constituencies. This is the system used in Scottish local elections since 2007, and recommended for use in Wales by the majority of the Sunderland Commission in 2002.

STV would represent a substantial change to how we currently elect our councils. In the 2012 local elections (i.e., not including Ynys Môn) there were a total of 880 council wards in Wales, electing 1,221 seats. Of those, 580 (66%) were single member wards; there are 162 (18%) two member wards; 76 (9%) three member wards; 20 (2%) four member wards; and two five member wards (Morriston and Sketty both on Swansea). All wards chose their councillors (except in those seats that went uncontested) through some form of categorical, plurality voting.

In the single-member wards, councillors were elected via the Single-Member District Plurality system (SMDP, commonly known as first past the post). Multi-member wards, through which the majority of councillors were chosen, used what might be termed Multi-Member District Plurality (MMDP): here, voters are given X categorical votes, X being equal to the number of council seats to be elected, and the X people with the most votes in each ward are those elected.

The virtues and vices of SMDP have been discussed numerous times (including on this Blog). Among the suggested virtues are simplicity and familiarity. There are also some arguments in favour of single-member districts: notably that they can be kept relatively small in size; and that there is a direct link between a single, individual representative and her/his area. If one is going to choose councillors from single-member wards, then SMDP is a perfectly defensible system to use. Personally I would marginally favour the Alternative Vote system, but there probably isn’t a great deal in it either way, and the simplicity and familiarity of SMDP do weigh in its favour.

MMDP is quite another matter, however. This is a system that combines all the vices of SMDP with none of the redeeming features. The use of multi-member wards means that what some defenders of SMDP for Westminster have termed the sacred bond between the single representative and their district is lost. The simplicity of SMDP – one person, one categorical vote – is also abandoned. At the same time, the principal vice of SMDP is maintained (even, in some scenarios, enhanced): the substantial, at times outrageous, disproportionality between the votes of the people and the party balance of the representatives who are elected. It is far from unknown for one party to win 100% of the representation from a multi-member ward on a minority (sometimes quite a narrow minority) of the vote. Repeated across different wards, this can make particular councils grossly imbalanced – at times looking almost like ‘one-party states’ – on a minority of the vote. MMDP is quite possibly the worst electoral system currently in use anywhere in the world. It is understandable that no Sunderland Commission member favoured retaining this system.

Implementation of Williams would, presumably, mean cutting Wales’ currently excessive number of councillors. (We have almost exactly the same number as in Scotland, despite being much smaller both in territory and population). Council wards would need to be re-drawn. The introduction of new wards for new councils would seem to be an ideal time to also introduce any new voting system. While the mixture of SMDP and MMDP currently used may be the ‘default’ option, I very much hope that the case for STV is re-visited. STV would promote greater proportionality; the make-up of councils would represent popular views more accurately than under the current system. There would thus be far fewer one-party states (unless the voters had chosen to give a single party an overwhelming mandate).

There may be other virtues to STV as well. The elections would be more likely to be genuinely competitive. In 2012, 8.0% of council seats in Wales went uncontested; in Scotland, where STV was used for the simultaneous council elections, there were no uncontested seats. Studies of STV where it has been used also find that it tends to promote strong local representation: the possibility of having to compete with other candidates from one’s own party, as well as those from others, and the need to win lower preferences if not the first preference of voters, places a high premium on local visibility and the personal reputations of representatives. In the context of a reform of local government that may well make councils appear more distant to some people, this is of value.

Reform of local government in Wales is very much on the agenda now. One valuable reform would be to get rid of the system by which the majority of our current councillors are elected – a system that is a strong contender for the title of the worst electoral system in the world.


  1. dp

    Single-member wards… would you have other comments pro and con? I’m looking for evidence of how they work in medium-to-large cities using FPTP.
    I am v doubtful as to their fairness. A winner-takes-all in wards with moderate-to-strong showings of more than one party means that a sizable portion of voters are shut out, whereas a multiple member ward means there’s usually scope for different representations. Add to that the necessity of cross-party working at ward level: the residents benefit from the lack of party monopoly, and the city benefits from a greater degree of integration.
    I’ve been around enough to see single-member councillors who are put in place by the machinery, and who don’t give a damn for the locale. It’s disaster.
    So I’m puzzled why anyone would think single member wards are in any way credible.

    • Roger Scully

      The main advantage of single-member wards, as far as I am concerned, is that they can be small, and so maximise the extent of local representation.

      Personally I favour multi-member wards. What I oppose is using a first past the post type of voting system for multi-member wards, which can often mean the same party winning all the seats on a minority of the vote. The STV system introduced for Scottish local elections in 2007 is much fairer between the parties, and also tends to mean that all seats are to at least some degree competitive.

Leave a reply

Your email address will not be published. Required fields are marked *