Debating the Welsh Assembly electoral system

Alan Renwick

Peter Hain, the Shadow Welsh Secretary, has proposed today that the electoral system for the Welsh Assembly should be changed from the current proportional system to a plurality system using two-member constituencies.  This comes in the wake of rumours over the weekend that the government at Westminster would like to move in the other direction, making the current system more proportional.

So what are the options being discussed, and what are the issues we should be considering?

The current system is generally known in the UK as the Additional Member System (AMS) and in the rest of the world as the Mixed-Member Proportional (MMP) system.  Of the sixty Assembly Members (AMs), forty are elected by First Past the Post in single-member constituencies while twenty are elected from party lists in five regions.  The list seats are allocated so as to eliminate – so far as possible – disproportionalities in the constituency results: the idea is that, overall, the seats should be distributed across the parties in rough proportion to the votes they have won.

The phrase “so far as possible” is important: because the list seats make up only a third of the total and are allocated in regions that elect only four members each, there are sometimes too few of them to compensate fully for the disproportionalities in the constituencies.  In the most recent elections, in May, for example, Labour secured 50 per cent of the seats on only 37 per cent of the list vote.  That means that the current system, while more proportional than First Past the Post, is still less proportional than a pure PR system could be.

If the rumours are correct, the UK government’s idea is to change the ratio between constituency and regional seats from 40:20 to 30:30.  Presumably the five regions would be retained; they would elect six rather than four seats each.  For the reasons just discussed, this would make the system more proportional: it would be possible to compensate far more for the disproportionalities in the constituency results.

Peter Hain’s proposal, by contrast, involves a move in the other direction.  He too wants thirty constituencies.  But he proposes to abolish the list seats entirely and, instead, to elect two AMs from each constituency.  Whichever two candidates in a constituency won most votes would take the seats.

Hain says (on his website):

“The only acceptable option given the AV referendum result is to have all AMs elected by first-past-the-post, and we believe that each of the 30 new constituencies should elect two AMs by that system”.

This statement is full of holes.  First, to suppose that the result of a referendum on the choice between two non-proportional systems for Westminster says anything at all about whether voters want a proportional system for Cardiff Bay is nonsense.  Second, it seems that Hain does not entirely understand the system he is proposing: it isn’t First Past the Post.  Actually, there are two systems compatible with what Hain has said, depending on whether voters are allowed to vote for one candidate or for two.  If only one, the system could retain a moderate element of proportionality.  It would be the Limited Vote system, used in some UK constituencies between 1867 and 1884 precisely with the aim of providing some minority representation.

If, as seems more likely, Hain’s idea is that voters should be able to vote for two candidates (as in some British local council elections), then we have what’s called the Block Vote system.  This system is typically even more disproportional than First Past the Post.  Most voters cast both of their votes for candidates from the same party.  In most constituencies, therefore, the same party wins both seats.  The extreme case is where you have a single constituency and voters have as many votes as there are seats to be filled – the result is often that one party wins all or almost all of the seats.

What are the issues to think about here?  One point that should be recognized is just how self-interested all of it is.  Labour would win more seats under a less proportional system: in this year’s elections, they won 70 per cent of the constituency seats.  The Conservatives and Liberal Democrats, by contrast, would do better under a more proportional system.  As I’ve already suggested, the suggestion from Peter Hain that voters rejected proportionality in the AV referendum is plain silly.  Equally, the Conservatives cannot advocate greater proportionality for Wales without contradicting their outright condemnations of PR for Westminster.

Second, the underlying driver (or excuse) for the proposals is that the Westminster constituencies are changing: the number of Welsh seats at Westminster will be cut from forty to thirty.  At the moment, the forty Welsh Westminster constituencies are also the constituencies for the Welsh Assembly.  Hain argues:

“In Scotland the decoupling of constituencies – where Westminster and Scottish Parliament seats now have different boundaries – has been disastrous.  If that happened in Wales, you would be likely to have a situation where one Assembly seat straddled three parliamentary seats, which leads to confusion for voters, political parties and their representatives.”

But failure to decouple would also create problems.  The number of Welsh constituencies at Westminster will not be fixed: in fact, it could change every five years.  Do politicians really want the size of the Welsh Assembly to change every five years too?

Third, Peter Hain also refers (quite rightly) to the fact that many Welsh voters have been unhappy with having two types – or, as is often said, two classes – of AM.  But the main source of dissatisfaction is that voters have no control over who is elected from the lists.  The obvious solution is to open up the lists – allowing voters to rank the candidates – not to abolish them.

So the real issue (and this is my fourth and final point) is how proportional the system should be.  Here there is a range of legitimate views: proportionality is good in itself, but it carries costs, particularly in terms of the ability of voters to decide who should form the government.  The degree of disproportionality that Hain proposes is manifestly indefensible in the Welsh context: it would give Labour permanent, secure majorities on a minority of the vote for the foreseeable future.  On the other side, however, is there really a demand for greater proportionality?  The current system allows single-party government when a party is riding high, while requiring the parties to cooperate with each other when voters are more divided.  There’s a good case for saying that it gets the balance about right.

Aside for the need to open up party lists, therefore, the case for reform has not yet adequately been made.

Alan Renwick’s current research into political reform debates in the UK is generously supported by the Nuffield Foundation, to whom he is grateful.  Any views expressed herein and all errors are, of course, his own.