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


Electoral Systems and the Electoral History of Wales: Labour

4 June 2018

This is the second blog post in a series, drawing on research from my forthcoming book on Elections in Wales, which examines the impact of electoral systems on the electoral history of Wales.


It is often impossible to be wholly definitive about the impact of electoral systems on election outcomes. We cannot know for sure what election results would have been like under a very different voting system. We can normally (at least roughly) model the mechanical effects of an electoral system: that is, to see how the same votes cast and counted under another system might have generated a rather different result. But electoral systems have effects on politics that go well beyond the mechanical process of translating votes into seats; their influence can permeate the broader political culture. How we elect our politicians can shape what we expect them to do. For instance, the idea that a parliamentarian is the specific, individual representative of a defined geographical area, with the responsibility to represent the interests of that locality and all the people within it, is usually absent from countries that elect their MPs from national or regional lists. Electoral systems can also shape strongly how politicians relate to their political parties – not least because different systems can mean that politicians depend to a greater or lesser extent on their own efforts, as well as those of the party, for their electoral fate. And the electoral system can strongly influence how a party seeks to fight the election: does it try to raise its vote across the board, or does the electoral system incentivise a focus on particular places (such as marginal constituencies)? It is very difficult, if not impossible, to model precisely all the effects of electoral system change.

However, the mechanical effects of electoral systems can at least be explored. And given the long history of one-party dominance in Wales, it makes sense to begin examining the impact of electoral systems by considering their contribution to this defining reality of Welsh political life. To what extent has the FPTP electoral system used in general elections exaggerated or reinforced Labour’s dominance of Welsh electoral politics? The figure below plots Labour’s general election performance in Wales from 1945-2017, showing the share of the vote achieved by the party and the share of seats that they won at every general election. While Labour came first in votes every time, its share of Wales’ parliamentary representation was always still greater. This finding should not surprise us – it is what the FPTP system tends to do. The bottom line in the chart shows the percentage difference between seats and votes won by Labour at each election. We see two upwards spikes, followed by respective periods of decline. Over the entire period the broad trend is, if anything, upwards: that is, the extent to which Labour has been over-represented in the number of Welsh seats it has won, relative to its share of the vote, has slightly increased over time. But what is certain is that the electoral system used for UK general elections has done nothing to undermine Labour’s dominance in Wales.


Labour General Election Performance, Wales, 1945-2017

(click on the figure for a larger version)

The AMS system used for NAW elections ought to produce a rather different relationship between votes and seats than does FPTP. Under AMS, sixty AMs have been chosen in each Assembly election. Forty are elected under FPTP rules within constituencies – with those constituencies always, thus far, having been based on the same boundaries as the ones used for Westminster. Four AMs are then chosen as regional members from each of five regions: North Wales, Mid and West Wales, South Wales West, South Wales Central and South Wales East. Crucially, the allocation of regional members takes account of which parties have won the constituency contests; this makes the overall election outcome more proportional. However, in Wales only one-third of the seats are allocated regionally – this contrasts with other places where versions of AMS are used, where the proportional element of the system typically comprises 40-50 percent of the seats. The implication of this is that while there is a proportional element to the Welsh Assembly electoral system, the system is some way from fully-fledged proportional representation.

We can see the impact of AMS on Labour in the next figure. This shows the difference in percentage of votes and seats won by Labour at all the general elections and Assembly elections from 1997 onwards. (The Labour vote share for Assembly elections in the figures is calculated as the average of the party’s share across the two ballots.)


% Seats minus % Votes for Labour in Wales, 1997-2017

(again, click on the picture for a larger version)


In every one of these elections, Labour was over-represented – its percentage of the seats won was greater than its percentage of the vote. But in all of the Assembly elections, this over-representation was lower than in any of the general elections. In short, AMS as applied to Assembly elections was still favourable to Labour, but rather less so than FPTP has been during the same period. However, the trend in terms of Labour over-representation has been broadly upwards for Assembly elections and downwards in recent general elections, so the impact of the two systems has converged to some extent.

In my next blog post, I’ll broaden the focus out from Labour, to look at the impact of electoral systems on the party system in Wales as a whole.


  1. Geraint

    Is it right to put the green line on the first graph? The two other data sets are percentages. The green line shows the actual number of MPs elected and over the period the number of MPs has increased from 36 to 40. I guess from the thrust of your arguement in your blog the difference it is not significant to the points you are making, but to me it is a fraction awkward.

Leave a reply

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