This Blog Post is the second in a mini-series I am running on Electoral Systems. In the first I briefly described the main systems currently used, or proposed for use, in Wales. The remaining posts in the series, starting with this one, will look at these systems rather more critically.
I start with ‘First Past the Post’, or SMDP. This system has the virtue of familiarity, having long been used for UK general elections. It also has the virtue of simplicity. What other virtues does it have? They are often suggested to include the following:
- Unlike closed party list systems, SMDP allows people to vote for a person rather than a party
- SMDP facilitates local representation
- SMDP produces clear outcomes to UK general elections, thus facilitating strong and effective government
The first alleged virtue is rather dubious. SMDP is every bit as much a closed list system as the party lists used for European elections in Wales; the only difference is a district magnitude of 1 rather than 4. Voters are still presented with the take-it-or-leave-it choice of a party’s chosen candidate, and cannot favour a particular representative of a party.
SMDP does facilitate local representation: everyone has ‘their’ local MP. Of course, different representatives can vary in the diligence with which they conduct constituency work – though we seem long past the days when some MPs would make an ‘annual visit’ to their constituency! Most representatives elected under SMDP take this dimension seriously. And in research on the electoral system change implemented for European Parliament elections in 1999, David Farrell and I found that moving from a constituency to a regional list system did seem to reduce the focus on local representation among British MEPs. But SMDP is not the only electoral system to facilitate local representation. The Alternative Vote, rejected in the 2011 referendum, would have maintained the single member link to a constituency; while STV arguably provides for even stronger local representation.
The third alleged virtue for SMDP is also starting to look dubious. There is significant research which severely calls into question whether single party government tends to be more effective. The evidence is complex, but it is very difficult to argue that single-party majority government is clearly superior in terms of delivering better government. Furthermore, SMDP manifestly failed to provide for a clear single-party victory at the last UK general election and, as John Curtice has pointed out, is becoming significantly less likely to do so in the future. For various reasons (including, but not limited to, the rise in the number of MPs who are neither Labour nor Conservative) ‘hung’ parliaments at Westminster have become much more probable.
When electoral systems are discussed, it is often in a way that implies that particular systems operate in ways that do not change much over time. But that is evidently not so for SMDP in the UK. SMDP, I would suggest, was a reasonably satisfactory electoral system during the era of strong two-party politics – the 1950s and 1960s. OK, the electoral system – in line with Duverger’s alleged Law – had arguably done much to create this form of politics, by squeezing out the Liberals from 1920s to the 1940s. But with both major parties winning close to 50% of vote much of the time during the 1950s and early 1960s, any distortions created by the SMDP system were modest in scale. (Although these distortions were sometimes still important. Perhaps the most notable one came in the 1951 general election, which Labour lost despite winning the most votes).
However, SMDP as a system has become ever less satisfactory as the hold of two-party politics on British society has declined. The symptoms of this decline well known. There has been a substantial fall in the vote share won by the two largest parties: from over 90% in all the four general elections held in the 1950s to merely 65.1% in 2010. There have also been substantial declines in the extent to which people identify with the two largest parties, and in party membership numbers.
The combination of an unchanging electoral system amidst a changing politics has produced increasingly large distortions of the public will in election outcomes. The blue line in the chart below shows the Gallagher Index of Disproportionality* for Britain in every general election since 1950. The trend of the line is clearly, and strongly, upwards. We have, in effect, an electoral system straining – with ever greater effort – to uphold a two-party politics that much of the British public long ago ceased to believe in. For this reason, SMDP has become substantially less appropriate as an electoral system for House of Commons elections.
The second (red) line in the above figure gives Gallagher Index figures for Wales at general elections. As can be seen, in Wales the line is generally higher (indicating greater levels of disproportionality) although the slope of the line is less consistently upwards. In Wales, SMDP has for a long time helped to uphold one-party dominance. Labour has won most votes at every post-war election, but with the help of the electoral system has won a majority of Welsh seats even when its vote share fell well short of 50%.
A vivid illustration of how SMDP tends to work in Wales came in 2010. Across Britain, the Conservatives 36.1% was enough for largest party status in the House of Commons, but some way short of a majority. In Wales, Labour’s 36.2% vote share enabled them to win 65% of the seats.
The 2011 AV referendum fiasco substantially damaged the prospects for electoral reform (at least for UK general elections) for some time. There currently seems little chance of change in the short- or even the medium-term. What this means is that the UK will retain an electoral system for House of Commons elections that is ever less suitable for the society over whom it operates; an electoral system that in my view has become substantially less fit for purpose.
* The Gallagher Index runs from 0 (where the percentage of votes won by each party in an election perfectly matches the percentage of seats they get) to 100 (in the somewhat implausible scenario that all parliamentary seats go to parties that won none of the votes!) In practice, most countries with explicitly proportional voting systems produce Gallagher scores in low single figures for their national parliamentary elections. A score approaching or exceeding 10 indicates significant disproportionality. For further details, see: Gallagher, Michael (2013) Election Indices Dataset at http://www.tcd.ie/Political_Science/staff/michael_gallagher/ElSystems/index.php.