(Updated below)
The preliminary results of the Scottish parliament elections are as follows (given as total seats as a sum of constituency and list seats).
-
SNP: 47=21+26
Labour: 46=37+9
Conservative: 17=4+11
LibDem: 16=11+5
Green: 2=0+2
independent: 1=0+1
The Scotsman has the detailed breakdown of constituencies and regions.
Speaking of breakdown, the voting and counting was riddled with problems.
The Electoral Commission announced it will conduct a full inquiry into problems with the new electronic counting system that resulted in several counts being suspended. As many as 100,000 ballots were also rejected because they were classed as spoilt papers.
These are two separate problems. The ballots are paper ballots–separate papers for the assembly (with two columns, one for the constituency vote and one for the regional party list) and the local council (ranked choice, STV). So any problem with the electronic counting should be easily resolvable: the ballots can be re-tallied by hand. However, the problem of spoiled ballots may have resulted from voter confusion: not understanding the different kinds of marks needed on each ballot (a single check in each column on the MMP ballot and ranking candidate choices on the STV ballot).
In a separate article in The Scotsman, Ken Ritchie, chief executive of the Electoral Reform Society, notes, “While the counting equipment has experienced teething problems in some areas, it is not the equipment that has caused people to make mistakes in the completion of their ballot papers.”
SNP leader Alex Salmond adds: “It is also the case that the decision to conduct an STV election at the same time as a first-past-the-post ballot for the Scottish Parliament was deeply mistaken.”
That’s right. Some degree of voter confusion may be inevitable in a first (which surely now will be the last) simultaneous use of categoric and ranked-choice ballots. Still, in examining ballot papers, some of these errors would seem recoverable.
For instance, if a voter marked a 1 (and other numbers) on the FPTP or party-list ballots, it’s pretty easy to know what the vote was intended to be (the lowest number on each side of the MMP ballot is the first–and only valid–choice on that part of the ballot). On the STV ballot, apparently voters were required to rank at least three candidates for the vote to be valid. So, if voters marked only one, there is not much that can be done. This is an unfortunate aspect of the law, and a worse decision (in my view) than the simultaneous use of MMP and STV.
In any event, I agree with Malcolm that, as bad as the failure to register so many votes in Scotland is, the bigger democratic failure yesterday was in the local councils of England. There, under FPTP and England’s increasingly multiparty system, large numbers of votes may have been literally counted (they will register in the final vote totals) but they will not count in the more fundamental sense of helping elect any candidates. The wasted votes are very high in many council races, with many councils having massive seat majorities for parties with small pluralities of the vote, or even less than a plurality.
Meanwhile, in Wales, Malcolm notes:
Conservatives should once again be proclaiming the benefits of a more proportional system, but are strangely silent. This time, they have gained 4 more constituency (FPTP) seats than 2003 – in part because of the electoral base that their ‘top-up’ regional list AMs [Assembly Members--ed.] have given them over the past years; despite being absent from parliamentary representation between 1997-2005.
- Labour: has gained an extra list seat to compensate a little for losing a number of constituency AMs. So even Labour is starting to get the benefits of the system.
__
Update: The BBC has the votes percentages.
First, the list vote (with the overall seat percentage):
-
SNP, 31.0 (36.4)
Lab, 29.2 (35.7)
Con, 13.9 (13.2)
LD, 11.3 (12.4)
Grn, 4.0 (1.6)
BNP, 1.2 (0.0)
It is interesting to compare the vote percentages in the nominal tier (here the number in parentheses is the percentage of constituency seats won)
-
SNP, 32.9 (28.8)
Lab, 32.2 (50.7)
Con, 16.6 (5.5)
LD, 16.2 (15.1)
Grn, 0.2 (0.0)
BNP, –
From these numbers, what jumps out is the extent to which the Labour party was advantaged by the FPTP portion of the system, winning a majority of the seats allocated that way on less than a third of the votes, and fewer votes in both tiers than the SNP. The list tier “top-up” restored the correct ranking of the two parties’ seats, relative to votes, but barely! Labour remains more over-represented than the SNP (advantage ratio of 1.21 for Labour, 1.17 for SNP).
It is obvious that there was considerable ticket splitting. That is, for all the attention to the spoiled ballots caused by voter confusion, the bigger picture is that the electorate as a whole appeared to understand the strategic implications of the two votes. (It is interesting that among the major parties, it was the third and fourth parties that did better in the nominal vote; that reflects districts where they had the chance at the local plurality against Labour or the SNP.)
The detailed Welsh Assembly results also suggest considerable ticket-splitting.
The unfortunate breakdown was all over my Google alerts today. FairVote actually has a large delegation over there monitoring the elections and meeting with key people.
One thing impressed me, though. Rather than convict the “Byzantine electoral system,” there was – at least among some papers – a pretty thoughtful level of discussion. Some papers talked about ballot design and the “vote for two” issue. Others talked about using two very different electoral systems side-by-side.
All that means, rather than reform rollback, Scotland will see an improvement next cycle.
Seed planted by Jack — 04 May 2007 @ 17:50
NZ had some issues with the introduction of MMP as well. E.g. if you voted in the wrong electorate, then your national-list party vote was invalid. The mechanics of voting have since been updated.
Seed planted by Errol — 04 May 2007 @ 19:23
There seems to have been a fair amount of trouble with the ballot counting machines, no matter that the returning officer and machine contractors want to discount it. The Scotsman reports mangled ballots and, in Islington, seven out of thirty machines breaking down. In addition, there was a problem with mail-in votes (seems to be the same sort of thing that has been troubling Australia for quite some time now). Finally, there was a problem of getting remote ballots to a place where they could be counted. The Arbuthnott Report cited this as a possible problem arising from too much tech, as I recall.
Just raging against the machine a little…
Seed planted by CCBC — 04 May 2007 @ 20:59
The thing about mail-in ballots in Australia is pretty much a furphy.
The ruling Coalition parties raised a number of issues in the Joint Committee on Electoral Matters review of the 2004 federal election. All were allegations of possible electoral fraud that would be familiar to anyone who’s heard similar allegations from a certain political party in the US.
The independent electoral commission investigated the allegations thoroughly and found there was no substance to them.
The Coalition went ahead and amended the Electoral Act anyway. The amendments would be familiar to anyone who’s seen similar proposals from a certain political party in the US.
Seed planted by Alan — 06 May 2007 @ 00:48
I see from later coverage of the Scottish vote that they didn’t learn from one the NZ’s errors – votes with no local vote had their list vote discarded. There is no excuse for this.
Seed planted by Errol — 07 May 2007 @ 01:32
If Scotland pursues the Arbuthnott Commission’s recommendation to use some kind of flexible list system in 2011, they will have to deal with the regional “overhang” problem: excess local seats won by a large party, beyond their share of the total vote in the region.
In any regional MMP model, the more regions, the greater the chance of regional overhangs. Yet if you want more accountability of list members, an “open” or “flexible” list system requires manageable lists, that is, smallish regions.
The problem is, if you have a fixed number of MPs per region, what do you do about overhangs? But if you add seats to balance the overhangs, you unbalance the regions. This helped lead the Ontario Citizens’ Assembly to choose province-wide closed lists. How will Scotland solve this?
One answer might be the German habit of rewarding turnout: instead of a fixed number of seats per region, allocate the party’s seats nationally, and assign them to regional list seats based on actual votes cast.
This has spread to Poland. In their 2004 election to the European Parliament they used a pure regional list system, but unlike France, they allocated seats between parties nationally — using highest average, favouring larger parties — but then assigned them within each party to each of the 13 regions — using highest remainder, favouring smaller regions, but also favouring regions with higher turnouts. With only 54 MEPs the regions average only four MEPs each, and since eight parties got over the 5% threshold, that’s some interesting math.
Should MPs represent electors or actual voters? Baden-Wurttemberg thought about this too, when they designed their best-runner-up list-free MMP system: the “best runner-up” is the party’s candidate with the highest number of votes in the region, not the highest percentage. Again an area with a higher turnout gets more representation.
Would Scotland let regions get more MSPs if the region had a higher turnout?
Seed planted by Wilfred Day — 16 July 2007 @ 12:17