Uncategorized

STV in Wales: a follow-up post.

As a follow-up to my previous post about STV, and in response to queries from a couple of people, I thought the following might be of interest to some readers of the Blog.

As mentioned in my previous post, the 2011 Welsh Election Study not only asked respondents immediately after the election how they had voted, but also fielded a question which asked:

“Please indicate how you would have voted in the National Assembly for Wales election if you had been asked to rank the parties in your order of preference. Put 1 for your most preferred party, then 2 for your second best party, 3 for your third choice etc. You may rank as many or as few choices as you wish.”

Most interesting, I think, about answers to this question is not who wins the first preferences (first preferences won by the parties were very close to the proportion of actual votes that they won in the Assembly election) but who is then chosen as the second preference.

Below I list the second preference choices by party: i.e. what those who said they would favour a party with their first preference then said they would do with their second preference. (The percentages given don’t sum up to 100: some people only indicated a first preference, while a handful gave their second preference to parties too minor to be listed here):

Of those indicating that the Conservatives would have received their first preference vote, second preferences were allocated as follows:

4.2% Labour

23.5% Liberal Democrats

13.9% Plaid

6.9% Greens

30.5% UKIP

 

For Labour first preference voters, second preferences were:

3.6% Conservatives

10.3% Liberal Democrats

44.9% Plaid

9.0% Greens

4.7% UKIP

 

For Liberal Democrat first preference supporters, second preferences were:

23.5% Conservatives

26.9% Labour

15.1% Plaid

21.0% Greens

3.4% UKIP

 

For Plaid Cymru supporters, second preferences were allocated:

1.6% Conservatives

30.9% Labour

21.6% Liberal Democrats

15.4% Greens

1.6% UKIP

 

For the (small number of) Greens, the proportions of second preferences were:

2.6% Conservatives

30.8% Labour

17.9% Liberal Democrats

17.9% Plaid

0% UKIP

 

Finally, for UKIP supporters, the profile of second preferences was:

37.8% Conservatives

13.5% Labour

5.4% Liberal Democrats

5.4% Plaid Cymru

4.1% Greens

 

So what can we make of these numbers? Overall, I think they reinforce what I said in my previous post. To do well in a preferential voting system, a party first of all needs to win lots of first preferences. But it also need to be fairly attractive to supporters of other parties to pick up vote transfers. The data suggests that while the Conservatives would do reasonably under STV in terms of first preferences, they would struggle to win many vote transfers. (I’ve only shown second preferences here, but things don’t get any better for them when you look at third or even fourth preferences). On second preferences, the Tories are not even the plurality choice among the diminished ranks of their Liberal Democrat coalition allies. They are only the most popular choice among (the then fairly small number of) UKIP supporters.

Labour would seem to do well out of STV. Not only does it gather a large number of first preferences; it is also the most popular second preference choice for Liberal Democrats and Plaid Cymru supporters, as well as among small number of Greens.

Plaid would also seem to do quite well out of STV: certainly, they are much the most popular second preference choice among the large pool of Labour supporters. However, this finding also re-emphasises a point I have made previously: there were lots of voters in 2011 who felt quite sympathetic to both Labour and Plaid, but Plaid was very poor at getting many of them to make it their first choice.

Comments

  • Tal Michael

    Very interesting. I have had another look at my analysis based on this – and under two-seat STV, I now think Labour 43, Conservative 23, Plaid Cymru 8, Liberal Democrat 5 and Independent 1 (I had missed the strong showing from an Independent in Torfaen – which put her ahead of the Conservative for the second seat). I am guessing the sample size of the poll is too small to tell us who Independent voters will put second (and it probably varies depending on who is the incumbent in that seat). What about BNP and Socialist Labour voters? Although they both got just 2.4% across Wales, there may be enough of them to affect the outcome in some seats…

  • Welshguy

    Second preferences from a party that is larger (locally) than one’s own are mostly useless – votes for that party are likely to go to electing their representatives rather than being redistributed to the second preference. So Plaid Cymru, for example, is unlikely to benefit from the large number of Labour voters who rank them as a second choice throughout most of Wales – only in Ceredigion are significant numbers of Labour votes likely to be distributed (similarly, UKIP’s gaining lots of second-preference Tory voters is unlikely to translate into any seats, as you’d generally expect the UKIP candidates to poll lower than the tories in first preferences and thus be redistributed).

    Interesting to note that people’s first preferences were not notably different from the way they voted in FPTP – part of the point of STV is that it allows you to vote for your genuine favourite, even if they are unelectable, without worrying about “letting the XXX in”. Then again I suppose this was just a quick poll and one would expect people’s voting patterns to change if the actual system changed.

    • Roger Scully

      Not quite true, I think, Welshguy. STV not only re-allocates the votes of candidates eliminated (those at the bottom of the pile); it also re-allocates surplus votes over-and-above the ‘quota’ needed for election.

      Therefore, in seats where the leading one or two Labour candidates would win election easily, their second preferences would still be of some relevance.

  • Mary Wimbury

    Interesting – is the data on second preferences published anywhere?

Leave a Reply

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