Local Elections Forecast for Scotland May 2017 (Pre-GE Announcement)

My forecast for the Scottish Local Elections in terms of change in share of the First Preference Votes from 2012 is SNP +9%, Labour -16%,  Conservatives +8%, Lib Dems -1% and Independents NC. In terms of change in seats, out of 1223 being contested, I forecast SNP +100, Labour -250,  Conservatives +150, Lib Dems -20 and Independents NC. The reasons are given below. This was compiled prior to Theresa May announcing a snap General Election to be held on June 8th. However, given that my forecast was largely based upon movements in Scottish GE opinion polls, the fact that the local elections will be held during a GE campaign might reinforce the influence on the local vote,

 

There are a number of factors that make forecasting the the outcome of the local elections just three weeks from now quite difficult.

  • In 2007 was the first election change to the Single Transferable Vote (STV) system, along with ward numbers being reduced, with three or four councillors per ward.
  • In 2007 the Labour Party did badly in the share of the of First Preference votes due to unpopularity of the Westminster Labour Government and also lost out due to STV. Compared to 2003, Labour’s share of the vote dropped 4.5% to 28.1%, but their share of council seats dropped 13.2% to 28.5%. The SNP were the big winners. From 2003 the SNP’s share of the vote rose 3.8% to 27.9% and their share of council seats rose 14.9% to 29.7% (source : tables 6 & 7 of Scottish Council Elections 2007 Results and Statistics).
  • In 2012 the Westminster Labour Party in Opposition were riding high in the polls. But in the Scottish Local Elections only improved their vote share on 2007 by 3.3% to 31.4% and their seat share by 3.7% to 32.2%. The SNP also improved on their 2007 results. The SNP vote share increased 4.4% to 32.3% and their seat share 5.1% to 34.7%. The Lib Dems suffered the biggest setback, with vote share almost halving to 6.6% and seat share reducing 7.8% to 5.8%  (source : tables 6 & 7 of  Lincoln Report on Report on Scottish Council Elections 2012).
  • Following the Independence Referendum of October 2014, the SNP, despite losing the referendum, increased their share of the Scottish Westminster voting intention opinion polls from around 30% to 47-48%.
  • Since the General Election (with SNP winning 50% of the vote and 56 of 59 seats) the SNP’s share of the Westminster voting intention has remained near 50%. The Conservatives have improved from 15% to around 25%, whilst Labour have declined from 24% to 15%. See Figure 1 below, lifted from the Electoral Calculus website.

  • The is a big gap in the polls between September 2015 and September 2016. But in May 2016 there was the Elections to the Scottish Parliament, with the constituency (and regional) vote shares of Con 22.0% (22.9%) , Lab 22.6% (19.1%), Lib Dem 7.8% (5.2%), UKIP 2.0%, Green 0.6% (6.6%) and SNP 46.5% (41.9%).  This bridges the gap in quite nicely, and is consistent with the polls.
  • In June 2016, the UK voted to leave the European Union by 51.9% to 48.1%. But in Scotland the vote was 38.0% to 62.0%. The SNP maintains a paradoxical position of wanting Independence from the UK, but to remain a part of the more opaque European Union. First Minister Nicola Sturgeon was hoping that Brexit would cause those Scots who voted to remain in both the UK and the European Union would come over to the Scottish Independence side. According to a large YouGov poll conducted last year and published in January, this switch has happened. But it has made no impact on the polls as there has also been a switch between those who voted for Independence from both the UK and the EU to support for remaining a part of the UK.
  • The STV system forces parties, if they wish to maximize the number of candidates elected, to optimize the number of candidates. My previous post was a long digression, based on the published candidate lists, on how of the three main parties, SNP and Conservatives appear to have adopted these optimization strategies. Conversely the Labour Party appears to have fielded candidates in line with the number of council seats held, failing to recognize that their support has about halved. As a result I would expect the Labour Party to experience a greater fall in share of seats than in share of the vote.

Forecast for Scottish Local Elections (pre-GE announcement)

Based on the above, I offer up quite a radical forecast for the Local Elections.

The Scottish National Party (SNP) look to have peaked in terms in popularity. Their vote share in the 2016 Scottish Parliament elections was slightly below the Westminster Elections of 2015. Estimation of the vote share in the Local Elections is a matter of allowing for how much the Independents will eat into the SNP share of the Scottish Vote. I think that a 41% share is a reasonable estimate. Share of seats I expect to rise by slightly less than vote share, which gives a rise of 100 seats. Note that with only 627 candidates, such a rise in seat numbers implies less than 1 in 6 of the SNP candidates would fail to get elected.

For the Labour Party, 2012 was an improvement on the 2007 result, but increase in the share of the vote was much smaller than the improvement in England or in Wales. That was due to the SNP taking up much of the fall in vote share of the Liberal Democrats and the Conservatives. Labour have been reducing in the opinion polls since the 2015 General Election and further since the Scottish Parliament election of last year. Given that Labour in the UK has been falling in the opinion polls to below 25% (see new lows with You Gov this last weekend), the share could be lower than 15% that I have assumed. Another factor is that due to the Independents, the vote share could also be lower. the forecast loss of 150 seats I believe to be quite cautious for two reasons. First, is that once a Party with wide spread coverage goes beyond a certain level of support, they fail to win as many seats as vote share, even with a proportional system like STV. The second reason is that Labour have too many candidates. There is an outside chance of seat share falling below 10%.

The Conservative Party are have increased their share in the Scottish opinion polls since last year. That resurgence is partly as a result of Brexit, with the party being the clear choice who support an Independent United Kingdom. But they are overcoming decades of increasing unpopularity. Whilst the opinion polls show the Conservatives at 25% or more, I believe that 21% is a more reasonable forecast. However, in terms of seats, that increase of 8% in first preference votes will be sufficient to mean a much larger increase in seats numbers. So this this time the share of seats will be similar to the share of the vote, whilst in 2012 there was a four point gap.

The Liberal Democrats, if anything, I believe will see another slight decline in their vote share. The Independents are the big unknown. But most of the seats held are in a small number of Councils. 125 out of 200 current Independent Councillors are in 7 out of 32 Council areas. In Orkney and Shetland all the council seats are held by Independents; in  Na h‐Eileanan an (Outer Hebrides) it is 21 out of 31; and in Highland 35 out of 80. I think the vote share will be fairly stable in May as well.

 

 

 

 

Labour in for a bigger thumping in Scottish Local Elections 2017 than polls suggest

The Labour Party are likely to suffer a bigger defeat in terms of seat losses than their drop in vote share. Under the Single Transferable Vote system if multiple candidates are on the ballot paper, then they split the vote. If there are more candidates on the ballot paper than is warranted by level of support (and number of candidates is greater than one), then too many candidates can result in less seats won. Recent opinion polls suggests Labour’s support is around half the level of 2012, yet in the councils where they are currently well represented, candidate numbers are about the same as existing Councillors, with multiple candidates in many wards. The Labour Party will suffer because they were not sufficiently ruthless in the face of cu

For local council elections Scotland adopted the Single Transferable Vote system in 2004. This May sees the third set of elections using the new system, the others being in 2007 and 2012. Under this system the local wards have a number of seats available. Just over half of wards have three council seats, with the rest having four council seats. But the majority of seats are in wards with four seats. There are up to five main parties (SNP, Labour, Liberal Democrats, Conservatives and Greens) and Independents fighting for those seats. The problem with trying to win a majority on a council is that the Party must try to win at more than one seats in every ward. But each voter only gets one vote, so multiple candidates of each party are competing for first preference votes with each other as well as candidates from other parties. If a Party believes that support is low in the area then they put forward just one candidate. If support is well over half of the electorate, then the Party might put forward three candidates, knowing that at least two candidates are likely to get elected. If the Party gets this wrong, and fields too many candidates, then its vote will be split. Most commonly fielding two candidates instead of one is a gamble. If got wrong the Party could either end up with either no seats instead of one. or one instead of two. This means a realistic prediction is necessary to optimize the number of seats. Elections Scotland has now published a candidate list for the forthcoming May elections. I have summarized the candidates per ward in figure 1.

The top section is a count wards where there a particular number of candidates. For instance in 27 wards the SNP have 3 candidates. The Green Party only has 1 candidate in each of 218 wards.

The middle section has the maximum council seats that a party can win, and the percentage of the 1223 council seats in Scotland. For the Independents the maximum seats is based on the assumption of only 3 seats per ward. Allowing for four or more candidates in four seats wards would make the theoretical maximum slightly higher.

The lower section gives the number of wards out of 354 with candidates and without candidates.

The results are surprising at first, particularly for the SNP who in the Scottish Parliament elections 2016 won 48.8% (63/126) of the seats, yet can win 51.3% of the seats available. Even allowing for about 14% of the vote in Local Elections going to Independents (source : Table 6  Lincoln Report on Report on Scottish Council Elections 2012) this appears to lack ambition. But comparison of the candidates to seats gained last time and position in the opinion polls in Figure 2 clarifies where the big anomalies lie.

The lower section of Figure 1 has been replaced by two more sections in Figure 2.

First is the seats won 2012 and the ratio of the maximum seats that can be won (identical to the number of candidates for the Political parties) to those seats won. In 2012 the SNP won 425 seats and in 2017 has 627 candidates. It therefore has a maximum seats to 2012 seats ratio of 1.48.

Second is the approximate value of recent opinion polls (less 14% to allow for the Independents)¹ and the ratio of maximum seat share to opinion polls share.

When comparing the figures between the parties the Liberal Democrats and the Greens do not really figure in the issue of having too many candidates. The comparison is between the three major parties.  It is here that the Labour Party clearly stands out. They have only got 15% more candidates standing in the local elections than Councillors elected in 2012. But, based on current opinion polls they have far more candidates standing as opinion poll ratings would suggest is ideal to optimize seat wins. The MSS / Polls ratio is partly exaggerated by the very low standing in the polls. But given that standing they should be adopting a similar strategy to the Liberal Democrats and the Conservatives. That is to only field more than one candidate in a ward where there is a reasonable expectation that the additional candidates will not compete against each other and lose seats. So why has Labour decided not defy the realities of the voting system? A strong reason is suggested in Figure 3.

In Figure 3 I have selected the 16 councils where Labour have (a) Candidates for at least a third of the seats available & (b) At least 10 Candidates. This is half the 32 councils in Scotland, but with 75% of Labour Council Candidates in 2017 and 82% of the seats Labour won in 2012. Numbers of candidates by council for the elections in May 2017 are compared with the seats won in 2012.

The number of candidates closely matches the seats won in 2012. This shows an expectation of winning significantly less seats in 2017 compared to 2012. However, it does suggest the reasons for adopting this strategy could be

  1. To promote the idea that Labour can still be the major party on the council, and be capable of a majority on 7 of these councils. This is especially important for Glasgow and North Lanarkshire, where Labour are currently in the majority.
  2. To help prevent second preferences being given to other parties.

However, if the Labour share of the vote on May 3rd is well below the 31.4% achieved in the local elections 2012, I predict the policy of aligning 2017 candidate numbers to seats held will result in a percentage of seats won will be significantly below the vote share. It is in those councils where Labour are currently strongest, the this disparity will be the widest. Top of the list of failures is likely to be North Ayrshire Council, where there are 17 candidates against 10 seats being defended. 

 Kevin Marshall

 

 

Notes

  1. I have assumed for General Election polling SNP 47%, Labour 15%, Conservatives 25%, Lib-Dem 6% and Green 4%. From the University of Strathclyde Scottish Opinion Poll Tracker the latest poll is for March 17. The figures are SNP 47%, Labour 14%, Conservatives 28%, and Lib-Dem 4%. Using these figures would make the Labour Candidate numbers even more out of line with the SNP or the Conservatives.