As per usual for all tables click to enlarge
In a previous post we analysed the rise in UKIP poll support. The bulk of increased UKIP support has come from disaffected 2010 Conservative voters, where nearly 1 in 10 Conservative 2010 voters now support UKIP. But how much further can UKIP rise? More importantly how can UKIP win Westminster seats?
We will have a look at the increased levels of UKIP support by modelling two scenarios of increased UKIP support. Each scenario models the shift in the vote by assigning proportions of other parties’ 2010 voters to UKIP. The basis of each scenario is as follows:
Current polling level of UKIP support – 8.2%
This first scenario is the current shift in voting intentions assessed in our previous post where we aggregated seven YouGov polls from mid to late April. In this aggregation UKIP support across Great Britain was at 8.2%. We shall refer to this scenario as the Current scenario. To recap the flow each of the major parties to UKIP was as follows
Hypothetical Ceiling of UKIP support – 19.5%
The second scenario attempts to capture all possible UKIP voters. These are voters who contemplate voting for UKIP but have not actually done so. If UKIP won all of these voters it would be the maximum level of their support. We shall refer to this scenario as the Hypothetical Ceiling scenario.
To attempt to model the hypothetical ceiling of UKIP support we shall use voters’ second choice data from the 2010 British Election Study. The 2010 British Election Study captured about 17000 respondents, so its figures are robust. We will sum all voters of other significant parties in 2010 that indicated UKIP as their second choice to create a hypothetical ceiling of UKIP support. We are accumulating all voters that may be persuaded to vote UKIP.
We include a table below calculating the hypothetical UKIP support ceiling.
THE aggregation of 2010 UKIP voters and voters of other parties that indicated that UKIP was their second choice creates a hypothetical UKIP support ceiling of 19.5%. This is well above the current polling figures of 8%, so further rises in UKIP poll levels are still possible.
It is important to note that most of the potential UKIP support comes from Conservative 2010 voters. The road to increasing UKIP support further from the current opinion poll level of 8%, is through these voters. However, nearly a quarter of potential UKIP support comes from 2010 Lib Dem and Labour voters.
We would suggest that obtaining 2010 Labour voters will be the hardest task for UKIP as Labour is in opposition and thus is far less likely to alienate their 2010 voters who stuck with the party at the last election when it was defending the baggage of 13 years in government.
On the other hand, Liberal Democrats are in Government and thus have shed their largest non-major party tag. Voters that have previously voted Liberal Democrat as a “none of the above” option now need a new home, which UKIP can provide. This distinction is very important in relation to winning seats as we shall see below.
Collapse of the Lib Dem Vote
For both of the above scenarios we have also made provision for the collapse of the Lib Dem vote by factoring that 40% of 2010 Lib Dem voters would switch to Labour. Avid readers of this blog may recall that when we previously assessed where the Lib Dem 2010 vote had gone we had determined that 28% had gone to Labour. We have adopted the higher 40% rate on the basis that the large 22% of 2010 Lib Dem voters who don’t know who they will vote for will split between the Lib Dems and Labour. We further assume that the Lib Dems will hold on to all of their other voters except the share in each scenario that we assign to UKIP.
The result that in both scenarios the Lib Dems hold between 50 & 60% of their 2010 vote (win 12-14% of the GB vote), which is a little optimistic on current polls, but is designed to hedge against a dramatic fall in the Lib Dem vote.
Use of Current Boundaries
We will now model these two scenarios using the current set of 650 Westminster constituencies. UK political watchers would be aware that current legislation has set up a process to contest the next general election on a reduced 600 constituencies. However, the new boundaries once finalised in late 2013 will require an approval by parliament which is not guaranteed. Furthermore the new boundaries, even if approved, will only come into force at the next general election. By-elections in the interim will be fought on the old boundaries which we should assess to see which constituencies would be vulnerable for a UKIP by-election gain.
Under none of the scenarios, did UKIP win any constituencies, but keep in mind this modelling is just a complex uniform swing. As election watchers know there is no such beast as a uniform swing in real life. Now onto some more detailed results. Firstly we address the Current Scenario
Current Scenario – Constituency Implications
Under this scenario, UKIP does not even come close to winning any seats. UKIP wins one pyrrhic second place in the super safe Labour constituency of Bootle with 8.6% of the vote. UKIP wins 75 third places in this scenario, up from 3 at the last general election.
The highest constituency vote it would obtain is in Boston and Skegness at 14.9%. However the constituency that the UKIP candidate comes closest to first place is Devon North where the UKIP candidate at 12.9% is 20.3% behind the first place candidate.
This is the first clue to the big conundrum that UKIP faces in winning seats. The constituencies where UKIP gains the most votes are not necessarily the seats where they have the best chance of winning.
We include two tables below which indicate the 20 seats with the largest UKIP share of the vote and a table with 20 smallest UKIP constituency deficits, the amount between the UKIP candidate and the leader in a given constituency. We have also included both metrics for the seats included in each table.
We have also colour coded each constituency by the party that currently holds the seat (not the projected winner under the model).
As you can see above in the Current scenario, UKIP polls the highest shares of the vote in mostly Conservative held seats. However it is mostly in Lib Dem held seats where UKIP are closest to the leading candidate.
Hypothetical Ceiling Scenario – Constituency Implications
In this more expansive scenario for UKIP, whilst they win no seats, UKIP comes within 10% of the leader in 45 constituencies. However, in all but one of these constituencies, this distance is more than 5%. It is these 45 seats that can be considered UKIP target seats, if the purpose is to win the seat.
While coming much closer to winning seats , UKIP also wins 134 second place finishes, 83 of these behind the Conservatives, mostly in the south of England, and 51 second place finishes behind Labour, mostly in the north of England. Under this scenario UKIP wins 420 third place finishes.
The largest UKIP constituency vote share is obtained in Christchurch at 28.4%, this would be a second place to the Conservatives at 40%, which puts it outside the 10% range indicated above. Once again the seats where UKIP polls the highest vote share are not necessarily the seats where they have the best chance of winning.
The closest UKIP come to winning a seat under this scenario is in Devon North where UKIP comes third on 23.3%, only 2.3% behind the Tories on 25.6%, Labour wins 23.7%, while the Lib Dems, who currently hold the seat finish fourth on 23.2%.
We again set out the tables of constituencies by vote share and distance to the winner below. For the latter table we have increased it in size to the 45 constituencies where UKIP polls within 10% of the leader. We will refer to the latter table as the UKIP target seats table.
In terms of UKIP winning seats at Westminster , the last table is the most important. This table lends itself to several important observations.
The largest UKIP vote shares do not equate to best chance to win seats
Firstly, what is important is not the size of the UKIP vote, but how far away UKIP is from the leader. Therefore seats where UKIP gains the most votes are not necessarily the seats where they have the best chance of winning.
In the above target seat table, constituencies which are in bold are where UKIP won 7% or more in the 2010 election. There were 20 such constituencies. It is indicative that only seven of these constituencies made it into the list of 45 constituencies where UKIP has the greatest chance of winning.
These constituencies are where UKIP’s largest potential vote approaches an amount where they can win the seat. If UKIP focuses its resources on these constituencies they can maximise their vote and come close to winning seats even if UKIP’s national vote share is less than 19.5%.
UKIP’s best chances are in former Con-Lib Dem 2-way contests that devolve into a four-way contest.
UKIP has the best chance of winning seats in former Conservative-Lib Dem 2way contests. In these constituencies under the UKIP hypothetical ceiling scenario both the Tory and Lib Dem vote shares are falling because Conservative voters are flocking to UKIP and Lib Dem voters are flocking to Labour turning these races into 4 way contest between the said parties. Devon North is the best example.
The major problem for UKIP in this situation is that the major argument against voting UKIP for otherwise Conservative voters would be that it would split the right-wing vote allowing Labour or the Lib Dems to win the seats. In a four way contest this is a powerful argument.
The only way to circumvent this argument that UKIP splits the right-wing vote to the benefit of the “left” is to threaten to win safe Conservative constituencies where no other party has a chance of winning. However to win such seats where the Tories would have over 50% of the vote, UKIP needs to take more than 29% of 2010 Conservative voters. This is beyond even the generous Hypothetical Ceiling scenario.
At lower turnout by-elections UKIP has a better shot of electing an MP
In general elections even under the most favourable scenario for UKIP, the Hypothetical Ceiling scenario, UKIP has a hard time winning any constituencies. However in by-elections UKIP has a much better chance in all the above listed seats if they can keep their supporters enthusiastic in an environment where all other parties struggle to turnout their voters.
Therefore the above list should also be seen as a UKIP by-election target list. Arguably, the UKIP’s best chance of winning seats is by winning a by-election to prove they are a genuine chance to win FPTP contests.
UKIP chances to elect MP’s is assisted by the Lib Dem collapse
The above table is contingent of 40% of Lib Dem voters defecting to Labour, which is a given in the current political environment. However in the medium term this may change, then all bets are off.
The Lib Dem collapse in support brings many current Lib Dem held constituencies seats into play for UKIP, especially in the South West. Furthermore the residual Lib Dem votes in these seats soak up a quarter of the vote which lowers the winner’s final vote share, making it easier for UKIP to chase down.
South West England is UKIP ground zero
Out of the 45 target seats, 22 are in South West England. Combined with UKIP obtaining their highest election vote shares in SW England, this region is the area where UKIP will most likely win a Westminster seat.
Limitations of modelling
Before we depart, we shall reflect on the limitations of this modelling of the possible levels of UKIP support.
Firstly, this modelling relies only on previous party share of the vote in 2010. It does not place any weight on demographics. Current polls have UKIP polling highest amongst voters over 60 years of age and doing better amongst working class voters. The former is consistent with Conservative demographic strengths, but not the latter. A uniform distribution of a share of the Conservative vote across the country may not take into account the relative UKIP strength amongst working class voters.
Secondly, the modelling is really just a complex version of the uniform swing. As there is no such thing as a uniform swing the above modelling is only a guide, not a prediction. Out of the group of 45 target seats listed above, the UKIP could pick a number of these seats up, if they receive a spike in their vote for whatever reason, (good candidate, bad sitting MP, important local issue, etc, etc). The reason to list the 45 target seats is to identify which seats the UKIP can snatch if circumstances go their way.
Thirdly, if new 600 constituency boundaries are implemented, we will need to review the above modeling for the new boundaries.
We are now in uncharted territory with a new party, UKIP earning a significant share of support. As UKIP does not have an intense geographical concentration of its vote like the SNP, Plaid or the Greens it cannot reasonably have a chance of winning seats until its national vote breaks, at least 15%. With the above modelling we have attempted to map out the most probable seats where UKIP can break into Westminster.