Not Another Revised Welsh Forecast!

I'm not going to do a full revision of the Welsh Forecast until just before Election, by which time I'm hoping another established polling company other than YouGov will have conducted a survey. That's not to say that I think YouGov are wrong but you can see in the Scottish polling that slight variations can make a big difference to the final outcome AMS elections, so it would be nice to have two companies to compare. In Scotland the key issue at the moment is how close the Scottish Nationalists can get to the magical majority figure of 65. In Wales it is how large Labour's majority will be. A few points either way could give Labour a majority of 2 or 10 seats, which would clearly make a big difference to what they could achieve in the five year term.

So, if this isn't a revised Welsh Forecast then what is it?! Well, it's more of a discussion on how last week's YouGov poll would affect the Revised Forecast I did 10 days ago. We have also received some comments, both public and private, on the state of certain constituency races which have influenced me a bit, especially if a few sources from different parties are saying the same thing!

In the revised forecast I didn't leave that many seats Too Close. With 3 ½ weeks to go I considered only five seats to have no clear favourite so it's worth having another look at those.

Constituency	Initial	Revised
Llanelli	PC Lean	Too Close	PC/	LAB
Montgomeryshire	Too Close	Too Close	LD/	CON
Aberconwy	Too Close	Too Close	PC/	CON/	LAB
Cardiff Central	Too Close	Too Close	LD/	LAB
Cardiff North	Too Close	Too Close	CON/	LAB

As the only changes in the constituency voting intention from the previous YouGov poll were Labour up two points and the Tories down one it hasn't conclusively switch any of the above seats, although clearly Labour's chances of taking their target seats seem higher than they did at the start of the campaign. Llanelli is the seat where Labour's march has been so keenly felt. To a point Plaid Cymru are camped in enemy territory here as the Westminster seat is solidly Labour, and has been since the 1920s. However, Labour were not polling well enough pre-March to take this seat on UNS projections so with their incumbent AM Plaid appeared favourites in this seat.

Since then the two party's fortunes have been very different as a 4% PC>LAB swing has been observed in the polls. Both parties appeal to left leaning voters in the South and so Helen Mary Jones is unlikely to be pleased with this development. Uniform swing sees her losing her seat by quite some margin and although her personal vote should mitigate some of her party's national drop in support it is worth remembering she lost a close battle for Llanelli in 2003. Jones is still the bookies favourites but this is one seat where the apparent PC>LAB swing could cost the Welsh Nationalists.

As the Liberal Democrats and the Conservatives were virtually unchanged in the new poll Montgomeryshire is still a bit of mystery. I have been told the Tories appear to have a better presence in the constituency and they are very slight favourites with the bookmakers. I have a sneaking suspicion this will either be a tight hold for the Lib Dems or they'll lose it badly on a large swing reminiscent to the one which ousted Lembit Opik from Parliament…so it's bound to be a narrow Conservative gain!

Aberconwy is very interesting seat for anyone considering placing any constituency bets on this election. Despite losing their incumbent AM Plaid Cymru are considering favourites in this three-way marginal. However, I have heard from Plaid supports that they are far from confident of holding this seat. In 2007 they won this seat with 38.5% of the vote but as they are 5 points down in the polls from that election and no longer have the Gareth Jones' personal vote a fairly significant drop in support seems certain.

So who out of Labour and the Conservatives are best placed to benefit? On Uniform swing Labour currently sweep to victory here and the bookies rate their chances much higher than the Tories. However, they had to change their candidate late in the election cycle which would have surely disrupted their campaign somewhat. You can get the Tories as long as 10/1 at the moment, which considering they were second in 2007 and have a prominent local candidate looks like value from a far. The caveat being they need to increase their vote to win which isn't apparent in the national polling. It's quite easy to find reasons why any of the three contenders won't win, which from makes working out who will somewhat difficult!

Finally on to Cardiff and Labour's chances in the two key marginals in the capital have surely increased since the campaign has started. They take Cardiff North quite comfortably on uniform swing and as their candidate is a former MP for the city the incumbency advantage usually enjoyed by the defending party is likely to be significantly weaker here. As for Cardiff Central, this would still stay with the Liberal Democrats on Uniform swing despite Labour polling almost 50% nationally. The Lib Dems have lost their incumbent but they do have a good local infrastructure to aid their defence. For Labour to win this they require a huge swing but given the current polling that is looking quite possible.

There are a couple of seats that I switched to Lean in the revision which are worth a quick look at.

Constituency	Initial	Revised
Carmarthen W & P'shire S	Too Close	LAB Lean	LAB/	PC
Preseli Pembrokeshire	Too Close	LAB Lean	CON/	LAB

Having dismissed Plaid's chances in Carmarthen West & Pembrokeshire South following their national drop in support I received a few comments from the ground there. Essentially they are all saying the Conservatives are not really in the running and the Plaid candidate, the popular regional AM Nerys Evans, seems the most likely to prevent a Labour gain. This seat is pretty much the opposite of the other three-way marginal Aberconwy in that each contender has plenty of positive reasons why they might win. Given Labour's poll lead and having the former AM for this constituency as their candidate I still think they are the likely victors but it sounds like Evans' personal support is keeping her chances alive.

I've heard suggestions that Labour are concentrating on the Carmarthen West seat to the detriment of their campaign in the next door target of Presili Pembrokeshire but as they easily win this on UNS it would take quite a poor local campaign to lose this one.

A few of Plaid held seats are seemingly under threat from Labour on strict Uniform Swing calculations but I'm still fairly sure they will stay with the Welsh Nationalists. Although they lose Ynys Mon on UNS they fact their leader in the Assembly is the sitting AM should help. Equally, it would be a pretty terrible night for Plaid if they lost Arfon or Carmarthen East & Dinefwr and I just don't see that happening.

As for the list seats, I'm not going to go into them in great detail as there is little significant change in the regional voting intention. The big change from the revised forecast is that UKIP have fallen back from their strong polling in March and would not win a seat in South East Wales on Uniform Swing. The cross-tabs indicate they are doing better in the region than my Welsh seat calculator gives them credit for but regional sub-sets are extremely unreliable.

So, where does all this leave the forecast? Here's the revised forecast from last Monday:

	2007	Initial	Revised
Labour	26	28-33	32-34
Plaid Cymru	15	12-15	10-11
Conservatives	12	10-13	11-14
Lib. Dems.	6	4-5	2-3
Independent	1	0	0
UKIP	0	0	1

Running through it all I don't think this poll has really changed much. The key point is that with only two weeks to go things are becoming more set and wild swings for or against any party are unlikely. Of course, the polling may have been over or under stating certain parties all along but significant changes in national vote share seem unlikely at this late stage.

On the evidence available I think Labour are seat for a majority of around 6 seats, so a best case scenario from the Initial Forecast and right in the middle of the Revised one. Plaid Cymru seem set to fall to around 10 seats as a reward for their four years as a junior coalition partner. As the Tories are holding the regional vote share around the same level as in 2007 then if the polling is correct they are set to increase their seat numbers. Essentially, they are going to benefit from the drop in support from the Liberal Democrats and Plaid Cymru on the regional list, which should more than make up for the expected constituency losses to Labour. The seat that did go to UKIP in the Revised forecast would now head to the Liberal Democrats so they are currently looking at 3-4 seats.

So there you have it. With less than two weeks to go until polling day I don't ask for much; just a Welsh poll from another established pollster and more of your comments on the ground!