Tuesday, 27 April 2021

What do the polls for Senedd Elections 2021 tell us ?


Can they accurately point to the final outcome?

Wales is more or less halfway through the Senedd election campaign. Hitherto it is proving to be a rather subdued campaign as well – although I doubt whether the candidates in the constituencies are of that opinion. Nevertheless, the noise and bustle seem to be missing.

Perhaps that was always going to be inevitable during the pandemic period.

Be that as it may, the pollsters on the other hand have been busy. Three polls in three weeks or so all giving contrasting sets of data.

The first poll was in the third week of March when the usual You Gov/ITV Wales poll showed that Labour could be heading for trouble. They were on 32%, the Conservatives on 30%, Plaid Cymru on 23% and the Welsh Liberal Democrats on 6%. The reasons for Labour's decline is for another time.

What was intriguing about this poll was that First Minister Mark Drakeford was seen as doing a good job by 57% of those interviewed and only 34% with a negative opinion. Over 70% of Labour and Plaid Cymru voters gave Drakeford the thumbs up. Plaid Cymru came in behind the Conservatives on the popularity contest for First Minister. Adam Price and his party just do not seem able to make a breakthrough according to the polls. The issue of independence for Wales is a similarly intriguing set of polls but again, that is for another day.

But despite that, the poll highlighted that Labour could lose seven seats – 5 to the Conservatives and 3 to Plaid Cymru. The Conservative gains would be in Cardiff North, Gower, Vale of Glamorgan, Vale of Clwyd and Wrexham.  Plaid Cymru projected to gain Blaenau Gwent, Cardiff West and Llanelli. Lee Waters has been very active on social media and he like Helen Mary Jones has also come in for criticism recently. How much that will impact the Llanelli result is speculative.

This synopsis would give a Senedd seat distribution of Labour on 22 seats, the Conservatives on 19, Plaid Cymru on 14, Welsh Liberal Democrats on 1 and the remaining 4 seats probably to the Abolish Welsh Assembly party.

By 20th April a new poll emerged, this time from Opinium for Sky News and this one gave a completely different result. This time Labour was forecasted to be within a couple of seats of returning to Government. Labour would win 29 seats, the Conservatives on 19, Plaid Cymru on 10, Liberal Democrats with 1 and Abolish the Assembly with 1. This one showed that Plaid had taken the main hit, down by 4.

Soon after there followed the second YouGov/ITV Wales poll confirming the Opinium poll and a Labour recovery. This time the poll result forecasted Labour with 26 seats, the Conservatives on 14 but Plaid Cymru with 17 replacing the Conservatives as the main opposition party in the next Senedd. Abolish the Welsh Assembly party remained on two seats and Welsh Lib Dems with their solitary one.

So what is going on with these polls?

The first thing to say is that every poll result has a margin of error in its findings of around +/- 4%. So for example, if a poll finds that Labour is on say 29% that means it could actually be in a range of support between 25% to  33%.

Then of course whilst pollsters do their utmost to get their say 1,000 interviewed sample as near as possible to cohere with the actual composition of the population of Wales the sample may not actually be truly accurate.

These two factors really make making seat projection from polls somewhat hazardous, particularly when you consider marginal seats and of course the 20 regional list seats. So possibly the pollsters should be reporting on the range of seats for each party in a given poll rather than being so specific in their findings.

There are aspects that I will return to before polling day such as likely voting turnout; that in this election because of Covid the figure using postal ballots will be probably much higher; then there is the matter of 16-17 year-olds voting for the first time and in what numbers will they turn out to vote; and finally the extent and impact of tactical voting in a number of key marginal seats.

So where are we at present?

First of all polling headlines could be misleading. But it does look that at present as if Labour will be just short of overall control. That is not unusual, because in four of the last five Assembly/Senedd terms it has had to rely on either formal coalitions or the support of non-Labour members to govern.

But there are still two weeks of campaigning left and as we have discovered in the last 36 hours unforeseen events can and will happen. Just Ask Boris Johnson!