Dr Kenneth Bunker, from the LSE, looks at recent polls in Northern Ireland, Scotland and Wales, and assesses their significance for the elections in May.
The 2016 election cycle is in full force. On 5 May, Northern Ireland, Scotland, and Wales will elect the members of their devolved governments. Pollsters are already looking at citizen preferences, and speculating on potential results. While few polls have emerged in Northern Ireland, enough have been fielded in Scotland and Wales to make some early predictions. In this article I review some of the major polls and their margin of application for the three countries. I also refer to implicit trends in the polls and lay some groundwork to interpret their results.
As mentioned above, Northern Ireland is bound to hold the election will the fewest pre-electoral polls in the cycle. Only a few pollsters have attempted to field polls. Indeed, in 2016, only one national pollster has taken a crack at it. This pollster fielded one poll in February and one poll in March, reporting that the Democratic Unionist Party (DUP) and Sinn Féin (SF) were technically tied in the front-running position. At the time of the polls, the DUP was estimated to obtain 27% of the vote, and SF was estimated to obtain 26% of the vote. The Ulster Unionist Party (UUP) and the Social Democratic and Labour Party (SDLP) were the other two parties with a significant share of preferences, obtaining 15% and 12% respectively.
Scotland and Wales, in contrast, have seen a significantly higher number of polls fielded. One explanation is because the elections are more attractive for national voters and international observers, or simply because they are more competitive. In Scotland for example, many eyes are on the nationalist party SNP which is expected to make an important advance, while in Wales the newfound rivalry between the Conservative Party and Plaid Cymru, seems to be awakening the interest of the press.
A second explanation is because the nature of the election is different. Polls in Northern Ireland only seek to measure the national vote intention since all members are elected in the same election, while polls in Scotland and Wales seek to measure vote intention at two different tiers since member are elected at both the constituency-level (candidate tickets) as well as at the regional-level (party tickets). This means that, in Scotland and Wales, polls have a wider range of application.
In Scotland, a stunning number of polls have been fielded since 2012, aiming to measure both constituency-level and regional-level preferences (over 70). Already in 2016, 10 polls have been fielded, with an average number of 1,030 interviewees. In all of these polls the Scottish National Party (SNP) occupies the pole position, with an average lead of 30% over the second favorite, in all cases the Labour Party (Lab) or the Conservative Party (Con). At the constituency-level the SNP obtains an average of 53.2% of the vote, while at the regional-level it obtains an average of 46.1% of the vote. The Labour Party moves between 20% and 19% respectively, while the Conservative Party consistently obtains 16.5% in both types of election.
In Wales, pollsters have been more active than in Northern Ireland, but not as much as in Scotland. Since the previous legislative election (2011), 17 polls have attempted to measure vote intention, at both the constituency-level as well as the regional-level. All of the polls show a clear lead for Labour over the runner-up, in most cases the Conservative Party, but occasionally Plaid Cymru (PC). In 2016, there have only been two polls so far. In both, Labour leads with a 9% distance over the Conservatives. Between both polls there is a minor variation for PC, moving from 19% to 22%. The UK Independence Party (UKIP), obtains an average of 16% of the vote, while the Liberal Democrats (LD) and the Greens (Gre) each orbit around 4%.
Before interpreting what these polls mean for each party in each election, it is necessary to make an important caveat. While the weight of the historical trend of polls is not to be taken lightly, it is always possible for numbers to suddenly change direction. Events that take place during the campaign trail can change the outcome of a race, especially when very close to the election. It is also important to consider that polls are snapshots of the moment, and as such should only be taken to interpret the window of time in which they were fielded. However, when the number of polls indicating the same results increases over time, it is more likely that they are approaching the true value of public opinion.
One more caveat should be made. Polls can only be used as reference margins for results, since they do not necessarily point at the winner. The party with highest preferences does not necessarily have to be the winner. It is possible that a party obtain more than 50% of the vote, but less than 50% of the seats. This will depend on the distribution of preferences across constituencies, and in the case of Scotland and Wales, also on the distribution of votes at the regional-level. However, it is highly unlikely that a party with a significantly robust lead, say around 30% in the polls, will lose the election. Thus, any interpretation of poll trends should be taken as a rough proxy, rather than as a confident forecast.
Taking into account the above, in the case of polls described, some of elections results can be anticipated, while others are still unclear. For example, we can hardly be confident about the result in Northern Ireland, given the restricted number of polls. In the two polls fielded in the past year the two leading parties (the DUP and SF) are tied at around 23%, while the two second largest parties (UUP and SLDP) are tied at around 13%. These numbers seem fairly consistent with the results of the previous election (2011), and as such lend these polls credibility. But this credibility only makes the outcome of the election more uncertain, given that the polls show the two pairs of relevant parties tied in close intervals.
Scotland, on the other hand, is more clear cut. The number of polls consistently pointing at the same results (low standard deviations) clearly indicate a victory for the SNP. The numbers are also consistent with previous electoral results, considering the win of the SNP in the 2011 Scottish parliamentary election, and their landslide victory in Scottish constituencies in the 2015 Westminster election. Everything seems to indicate that they will by reducing the margin of support for Labour and the Conservatives. Furthermore, even though the sum of votes for Labour and the Conservatives add up to around 40%, it is likely that they will both lose in a majority of constituencies where they the SNP fields candidates. While the SNP victory is likely, the extent of the win is still unclear.
In Wales, if nothing ground-breaking occurs, Labour will win the election. They have won every election since the devolution of government in 1999, and will probably repeat their victory this year. The more interesting question is whether the Conservatives will come in second (as they did in 2011), or if it will be Plaid Cymru (as they did in 1999, 2003, and 2007). The numbers favour the former, who obtained an average of 24% of the vote in the immediately previous election. In the same election, Plaid Cymru took 18% of the vote, and surrendered its historical place as the second largest party in the National Assembly. As in the case of Northern Ireland, polls in Wales seem to be confirmed by the results of the previous election.
Summing up, in Northern Ireland nothing is clear. Both the DUP and SF have a chance at leading the legislature. Likewise, the third place will be disputed between the UUP and SLDP. In Scotland, the SNP will win the election and lead the Scottish Parliament for the third time in a row. The important question is whether they will be able to repeat their 2015 Westminster win, and significantly increase their representation throughout the region. Similarly, in Wales, the important question in not as much about the winner, but about the second place. Here, Labour will likely win for the fifth consecutive time, while the second place will be intensely disputed between the Conservatives and Plaid Cymru.
This post represents the views of the author, who writes in a personal capacity, and not those of Democratic Dashboard or the LSE.