Official Report (Hansard)
Date: 08 May 2012
PDF version of this report (173.53 kb)
Assembly and Executive Review Committee
Northern Ireland Act 1998: Review of Parts III and IV
The Deputy Chairperson: We will move on to tab 6, which is the research paper entitled 'Electoral Constituencies'.
Mr Tim Moore (Research and Information Service): Thank you. This research paper addresses three issues: decoupling the Assembly constituencies from the Westminster election constituencies, the possibility of using the local council boundaries as the basis for Assembly constituencies, and the question of what you would do if you were to start from scratch. To be honest, the final issue is the one we found most difficult to answer, and we have not really found an answer to it.
I will take members through the paper. It is probably helpful to set out the legislative framework before the recent legislation, the Parliamentary Voting System and Constituencies Act 2011, came into force. The Parliamentary Constituencies Act 1986 set out the rules for the redistribution of seats across the United Kingdom. As enacted, there were certain constraints placed on the number of seats that there could be in any one part of the United Kingdom. In Great Britain, there should not be substantially greater or fewer than 613 seats; the number of constituencies in Scotland should not be fewer than 71; in Wales, that number should not be fewer than 35; and the number of constituencies in Northern Ireland should not be greater than 18 or fewer than 16, and should be 17. Of course, we have 18 seats currently. I think that that happened back in 1995, when West Tyrone became the new constituency. Perhaps a member of the Committee knows more about the reasoning behind that than I do.
So, that was the position as enacted. The Scotland Act 1998 removed the number of seats in Scotland in anticipation of a reduction in the number of Scottish MPs and changed the rules about the number of seats and how you would allocate them. The fifth boundary review reduced the number of seats in Scotland by 13, taking it from 72 to 59. The important point to make about that legislation is that it specified minimum numbers, but that there was scope for some variation above those numbers, or, in the case of Northern Ireland, between a lower and higher number. There was scope for the Boundary Commission to come up with numbers that it thought were most appropriate.
However, in 2011, the Parliamentary Voting System and Constituencies Act came into force, which totally changed the way in which seats would be allocated to the different parts of the United Kingdom. The Act set the number of seats for the whole of the UK at 600. Four of those were historical seats, which left 596 to be distributed among the different parts of the UK. That legislation introduced a big change: it stated that those 596 seats would be distributed using a mathematical formula, which removed a lot of scope for the Boundary Commission to make arguments that it thought were more reasonable. One would apply what is known as — and I am sure that my pronunciation will be wrong — the Sainte-Laguë formula. That sets out how 596 seats would be allocated to different parts of the United Kingdom.
Table 1 in the research paper shows the impact of applying the formula to the population as at 2010. The legislation specified that you would take a period of two years and 10 months before the Boundary Commission was due to report. We know that the Boundary Commission is to report in October 2013. Therefore, going back to December 2010 and based on the electorate at that time, the Sainte-Laguë formula gives you the number of seats. This is how we have come to the new allocation of seats, which gives Northern Ireland 16 constituencies. Sainte-Laguë is a bit like d'Hondt. There are iterations. The first seat goes to England, and you then work through the rest. For Northern Ireland to get its sixteenth seat, the number was in the low 560s. It would be close to the end. It may not have got the sixteenth seat. Perhaps it would be close to 15 when you go through the iterations.
As the paper states, we are allocating seats on the basis of a mathematical formula, with very little scope for the Boundary Commission to come up with its own numbers. Therefore, we can project, based on numbers, what would happen in a number of scenarios. The statistics team in RaISe tried to do that, and we have set out the results of our projections in the paper.
The first point to note is that the numbers you put into the formula depend on the registered electorate. I know that Ray briefed the Committee and that Members will be aware that in Northern Ireland there is individual electoral registration, whereas, in GB, it remains household registration. When Northern Ireland moved to individual electoral registration, the numbers on the register dropped, but they came back up again. We have tried to reflect in the models what would happen in England when it moves to individual electoral registration. There are three scenarios, and, in a complicated way, the way the methodology lists the scenarios is the reverse of how they are shown in the tables. I will talk members through each of the three scenarios.
The first thing we tried to do was to take the average registration over the past five years and apply it to the projected populations to determine what the numbers and outcome would be in respect of seats? Therefore, we took the five years, fixed the proportion of people who are on the register and applied that. That is scenario three. We found that, over time, you would see Northern Ireland losing a seat, and that the number would continue at 15. That is scenario three. That is working on the basis of the levels of registration as they are at the moment. Then we worked out what would happen if individual registration in England had a big impact and it lost 10% of the electorate, as we did when we first went to individual electoral registration. That is scenario two. Over time, the electorate in GB would reduce, so the seats in Northern Ireland would increase to 17, and, eventually, would come down to 16. That really reflects the change that you would get if individual electoral registration in England, Wales and Scotland had the same impact as it did here.
Finally, in the third scenario, we worked out what would happen if something between those two scenarios happened; for example, if there were a 5% drop in England. In theory, we would then drop down to 15 in 2015, but there probably would not be a boundary review report anyway at that date.
The Deputy Chairperson: Tim, I am sorry, we are going to have to suspend the session.
Committee suspended for a Division in the House.
On resuming —
The Deputy Chairperson: Tim, we stopped you in full flow. You can carry on.
Mr McCartney: You can start again from the beginning. [Laughter.]
Mr Moore: I will try to pick up where I left off. I was briefing the Committee on the research paper, which looks at the impact of the Parliamentary Voting System and Constituencies Act 2011. It contains scenarios that we worked out by putting some projections into the formula that is in the Act. I will recap very briefly. Projecting forward on the basis of registration, as it is, we see the number of Westminster seats in Northern Ireland reducing to 15 and staying at 15. We then modelled figures on the basis that individual registration in England would reduce registration levels to those comparable with here when it was first introduced here. That sees the number of seats here rise to 17 and eventually reduce to 16. Scenario 1, which is method C, shows that if something between those two scenarios happens, the number of seats would stay at 16. Perhaps that is the best way to say it. In theory, the number would go down to 15, but there would be no Boundary Commission review in 2015, so, by the time of the next one, it would be at 16. Those figures, and our projections, are not what we are saying will happen; we are simply saying that this is what could happen under certain circumstances. We are showing that there is the possibility of variation in the number of seats. However, we are not saying that this will or will not happen.
The paper goes on to look at the process of decoupling. As members know, Scotland and Wales have decoupled their Parliament and Assembly seats from their Westminster seats. That was particularly important in Scotland, where, as I have said, the number of seats was reduced by 13 in the most recent Boundary Commission review. The number will be reduced again under the recent proposals.
The Scottish Parliament (Constituencies) Act 2004 decoupled the seats. In place of the link to the defined parliamentary constituencies, the new legislation stated that the number of constituencies would be 73. Of those constituencies, two were fixed and the rest were defined by using the existing Westminster boundaries. They took the boundaries as they were, and said, in the legislation, that those would be the Scottish parliamentary boundaries.
In addition in Scotland, the Boundary Commission was given the power to review the Scottish parliamentary boundaries but not the number of seats. The commission could not move the number of seats up or down, but it could shift the boundaries around if population shifts occurred. Therefore, the Scottish Parliament knows the number of seats and what they will be, but they do not know not exactly where the boundaries will be. The rules that the Boundary Commission must apply when moving the boundaries are set out in the paper. I will not go into those rules.
The Boundary Commission in Scotland has made its first report on the parliamentary boundaries there. The report goes to the Secretary of State and gets laid before the Parliament. However, the Parliament has no power to amend the recommendations, nor, in fact, does the Secretary of State, who must take the recommendations as they are. The order that brings the boundaries into force is made through Westminster. The point to be made about that is that although you may decouple from the Westminster constituencies, this is still not a devolved matter. It is not within the Scottish Parliament's gift to determine the number of its seats.
As the paper points out, a broadly similar process took place in Wales, which has now decoupled. That happened under the 2011 Act. They also locked down the existing constituencies as their parliamentary seats. That may be an issue for the Committee to think about. When Scotland and Wales decoupled, they went with the existing Westminster seats. Part of the Committee's deliberations, I understand, has been about Northern Ireland going down to 16 seats, but the Boundary Commission is not due to report until October 2013, and those 16 seats do not exist. I am not sure whether there is a technical issue around that, but it may be something that the Clerk and the Committee may like to consider. If that were to become the preferred option, how would you go about drafting the legislation?
Another issue that we were asked to look at was what would happen if we looked at the new local council areas as the constituencies and allocated Assembly Members across those. We have tried to do that. The Local Government (Boundaries) Order (Northern Ireland) 2012 was laid last month. We conducted a simple exercise, which was to use data from 2009 to look at the historical electorate, work out the percentage of the overall electorate for each of the new areas and allocate the number of Members based on that percentage. It is a straightforward process, and, in the paper, you can see how the numbers divide up. You can see that, because the size of the electorate differs significantly in each areas, so would the number of Members in each district, if you were to say that each district would have Members allocated in relation to the size of the electorate. In a 108-Member Assembly, the district with the highest number of Members would be Belfast, with 18, and the lowest, in an 80-Member Assembly, would be the district of Fermanagh and Omagh with five Members. That is not to say what should happen but to show what could happen almost as a starting point. It shows that how you then subdivide some electoral areas is an issue, because Belfast appears to be extremely large.
The next part of the research paper looks at district magnitude, which is the issue of the number of seats in each constituency. We were asked to look at the situation with the Dáil, where different numbers of Members are drawn from different constituencies. The point being made here is that, although six Members of the Assembly are taken from each constituency currently, it does not have to be like that. You could take a different number from constituencies, and that is exactly what happens in the Dáil. The research paper contains information on the existing situation there and on the proposed changes there.
We have done very little on the final part we were asked to look at, because we could not find material on it. It was to ask, if you were to start with a blank sheet, what constituencies would you arrive at? There is no formula for doing that. We have included some comment, but the only piece of agreement that we found is in the last sentence of page 13, which states:
"Most scholars agree that district magnitudes of between three and seven seats tend to work quite well, and it has been suggested that odd numbers work better than even numbers, particularly in a two-party system."
We do not have a two-party system, but that was a far as we could go to consider whether there was some machine into which you could put the numbers if you wanted a perfect system. It does not look like there is. I can now take questions from members.
The Deputy Chairperson: Thanks very much for that. Have members any questions?
Mr McDevitt: I am sorry, Chair, I am being called back to the Chamber.
The Deputy Chairperson: There are no questions. Thank you very much, Tim.