Parliamentary Voting System and Constituencies Bill - Political and Constitutional Reform Committee Contents


Written evidence submitted by David A G Nowell (PVSCB 23)

— A modest reduction in seats is welcome in principle. — The reliability of electoral registers.

    — The overall basis on which seats will be allocated.

    — The need not to straddle European electoral regions.

    — Problems inherent with constituencies with very large areas.

    — Ignoring significant geographical and historical ties.

    — Permutations and combinations in relation to boundary reviews.

    — The potential need to divide polling districts.

    — The nature of Alternative Vote election results.

A MODEST REDUCTION IN SEATS IS WELCOME IN PRINCIPLE

  1.  Ideally the size of the House of Commons should be reduced to 500 seats, in tandem with the introduction of elected regional assemblies for the rest of England outside of Greater London. This proposal was set out alongside an analysis of the different voting systems and the effect of single and multi-member constituencies on electoral geography in my essay, "There is more to representative democracy than simple proportionality" (pages 62-89) in Democracy the missing element, a compilation of essays by distinguished writers making the case for the multi-member constituency, introduction by Charles Kennedy MP, published in 2008 by Democrat Action Group for Gaining Electoral Reform. Without this reform the United Kingdom remains exceptional in not having universal regional government. So while the proposed reduction to 600 seats would not be met with the introduction of English regional government, this would be a relatively modest step, which in itself I do not see is worth opposing.

THE RELIABILITY OF ELECTORAL REGISTERS

  2.  The proposal to use the electoral register published on 1 December 2010 as the basis for the allocation of seats has not been thought through. To begin with, no publicity campaign has been considered to remind people to register to vote: this is a legal obligation, and under-registration skews the numbers used for allocating constituencies. However, even estimates of the unregistered population can be flawed. For example, in the last boundary review for wards in the Borough of Barnet, council officers wrongly assumed roads with more than 3% of unregistered households should have their electorates adjusted upwards to reflect this, so such neighbourhoods were not disenfranchised. But they forgot that certain areas like Finchley contain large numbers of Japanese citizens who cannot register to vote and took no account of this.

3.  The reliability of electoral registers is further undermined by lax checks, and many EU citizens either don't bother to register or don't tick the right box and end up with full voting rights. Things are further complicated in areas with high population turnover, as the previous registration is not always cancelled even if somebody else moves into the property. As a local political activist, I happened to spot friends who had remained on the register for three years after they moved out, until I alerted them. While it is assumed student populations tend not to register to vote, they are entitled to register both at their parents' address and at university. I would suggest double registration is more likely among more affluent students, as their university halls of residence and colleges tend to register them automatically while they maintain their out of term registration. Furthermore, second home owners and MPs routinely register at both their main residences. Since there are roughly a quarter of a million second homes in the UK, this could be a significant but unknown number, as electoral registration forms don't ask for these details. Furthermore, this undermines the policing of general elections to ensure people only vote once, as no system is in place (even with postal votes sent to an address in another local authority), to undertake spot checks using the marked register and prosecute any offenders.

THE OVERALL BASIS ON WHICH SEATS WILL BE ALLOCATED

  4.  While I will go into greater detail in the subsequent paragraphs, the current proposals ignore the need to maintain certain geographical and historic ties, and fail to understand the underlying nature of providing balanced electoral geography inherent in any single seat electoral system. Most fundamentally, as the numbers of seats in a boundary review increases, the number of potential permutations and combinations rises exponentially. In particular, this would make it almost impossible for the boundary commission in an undivided England to adjudicate between rival submissions, which would straddle completely different sets of local authority boundaries. Furthermore, the Isle of Wight should be a named exception, and the residents of Cornwall would probably even prefer to be under represented with only five parliamentary constituencies than let the area around Bude share its representation with Westward Ho!. While provision is made for constituencies larger than 12,000 square kilometres, this is far too tight a ceiling when dealing with the extremely lightly populated Scottish Highlands and Islands beyond the two already named exceptions.

5.  As drafted, the Bill only makes special provision for Northern Ireland. Under the proposed system, it is likely (especially in Wales) that it could be difficult to satisfy the exacting requirements for all these seats to be within  5% of the UK target of around 76,200 electors. These limits are likely to be approximately 80,000 and 72,400 electors, using the last general election as a guide. Given the relatively small allocation, the average electorate for each seat in Wales could fall significantly out of line with the UK, as the number of seats within the principality has to be a whole number. Were this to happen without the current Bill being amended, the ensuing boundary review could be increasingly boxed in. The greater this discrepancy, the greater the likelihood that more seats would appear to be geographically contrived simply to meet this target. In any case, as I will explain, given the need beyond named exceptions for at least one extra seat for the Scottish Highlands and Islands, it would make sense for a handful of additional seats to be distributed to Northern Ireland (2), Wales (2) and Scotland (1) before the formula set out in the bill was used to allocate the remaining unnamed constituencies. This along with other measures I will suggest should dispense with the need for special arrangements for allocating seats in Northern Ireland.

THE NEED FOR SEATS NOT TO STRADDLE EUROPEAN ELECTORAL REGIONS

  6.  Within England it would be sensible for constituencies to be allocated wholly within European electoral regions, in case elections to the European Parliament coincided with a parliamentary by-election and to allow reliable comparison between these elections and general elections. Trying to run two separate elections on the same day with two different franchises, in a parliamentary seat straddling two regions, would be a recipe for confusion both for political parties and election officials. Furthermore, any MP representing such a constituency would have to liaise with two set of MEPs. In any case, the numbers of seats within each separate boundary review has to be constrained to reduce the potential number of permutations and combinations and allow for the boundary commission to undertake a fair and impartial process, and so it would be extremely illogical to allow parliamentary seats to straddle these regional boundaries.

PROBLEMS INHERENT WITH CONSTITUENCIES WITH VERY LARGE AREAS

  7.  As drafted, the Bill before parliament only allows for exceptions from the  5% UK wide target if a seat exceeds 12,000 square kilometres in area. Clearly this rule has the current Ross, Skye and Lochaber constituency in mind at 12,779 sq km with 51,836 electors, but unfortunately ignores the requirements for any seat starting in Caithness. In order for this seat to be allowed to deviate from the target electorate it would have to extend southwards to include the Black Isle and a large portion of the Highland Council Wester Ross, Strathpeffer and Lochalsh ward (4,948 sq km). This would further reduce the electorate of any seat starting with Skye, which would then have to extend eastwards into the Badenoch and Strathspey ward (2,330 sq km) to the south of Inverness in order to reach the 12,000 sq km threshold, as enlarging it southwards beyond Oban and Mull would be even more problematic.

8.  In any case, Argyll and Bute (67,165 electors) which includes Oban, like Orkney and Shetland (33,085 electors) along with Na h-Eileanan an Iar (Western Isles—22,266 electors), should be a named exception, as at 6,909 sq km it is relatively small. The current constituency (which has already been enlarged since 2005 to include the whole of the Argyll and Bute Council), extends for over 160 km encompassing at least nine inhabited islands linked by ferries, along with the Mull of Kintyre all the way round to Helensburgh. Adding Arran would make little difference as this island only has a population of around 5,000 and would not bring it up to the likely 95% threshold required for its electorate. Thus exceptions should be made for any seat exceeding 8,000 sq km given that the current Caithness, Sutherland and Easter Ross constituency is 8,752 sq km with 47,257 electors.

IGNORING SIGNIFICANT GEOGRAPHICAL AND HISTORICAL TIES

  9.  The Isle of Wight, with 109,966 electors, has a very distinct local identity so should remain undivided. Once the Boundary Commission has in consultation with the islanders decided whether to allocate one (or potentially two seats if the population were to continue to increase, as its electorate was only 90,961 in 1979), it should be excluded from the calculation of the UK average electorate, along with other named exceptions. Insisting on the island sharing a seat with the mainland would most likely result in the creation of an Eastern Isle of Wight and Southern Portsmouth constituency or Spithead for short, given that this end of the island is quite urban.

10.  Cornwall with 418,741 electors means that, given an UK average of 76,186 using the last general election as a guide, this culturally unique county with its own language would get 5.5 seats. Arrangements both with five and six seats would, with average electorates of 83,748 (109.9 %) and 69,790 (91.6 %) respectively, fall well outside the likely limits of around 80,000 and 72,400 electors. On this basis Cornwall should be a named exception with a fixed number of seats and thus removed from the calculation of the average electorate for the UK as set out in the Bill.

  11.  Beyond these two most notable cases, other slightly less significant geographical and historical ties can be coped with in the way in which European electoral regions should be subdivided. This is necessary in order to further reduce the number of realistic permutations and combinations of different constituencies to a reasonable level within each review area, and thus allow the national boundary commissions to judge fairly between rival submissions before making their final recommendations.

  12.  In any case, due to historical considerations, many countries do not allow their equivalents of parliamentary constituencies to straddle certain boundaries, such as De«partments in France and La­nder in Germany. Most notably, the constitution of the United States of America allocates congressional districts to individual states, on the basis of the national census. This means that there are great differences between smaller states in the numbers of people in each congressional district, depending on whether these states have reached or just fallen short of the number needed that time round to be allocated a second, third or fourth seat in the House of Representatives.

PERMUTATIONS AND COMBINATIONS IN RELATION TO BOUNDARY REVIEWS

  13.  As the number of constituencies rises within a given area the number of possible permutations and combinations increases exponentially, including almost endless possibilities for straddling local authority boundaries in so many different ways that the whole process becomes impossible to manage. For example, in the last parliamentary boundary review for England, which came into effect at the 2010 general election, Greater London was divided in two, given that the River Thames acts as a natural barrier within this region. However within both halves the Commission quickly established that a number of boroughs had electorates near enough to whole multiples of the target electorate for new seats to be formed from combinations of wards exclusively within such boroughs. Once these had been eliminated, this made concentrating on submissions dealing with new constituencies that would have to straddle other borough's boundaries much more straightforward than might have been the case. If the requirement for every constituency to be within very tight limits were to override all other considerations, then boroughs with whole numbers of seats within these limits could nevertheless have to be straddled with neighbouring boroughs in order for all the new constituencies to meet these rules.

14.  This can be avoided to a certain extent if the European electoral regions are subdivided before the start of submissions to the boundary commission. Once they are allocated a given number of seats using the formula set out in the bill, or a slightly amended regional method with the proviso mentioned below, the average electorate within each sub-regional division should be used to set its own  5% limits for electorates within that area. This, combined with slightly greater than currently proposed  8% (about 6,080 electors) national limits, would rule out more extreme variations between constituencies representing around 98.6% of the United Kingdom's total electorate who are not covered by my suggested exceptions. This would reduce the likelihood of potential seats being too tightly constrained by how far their electorates can depart from sub-regional averages before coming up against one of the UK limits. However, care would still have to be taken not to draw up mathematically unviable sub-regions within the rules, in terms of the average number of electors for a given number of seats.

THE POTENTIAL NEED TO DIVIDE POLLING DISTRICTS

  15.  In recent decades parliamentary boundary reviews have traditionally been based on council wards which have not been divided between proposed constituencies, even if over time their own local electoral reviews often result in new wards straddling existing parliamentary boundaries. For electoral administration, local authorities often place a prefix in front of the letters used to designate polling districts in ward sequence across a whole authority to indicate the parliamentary constituently to which it belongs. A suffix is used to divide up polling districts split between existing constituencies, with the prefix used to indicate the new constituency arrangements that will come into force at the next general election. Thus it is possible to conduct a by-election efficiently using the same set of registers.

16.  As the bill requires seats to be within  5% of the UK average electorate, likely to be around 3,800, it will not be possible to leave wards undivided, especially in urban areas, and in Scotland since the introduction of Single Transferable Vote elections for local government. Birmingham is the most extreme example with wards ranging between 16,012 (Sheldon) and 20,208 (Springfield) local government electors. Thus in many areas it would be assumed that the smallest unit from which new constituencies could be devised would be the polling districts used to sub-divide wards purely for administrative convenience. However, my own polling district (CAB) in the Borough of Barnet just happens to have an electorate of around 5,000 which I doubt is that exceptional. Thus in certain places, even using polling districts could prove extremely difficult. To expect people to manually sub-divide a number of polling districts over a whole review area would be unreasonable—as a local activist it was bad enough doing this for the last local government boundary review for Barnet, when council wards can be devised from scratch.

  17.  Given the short time-frame proposed for electoral reviews, to level the playing field for parties making less well resourced submissions the respective boundary commissions should provide maps and detailed listings of electorates in every polling district in each review area. These numbers will have to be carefully compiled to discount people ineligible to vote in Westminster elections, including those under 18 on the relevant date. Since they are allocated polling numbers, it is not just a case of quoting the highest polling number for each polling district. Then if proposals with split polling districts have to be considered, a lot of the other ground work will have been covered by the collective resources which I have suggested should be made freely available. Even going down to polling district level in many places would mean that the number of potential permutations and combinations would be greatly increased within a given boundary review.

THE NATURE OF ALTERNATIVE VOTE ELECTION RESULTS

  18.  To correct a misunderstanding in the explanatory notes provided for the bill, if the Alternative Vote is adopted after being approved by a referendum, it is quite possible using this system for winning candidates to still be elected with fewer than half the total votes. This is possible because voters can decide whether or not to make further choices beyond their first or subsequent preferences, and thus votes are usually lost at each stage of the count as candidates are eliminated and votes are not fully transferred. The counting system set out in the bill uses exactly the same method as single seat elections using Single Transferable Vote. While this is clearly stated by the Electoral Reform Society website, these pages make the same mistaken assertion that winning candidates will always gain the support of more than half the people who voted. However, real examples of how Alternative Vote elections would be counted can be seen in recent by-election results for the Irish Dáil on the ElectionsIreland.org website. In these single seat elections, only one of the last six winners has got over half the votes and "Made Quota" (as it is put) rather than simply being "Elected" with the highest remaining vote.

1 September 2010





 
previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2010
Prepared 20 October 2010