Chapter Seven: The Case For and the Functioning of a Mixed System
108. What system do we think may better fulfil our four reference requirements and our additional two tests? If we do not go in an STV direction the alternative must be a variant of the Additional Member System (henceforth AMS). As was previously expounded (paras 55-61) this has worked very well in Germany for half a century. What we propose for Britain is however substantially different from the German system both practically and conceptually. The practical differences are such that it may indeed be better to describe it for reasons which will emerge later as AV with Top-up members rather than AMS, although the two, while distinct, are recognisably of the same broad family. The conceptual differences arise from a contrast of origins. The modern German system stems from the old strict proportionality of the Weimar Republic and proceeds by substantial modification (a threshold of 5% to avoid a multiplicity of splinter parties, an obligation upon a Bundestag which defeats a government to provide another one in its place before the vote of no confidence is valid, etc.) in order successfully to avoid the governmental weakness of Weimar. Our proposition for this country stems essentially from the British constituency tradition and proceeds by limited modification to render it less haphazard, less unfair to minority parties, and less nationally divisive in the sense of avoiding large areas of electoral desert for each of the two major parties.
General Advantages of a Mixed System
109. A principal advantage of such a mixed system is its flexibility. According to the proportion of top-up members which is fixed upon, and according somewhat also to whether they are elected nationally, or from big regions, or from a more local grouping of constituencies, varying degrees of priority can be given to proportionality on the one hand and to the constituency link on the other. This flexibility has enabled the Commission to steer to a point closest to fulfilling all four of our terms of reference, which, as well as greater proportionality and the maintenance of a constituency link, are an increase in voter choice and stability of government.
110. The essence of the system is that the elector would have the opportunity to cast two votes, the first for his choice of constituency MP, the second for an additional or Top-up member who would be elected for the specific and primary purpose of correcting the disproportionality left by the constituency outcomes, and could thus be crucial to determining the political colour of the next government. The second vote can be cast either for individuals or (as in Germany) for a party list without regard to the individuals on it. For reasons we develop in paragraphs 137-9 we greatly prefer an 'open list', giving the voter the ability to discriminate between individuals, to a closed party list. The counting of the second votes must be done in such a way that the central purpose of the 'Top-up', which is leverage towards proportionality, is maintained. This means that account must be taken, not only of how many second votes a party has received, but also of how many constituency seats in the area it has already won. The allocation of Top-up seats would proceed as follows:
- After the total number of second votes cast for each party have been counted, these numbers are then divided for each party by the number of constituencies gained in the Top-up area by that party plus one (adding one avoids the impossibility of dividing by zero and ensures that the party with the highest ratio of votes to seats receives the Top-up seat.)
- A Top-up member is then allocated to the party with the highest adjusted number of votes.
- Where there remains a further Top-up member to be allocated this process is repeated but taking into account any Top-up members already gained by each party. Parties should not be eligible for Top-up seats unless they have contested at least 50% of the constituencies in the Top-up area.
111. Voter choice is manifestly enhanced by the ability of electors under the new system to cast their two votes in different political directions and thus to escape from the dilemma outlined earlier that, under FPTP, they have either to subordinate their view of who is the individual candidate best for the constituency to their choice of government for the country, or (less frequently in practice as all the evidence shows) vice versa. Thus, to take a concrete example, many Conservative voters of the Tatton division would at the last election have been able to balance their vote for the Labour and Liberal-supported independent candidate by using their second vote for a Conservative additional member from Cheshire. Martin Bell would still have been elected, but natural Conservatives could have eased the strain of a vote for him being a vote against John Major.
112. From the point of view of stability of government there is no evidence that an additional member system, even in the extreme form of a 50:50 division between them and the constituency members, as practised in Germany, produces less stability of government than does FPTP. Furthermore there is no electoral system which is a guarantee against occasional periods of instability, as witness the already-cited FPTP results in Britain in 1922-4, 1950-51 and 1974. And, to cast the net of comparison wider, 'majoritarian' systems (very similar in effect to FPTP) have produced in France several periods of co-habitation (a government of a different political orientation from the President) and in the United States of a President with a hostile Congress.
113. The Commission has therefore seen the essence of its task as being to use the flexibility of a Top-up system to strike such a balance as best to reconcile the four requirements of our terms of reference with our view of fairness, both of representation and of proportionality of power (as set out in paragraphs 6-8), and to do so in a way which offers a reasonable chance of our work being fecund rather than sterile.
The Status of MPs in a Mixed System
114. AMS or its variants involves electing MPs by two different methods, and thus having, as some might put it, two classes of MPs. This is however unavoidable, unless either the search for greater fairness (and rationality) is to be abandoned or a solution is to be sought through universal STV or having all members elected on a list system, both of which solutions the Commission believes would be less acceptable.
115. Some (mainly politicians) have raised the perceived problem of two categories of MPs. We do not see this problem as formidable. In the first place it is not really such a break with the British tradition as may superficially be thought. Throughout the nineteenth century there was a considerable difference of category between county and borough members. In addition the Scottish and Irish members were mostly elected on a different franchise from the English and Welsh ones. And some constituencies were always regarded, at any rate by some politicians, as having a greater prestige than others. Some saw the City of London as being peculiarly appropriate for great financiers or, on two occasions, for party leaders, and Lord Randolph Churchill several times tried to escape from what he regarded as the mediocrity of South Paddington to the romance of a 'great industrial borough'. The university seats, different both in electorate and in electoral method, persisted until 1950. And we are about to move into a position in which some MPs will represent areas with devolution and hence will have more restricted constitutency duties. Furthermore there has long been a difference of practice, if not of theory, between those who entered Parliament primarily to seek a national role, often switching from one constituency to an utterly disparate one in order to achieve it, and those who sprang out of a particular locality, found their greatest satisfaction in representing and serving it, and could not easily have been imagined, by themselves or others, as migrating to a seat in a different part of the country.
116. A major disadvantage of the German system, if transported wholesale to Britain is that there would be too many list members. There is the equivalent of one list member for every constituency, and as many of them aspire to become directly elected constituency MPs, they concentrate their hopes and effort upon a particular constituency, in effect making themselves a shadow member for it, but with the substantial advantage over an adopted parliamentary candidate that they have all the advantages - access to ministers, full parliamentary expenses and salaried time when the Bundestag is not sitting - of the directly elected MP but without the constituency responsibility. By the criterion of a level playing field for the next elections this may be fair, but it is also inimical to the best traditions of an MP performing at least a semi-impartial role in his or her constituency between elections and endeavouring to serve all constituents - those who supported him or her and those who did not - with equal diligence. If there is a rival and equally active MP of an opposing party on the scene this link is almost inevitably weakened if not broken.
117. Another disadvantage of such a high proportion of list members - to be set against the highly proportional outcomes which it secures - is that, particularly in scattered rural areas and doubly so if more rigid equity in constituency populations is sought, it loosens the local link. Apart from more extreme examples in Scotland and Wales, it would mean, to take a specific example from a fairly populous county of England, that there would only be one seat for the whole of north and west Devon stretching from Tavistock to Ilfracombe. And there is a third, and to many the principal disadvantage of a 50% 'top-up', which is that it would make coalitions if not inevitable very much the norm.
Proportionality and Stable Government
118. It therefore appears to us that anything like 50% of Top-up members is undesirable. Fortunately it is also unnecessary. Studies done for the Commission show that a substantial degree of proportionality could be obtained for the country as a whole with a top-up of 15-20%.
119. The next and central question which arises is why it is substantial proportionality and not complete proportionality that we are seeking. This necessarily flows from our terms of reference which require us to balance four competing criteria, of which broad proportionality is only one. We are required to steer a path which, so far as possible, reconciles this with the other three criteria.
120. These external constraints are re-inforced by the consensual view of the Commission on the future political framework which we regard as desirable in itself and most likely to be acceptable to a majority of the British people in a referendum. As argued in Chapter 4 we do not recoil with horror from the very idea of coalitions, regarding them, on the basis both of British and of some foreign experience, as capable of providing effective and decisive governments. Their quality depends to some considerable extent on whether the coalitions are 'honest', defined in the sense that those within them agree with each other more than they do with those outside (sometimes indeed they may agree more than do those within single-party governments), rather than mere patchworks of opportunism.
121. This does not mean that permanent coalition is desirable. We would prefer, and certainly regard as more compatible with the totality of our terms of reference, that when there is a strong surge in one political direction or the other, single-party governments, even if with somewhat under 50% of the vote, should stand out like mountainous land masses rising above the surface of the ocean. This should clearly have been the case with the Attlee government in 1945 and with the Macmillan government in 1959. No purpose of justice or efficacy would have been served by either being forced to coalesce with the very small Liberal party of those years.
122. The position becomes more complicated with the decline of the two-party duopoly from 1974. Clearly in these past 24 years when governments found it easier to get big majorities with substantially lower percentages of the vote, any reformed system would have involved some coalitions. Particularly in 1974-9 and in 1992-7 they may well have been healthier for effective and responsive government. Nevertheless we would not wish to propound a system which would involve persistent coalition. Reverting to the comparison made with the German system, one aspect which we find difficult to defend, in spite of the striking overall economic and political success of German government, has been the permanent hinge position of the very small Free Democratic Party. In spite of having a voting strength on average barely a third of that of Alliance/Liberal Democrats since 1983, it was continuously in office from 1969 to 1998, with a perpetual grip on the Foreign Ministry and of two or three other cabinet seats as well. The pattern is being unfrozen by the rise of the Green party, which has recently moved to working inside rather than outside the system, and which in alliance with the SPD has excluded the FDP from the next government. 123. Nevertheless, such a period of 29 years of almost guaranteed continuous if subsidiary power, for thirteen years with one ally, for sixteen years with another, obtained with an average of 8.7% (down to 6.2% at this year's election) of the vote seems to us to be an anomaly comparable (in the opposite direction) with the SDP/Liberal Alliance receiving only 3.5% of the seats for 25.4% of the vote in the 1983, or the Conservatives winning, mostly with big majorities, four consecutive elections with an average of only 42.6% of the vote. The Commission does not wish to design such a position of constant privilege for a hinge party, and it does not believe there is anything inherent in an additional member/Top-up system which makes it do so.
124. The Commission believes that it can iron out the gross anomalies, such as the Conservative and allied landslide of 1931, when a loss by Labour of 7% of the vote meant that it was deprived of 83% of its seats, or of Labour in February 1974, per contra, securing the right to form an independent government with only 37.2% of the votes, and fewer moreover than those of the Conservatives, without producing any likelihood of a stagnant and unhealthy prospect of constant and unchangeable coalition. That is how it sees its mandate of greater proportionality accompanied by respect for the other criteria.
Method of Electing Constituency Members
125. A further question to be determined is whether, within an Additional Member/Top-up scheme, constituency MPs should be elected by FPTP or by AV. AV, it will be recalled, was shown in paragraph 81 as having many advantages, not least that it clearly increases voter choice (point iii of our terms of reference) and that it counteracts the growing tendency for many MPs to be elected by a plurality and not a majority of their constituents. To re-inforce the point made at paragraph 36 no fewer than 312 of the present 659 Members of the House of Commons were elected with less than 50% of the vote and of those 49 were elected with less than 40%.
126. Under our system, AV would have a number of positive features which persuade a majority of us that it would be superior to FPTP as a method of choosing constituency representatives. First, there will be many fewer 'wasted votes' in the constituency side of the election, and far more voters will potentially influence the result. This, we hope, will encourage turn-out and participation. Second, it would encourage serious candidates to pitch their appeal to a majority of their constituents, rather than just seeking to target a hard-core minority of the party faithful. This should lead to more inclusive politics than FPTP. Third, because second and subsequent preferences may count, it will discourage individual candidates from intemperate attacks on their rivals, since they will be hoping to gain their second votes and will not wish to alienate their supporters. This should contribute to the more consensual and less confrontational politics to which the majority of the public appear to aspire.
127. On top of these arguments, the use of AV has one other and crucial advantage. AV counters one important objection to electoral reform. This is the tendency to transfer power from voters to the subsequent deals of politicians. The recent example of New Zealand is widely cited in this regard. New Zealand is an example of the potential disadvantage of using FPTP for constituency elections under a mixed-system. For using FPTP means that each party in each constituency will seek to confront all others in order to maximise its own seats in the election, doing any necessary deals only after the polls have closed. By contrast, the use of AV in constituencies militates strongly against this.
128. There can be an element of chance about individual constituency outcomes under AV, as there can be under almost all systems but it is probably less than that which exists under FPTP once tactical voting has become a factor, as it recently has in Britain. Those Conservative MPs who at the last election were faced by a clear leading challenger, whether it was a Labour or Liberal Democrat one, tended to lose their seats. Those who faced almost evenly divided opposition held them. That is a real example of randomness. Nor in our view is AV too complex for voters. AV is widely used in voluntary elections from the Church Synod to many local societies, and is widely understood.
129. On its own, AV, which we also noted had the advantage that it could be implemented quickly without any change to constituency boundaries, was rejected on the ground that in the circumstances of 1997 (and maybe of the next election) it would be unfair to those who support the Conservative party, and that it is particularly important that parties in adversity should not be treated unfairly. It is also important to remember that in 1983, 1987 and 1992 AV would have brought no similar benefit to the Labour party over the Conservative party, 130. The duty of the Commission, however, is to design a system which is of validity not just for the last or even the next election alone but for a much longer timespan. This gains particular force if the election after next, because of the burden placed upon the Boundary Commission by the need to reduce by approximately a fifth the number of constituencies, is the earliest point at which there is a realistic chance of our recommendations coming into operation. While we regret that there cannot be an earlier implementation we feel that by then the salient force of the 1997 election, which was the over-riding desire to end the long period of Conservative government, and which may also have some residual force at the next election of 2001 or 2002, must surely have expended itself.
131. This, in combination with the still more important corrective mechanism of the Top-up, removes the decisive objection to AV on its own, which was its potential short-term unfairness to Conservative party supporters. In these circumstances the general view of the Commission is that the arguments outlined above swing in favour of AV for the constituencies. To sum up, AV extends voter choice while FPTP limits it. AV will substantially reduce the numbers of wasted votes because voters will be able to influence outcomes through second and sometimes third preferences and thus correct one of the mischiefs of the current system which is effectively to disenfranchise voters whose first preference is for small parties or independent candidates. And AV also guarantees that MPs are elected by a majority of those who vote in their constituencies, which majority is thereby given a stake in the validity of the representation. Lord Alexander has strong doubts about AV, and sets out his reasons, which are both conceptual and practical, in the attached note of reservation. He makes clear, however, that he supports wholly the recommendation of a 'top-up' with a great majority of constituency members but he considers that those members should continue as at present to be elected under FPTP.
Electing Top-up Members: A Local Solution
132. The next issue is the geographical basis on which the list members should be allocated. We have wished to do this on a basis which is at once devolved and pays regard to historic local entities. We propose that for the whole of the United Kingdom there should be 80 Top-up areas. Two of them would be in Northern Ireland (see para 141 below for our recommendation that under the AV plus system Northern Ireland would not require a different system), eight in Scotland and five in Wales. In the case of these latter two countries there has already been an allocation of Top-up areas for use in the elections to the Scottish Parliament and the Welsh Assembly. We see no reason to complicate the position by departing from these.
133. This would leave 65 Top-up areas for England. What follows is offered for illustrative purposes. Outside the metropolitan areas we propose that these should be the "preserved" counties, although with the four largest in population - Kent, Essex, Lancashire and Hampshire - each split into two. London might be split into seven Top-up areas, conforming as far as possible to borough boundaries. In the ten other metropolitan areas of England there would be sixteen Top-up areas. In South Yorkshire, for instance, there might be the two Top-up seats of Sheffield with Rotherham and Barnsley with Doncaster. In West Yorkshire there might be three: one for Leeds, one for Bradford with Halifax, and one for the remaining urban areas comprising the towns of Huddersfield, Wakefield and Dewsbury. The full illustrative list, not only for England, but for Scotland, Wales and Northern Ireland is appended. All these 80 (in the UK) areas would have at least one Top-up member. The more populous ones and both Top-areas in Northern Ireland would have two. Where the dividing line is placed clearly depends on the exact proportion of Top-up seats which is thought desirable, and this is dealt with later in paragraphs 151-161.
134. The strong advantage, as it appears to the Commission, of this devolved county/city allocation is (i) that its would help to restore some cohesion of representation to the recently weakened traditional localities of Britain; and (ii) that one or two additional members locally anchored to quite small areas comprising a maximum of 12 and an average of eight current constituencies put together are, we believe, more easily assimilable into the British political culture and indeed the Parliamentary system than would be a flock of unattached birds clouding the sky and wheeling under central party directions.
The Role of Top-up Members
135. It is highly desirable that they should not be perceived as second grade MPs. It will of course continue to be for the members elected for constituencies to represent those constituencies at Westminster. We see the Top-up members as serving a new role in representing in the House of Commons the broader interests of the counties and cities. Top-up members will widen the opportunity for and access to parliamentary representation in two ways. It will be open to Top-up members at Westminster to represent the interests of their wider constituency on the hinge between the local authorities areas and Whitehall. And without prejudice to this broad responsibility to the wider constituency, the existence of Top-up members at county and city level across the country will provide representation in Parliament for minority political opinion, which because of the existence of electoral deserts created by the existing system, is currently extruded from large swathes of the country. The scope and breadth of these responsibilities should be such as to give to MPs elected as Top-up members fully equal status in the House of Commons to that of those elected in a constituency.
136. Top-up or additional members are strongly likely to come from a spread of parties. Had the 1997 election been fought under our recommended system, for instance, there would have been more Conservative than Liberal Democrat Top-up members returned. Obviously when a party does as well in direct constituency elections as the Labour party then did there is not much scope for them to receive Top-up members as well. Even so they would have gained a few in the southern counties and at a more normally balanced election this few would have been considerably increased.
137. While we do not think Top-up candidates should be legally prevented from simultaneously fighting individual constituencies (or vice versa), even more strongly do we feel that it should not be obligatory upon them to do so, nor indeed desirable. For this reason, amongst others, we differ from one of the recommendations of the admirable report of the Hansard Commission on Electoral Reform of 1976. This Commission was set up under the auspices of the Hansard Society and the chairmanship of the Conservative constitutional historian Lord Blake, with a distinguished membership including the then Chairman of ICI (Sir Jack Callard), Professor Ralf Dahrendorf, Lord O'Neill of the Maine, Baroness Seear and Richard Wood MP. It recommended an Additional Member System, and we have been much influenced by its powerfully presented argument. We differ however from its subsidiary proposal that list members should be chosen from the 'best losers'. We think that the concept of losers being transformed into winners, like base metal into gold, would not be an easy one to explain. Furthermore, the proportionally corrective function of additional members cannot always be performed by choosing 'best losers'. The second or third 'best loser' would quite often have to be chosen. And beyond that it would make a constituency contest obligatory for any additional member. We are in favour of greater flexibility on this last point.
|
