Chapter Three: The Current System
18. Next we set out what appear to us to be both the qualities and the defects of the existing First Past The Post or simple plurality in single-member constituencies system (henceforth referred to as FPTP; a brief description of this and the other main electoral systems is provided on the inside of the front cover of the report).
The Virtues of FPTP
19. First the virtues. It is the incumbent system. It is familiar to the public, votes are simple to cast and count, and there is no surging popular agitation for change. It usually (although not invariably) leads to a one-party majority government. It thus enables electors, while nominally voting only for a local representative, in fact to choose the party they wish to form a government. It then leaves each member of Parliament with a direct relationship with a particular geographical area, on a basis of at least nominal equality in the sense that they are all elected in the same way. It also enables the electorate sharply and cleanly to rid itself of a unwanted government.
The case can be expanded in the following ways.
- By giving to all MPs each a unique position in their constituency for the period of their incumbency it encourages them to try to serve all their constituents well, and however partisan members may be at Westminster, to practise a more even-handed approach in their base.
- The single-party government outcome may be seen as assisting quick decisions - although there are one or two examples to the contrary - and the implementation of a sustained line of policy.
- Where a government fails, or at least disappoints, it can easily be punished by the electorate.
- By its 'winner takes all' and 'loser (particularly second or third losers) gets very little' effect it encourages parties to broaden their appeal and thus discourages extremism. (It can also be said, however, that in certain circumstances it encourages extremists to infiltrate moderate parties because the system gives them so little to gain on their own.)
- It offers to unorthodox MPs a degree of independence from excessive party control, provided (as many of them do) that they can retain the support of their local organisation.
Historical and Political Context
21. These are by no means negligible virtues, partly springing out of and partly providing the reasons why the system has persisted for a long time in Britain (although not, in exactly its present form, as long as is widely thought). There are one or two glosses which need to be put upon this list of virtues before more fundamental criticisms are considered. First the single member constituency is not an inherent part of the British parliamentary tradition. It was unusual until 1885, and only became the rule in 1950. Until the first date most seats were two-member, one (the City of London) four-member, supplemented by thirteen three-member ones in the large cities. These last were created by Disraeli's 1867 Reform Act, each elector having only two votes, the limitation introduced with the deliberate intention of providing for minority representation. Until 1950 a number of two-member boroughs persisted, in which it had been often the case that the two members were not of the same party; this was indeed the way in which most members of the early Labour party, frequently in double harness with a Liberal, secured their entry into Parliament. There were also the twelve university seats, three of which were two-member and one three-member, all of these multiple ones elected on a system of a Single Transferable Vote.
22. Second the FPTP system, although familiar, certainly could not be said in recent decades to have produced a House of Commons the functioning of which commands strong respect. There has been a long history of attempts to replace or at least substantially to modify the system. Many of these go back well into the nineteenth century. There were two high points of such attempts. First, the 1917 all-party Speaker's Conference which unanimously recommended a switch to a Single Transferable Vote system in the cities and large towns, accompanied by the use of the Alternative Vote in the counties. The various propositions foundered in a series of cross party currents with unfavourable votes in a not very well-attended wartime House of Commons. Then in 1931, under the second Labour government, a bill for the introduction of the Alternative Vote got through the House of Commons, but was rejected by the Lords and was lost with the break-up of that government in the following year.
23. The third occasion when there was a surge of criticism of FPTP was in the mid 1970s, when, following a perverse general election result in February 1974 (the Conservatives had a lead of 0.7% or 226,000 over Labour, but secured fewer seats, and the Liberals got only 2% of the seats for 19% of the vote), about a hundred Conservative MPs (in step with the CBI resolution of 1977) pronounced themselves in favour of electoral reform, the enthusiasm of many of them fading away during the long period of Conservative power in 1980s. However such fluctuation of view in accordance with changing party need has by no means been peculiar to the Conservatives. The Liberals were indifferent to the issue during their ten years of early twentieth century power, and as late as 1917 the London Liberal Federation even produced a pamphlet entitled The Case Against Proportional Representation. The Labour party showed matching hostility in the years from 1945 to 1979 when they enjoyed somewhat more than an equal share of power. The Labour party renewed its interest in the late 1980s which led to the admirable analysis of the Plant Report. There is enough here to prompt the cynical thought that there has been an element of 'The devil was sick, the devil a monk would be, the devil was well, the devil a devil he'd be' about the attitude of all parties to electoral reform. Their desire to improve the electoral system has tended to vary in inverse proportion to their ability to do anything about it.
24. The Liberal Democrats, in contrast with one side of their early twentieth century ancestry, are of course strongly in favour of electoral reform, and have a great interest in its effect. It is inevitable that when a system has heavily discriminated against a particular party, as FPTP has undoubtedly done against the Liberal Democrats, they are likely to be substantial beneficiaries of a change. But to have an interest in an outcome does not necessarily vitiate advocacy of it. Churchill had a great interest in victory in the Second World War. But that does not mean that the quality of his rallying cry should be dismissed on the ground that 'he would say that, wouldn't he'. Furthermore the Liberal Democrats are against the simple introduction of the Alternative Vote, despite the fact that it would be of substantial benefit to them.
25. We are of course aware of the considerations of realpolitik which have informed the bigger parties in framing their positions at and since the 1997 General Election. However, the narrow interests of the two parties are by no means obviously reflected in their positions. The Conservative party, which has often done well out of FPTP in the past, although not particularly so in 1992, was hit on the head by it in 1997. Not only were Scotland, Wales and all the big provincial cities (with the solitary exception of the Sutton Coldfield appendage of Birmingham) rendered an absolute electoral desert for them (in spite of their polling an aggregate of 1.8 million votes - or 17% of the total in these areas), but their overall reward of only 25% of the seats for 31% of the vote meant that the rest of the country was not adequately rich in compensating oases. Yet the Conservative evidence to this Commission shows that their faces are now firmly set against any change from the system which has temporarily treated them so harshly. The members of this cross-party Commission wonder, however, how fully the Conservative party has appreciated the longer-term nature of a bias against them which has recently entered the FPTP system and how additionally difficult this will make a quick and major resurrection. This point will be explained when we come to one of the broader deficiencies of FPTP in paragraphs 40-43 below.
26. The Labour party, per contra, has after many thirsty years had a cornucopia of luscious psephological fruit emptied over its head. FPTP, aided by some mutual tactical voting from and to the Liberal Democrats, has rewarded it with 63.6% of the seats for 43.2% of the vote. On a 'what we have we hold' basis 1997-8 might be expected to be the most improbable period for the Labour party leadership to contemplate electoral reform. Yet, perhaps on grounds of wider statesmanship, perhaps with a shrewd instinct that when you have as much as this you are historically very unlikely to hold anything like the whole of it, the Labour government, which is already legislating for a more proportional system for the European Parliament, the Scottish Parliament and the Welsh and London Assemblies, has set up this Commission, with the strong presumption that, if well argued, its recommendations will at least be taken seriously. If this disposition persists this Labour government will have the unique distinction of having broken the spell under which parties when they want to reform do not have the power and when they have the power do not want to reform. As a result of this knot the existing electoral system, in many ways irrational, and, to judge from most opinion polls on the subject, not particularly loved either, has persisted.
The Defects of FPTP
27. The deficiencies of FPTP are principally the following, many of which derive from a natural tendency of the system to disunite rather than to unite the country. This tendency shows itself in several ways.
28. FPTP exaggerates movements of opinion, and when they are strong produces mammoth majorities in the House of Commons. Since the war it has done this for Labour in 1945, 1966 (less sweepingly) and 1997, and for the Conservatives in 1959, 1983 and 1987. While there is a considerable case for some clear cut results, there are also disadvantages to 'landslide' majorities, which do not in general conduce to the effective working of the House of Commons. Landslide majorities, our researches suggest, are regarded with considerable suspicion by the wider public, perhaps more so even than coalitions. It is also the case that recent large majorities (both in 1987 and 1997) have been secured with a smaller percentage of the total vote (42.3% and 43.2% respectively) than in 1945 (48.3%), 1959 (49.4%) and 1966 (47.9%). This is of course largely a function of stronger support for a third party.
29. The FPTP system is peculiarly bad at allowing this third party support to express itself. Half a century ago, at the great mass plebiscite of 1951, when 82.5% voted and 96.8% of them voted for one of the two big parties, it was a negligible problem. The fact that the 2.5% who voted for the third party achieved an even lower percentage of the seats, barely 1%, was not a serious distortion. Already by 1974, as had been seen, this 2.5% had grown to 19.3% of the vote, but still yielded only 2.2% of the seats. And in 1983 the third party, then known as the Alliance, got 25.4% of the vote and 3.5% of the seats. Even in 1997, when the third party benefited from tactical voting, it still got only 7% of the seats for 16.8% of the vote.
30. This under-representation of a relatively strong minority party is very much a function of that party's appeal across geographical areas and occupational groups. When a party has a narrow but more intense beam, as with Plaid Cymru but less so for the Scottish Nationalists, its representation, although by no means perfect under the present system, approximates more to its strength. This is perverse, for a party's breadth of appeal is surely a favourable factor from the point of view of national cohesion, and its discouragement a count against an electoral system which heavily under-rewards it.
31. The same properties of FPTP tend to make it geographically divisive between the two leading parties, even though each of them can from time to time be rewarded by it with a vast jackpot. We have already seen how the 1997 election drove the Conservatives out of even minimal representation in Scotland, Wales and the big provincial cities of England. During the 1980s the Labour party was almost equally excluded from the more rapidly growing and more prosperous southern half of the country. South of a line from the Wash to the Severn estuary and outside London there were, in both 1983 and 1987, only three Labour seats. It was also the case that as a result of both of these elections there was no Labour MP for a predominantly rural English constituency. This, also, is a bifurcation which has recently become increasingly sharp. In 1945, for instance, there were three Labour members for Norfolk county divisions, which were then more rural than they are today. And in 1955 there were, unbelievable as it now seems, eight (out of fifteen) Conservative MPs for Glasgow. Such apartheid in electoral outcome is a heavy count against the system which produces it. It is a new form of Disraeli's two nations.
32. One thing that FPTP assuredly does not do is to allow the elector to exercise a free choice in both the selection of a constituency representative and the determination of the government of the country. It forces the voter to give priority to one or the other, and the evidence is that in the great majority of cases he or she deems it more important who is Prime Minister than who is member for their local constituency. As a result the choice of which individual is MP effectively rests not with the electorate but with the selecting body of whichever party is dominant in the area. Unless the electorate is grossly and rarely affronted (as appeared to be the case in the Tatton division of Cheshire at the 1997 election), individual popularity in any broad sense hardly enters into the process at all. This is not an inbred deficiency in all voting systems. Both the Additional Member System (as in Germany, see paragraphs 55-61 below) with its two votes and the Single Transferable Vote in multi-member constituencies (as practised in the Republic of Ireland, see paragraphs 50-54 below) allow the voter to combine influencing the choice of government and expressing a preference between individuals as local representative.
Voter choice and 'Making Votes Count'
33. The next criticism of FPTP is that it narrows the terrain over which the political battle is fought, and also, in an associated although not an identical point, excludes many voters from ever helping to elect a winning candidate. The essential contest between the two main parties is fought over about a hundred or at most 150 (out of 659) swingable constituencies. Even in a landslide election such as 1997 Conservative vulnerability or Labour hopes did not extend beyond the larger range, and in most elections the range has been even more narrowly confined. This indeed was explicitly recognised by what is regarded on all sides as the exceptionally efficient Labour machine in 1997. They concentrated their resources on what they had identified as the vulnerable hundred with all the clinical precision of the German general staff going for weak points in their 1870 or 1940 advances. Outside the chosen arena voters were deprived of (or spared from) the visits of party leaders, saw few canvassers, and were generally treated (by both sides) as either irrevocably damned or sufficiently saved as to qualify for being taken for granted.
34. To some extent the challenge of the third party provides an antidote to such complacency, sometimes threatening Labour in what would previously have been regarded as safe inner-city seats, and doing the same to the Conservatives in the far West Country and Wessex. This point can however hardly be called in aid of FPTP, for one of the most salient characteristics of the system is that it makes it as difficult as possible for a third party to win seats and thus does its best to render that threat innocuous.
35. The semi-corollary of a high proportion of the constituencies being in 'safe-seat' territory is not merely that many voters pass their entire adult lives without ever voting for a winning candidate but that they also do so without any realistic hope of influencing a result. In these circumstances it is perhaps remarkable that general election turnouts remain at or a little above a relatively respectable 70%, well down on the 80% plus of Britain in the early 1950s or of Germany last month, but a little higher than the Republic of Ireland or France and well up on the United States. As the Home Affairs Select Committee has recently argued, we should not be satisfied when 3 in 10 voters (although some of them are disenfranchised by an out-of-date electoral register) fail to use the five-yearly opportunity to influence their choice of government. Nevertheless we do not believe that this problem should be solved by compulsory voting.
36. Although FPTP is often referred to as a 'majoritarian' system this is an increasing misnomer at the constituency level. To a growing extent it is a 'plurality' rather than a 'majority' system. In the four elections of the 1950s an average of only 86 or 13.5% of MPs were elected without having the support of a majority of those voting in their constituency. In the two elections of the 1990s these figures have risen to an average of 286 or 44%. The change is of course a function of the growth of support for the third party (and the fourth in Scotland and Wales). But as a fundamental weakness of FPTP is that it is inherently ill-at-ease with anything more than a two-party pattern, this can hardly be regarded as an adequate excuse. It is a heavy count against a system which claims the special virtue of each MP being the chosen representative of his or her individual constituency if, in the case of nearly a half of them, more of the electors voted against than for them.
Perverse Results
37. There is also not merely the regular divergence from a majority but occasionally from a plurality in the country as a whole. The perverse result of the first 1974 election has already been referred to. There was also the arguably equally perverse one of 1951, when the Conservatives, although polling 250,000 less votes than Labour, won a small overall majority of 17 seats and skilfully built 13 years of power on this slender base. The irony of that result for Labour was that in terms of crosses on ballot papers it was their best result ever. Both in absolute numbers and percentage of the votes cast they did better than they had ever done before, or have ever done since - better than in 1945, better than in 1997 - and yet they lost.
38. It may be said that, if two elections of the fifteen since the war have produced perverse results, that is in itself unfortunate, but it nonetheless means that thirteen have given the victory to the party with more votes than any other and that is on average not at all bad. However risks have to be measured by their consequences and not merely by their incidence. Two rainy days out of fifteen would certainly be an acceptable risk for the planning of a picnic, but an air journey which has two chances out of fifteen of ending in a crash would most certainly not be. Nor, in the days of controversy about the death penalty, would for most people be a two in fifteen chance of hanging the innocent. A false election verdict might be regarded as about halfway between the two categories, which is well short of saying that two distorted results out of fifteen do not matter. Nonetheless, in fairness to FPTP, it should be noted that other electoral systems can also produce occasional irrational results.
Wider Representation
39. There is some, but not overwhelmingly strong evidence that FPTP is less good at producing parliamentary representation for women and for ethnic minorities than are most more proportional systems. In New Zealand, for example, (which we discuss in more detail at paragraphs 67-73) the proportion of women (30% of MPs there are now women), Maoris and ethnic groups increased dramatically following the introduction of a proportional system. And in Germany where a similar system is used the proportion of women in the Bundestag is 26%. Both are significantly higher even than the current UK figure of 18%, itself a great improvement upon the less than 10% upon which it was stuck for half a century. But the point should be noted without giving it a weight which it cannot bear. We can equally point to examples where a more proportional system has not been so successful in this area. In Ireland, for example, under the Single Transferable Vote rather than an Additional Member system, women make up only 13.9% of the Dail. We believe that, ultimately, under any system, it is the political parties who are responsible for candidate selection, and the matter is in their hands. Nevertheless, a party which has the will to increase female or minority representation might find it easier to do so under a system involving lists or slates of candidates than it would with a system which makes use exclusively of single-member constituencies.
Bias
40. A more certain, and in this list final, criticism of FPTP is its tendency to develop long periods of systemic bias against one or other of the two main parties. These periods of bias (apart from that against a widely-spread third party) are not necessarily permanent but while they last they are very difficult if not impossible to correct. They are in this respect rather like a little ice age or period of global warming.
41. Bias essentially arises when a given number of votes translates into significantly more seats for one party than for the other. For the post-war period until about 1970, as the graph below illustrates, it ran in favour of the Conservative party and against the Labour party. It was largely a consequence of Labour piling up large unneeded majorities in its heartland seats (of which the old mining constituencies were the most conspicuous examples) while failing to pick up a full share of the key voters in the marginal seats. In the 1970s and the early 1980s there were fluctuations around an approximate equality. In the two elections of the 1990s, however, the bias of 1945-70 has drastically reversed itself. The number of votes achieved by the Conservatives in 1992 was not substantially different from that achieved by Labour in 1997. But the former election yielded the Conservatives only what proved a shaky and erodable majority of 21 (and one over Labour of 65) whereas the latter gave Labour an overall majority of 179 (and one over the Conservatives of 255). The discrepancy arises from a mixture of causes, ranging from the over-representation of Scotland and Wales (from which the Conservatives are now wholly excluded), through some inequality in the size of English constituencies, the Boundary Commission being almost inevitably a bit behind the game, and the impact of the Liberal Democrats being now (much more than in the 1980s) favourable to Labour than to the Conservatives, to the most important but most elusive factor, which is that the lowest percentage polls are in Labour (often inner-city) seats, and that in consequence a given number of Labour votes now produces more seats than the same quantity of Conservative votes.
42. The combined strength of these factors is such that there is now an almost unanimous psephological opinion that at the last election an equality of nation-wide votes between the two parties would have produced a seat lead of circa 76 for Labour, or, put another way round, the Conservatives would have required a lead of approximately 6 1/2 % to give them an equality of seats with Labour. In order to obtain an overall majority, taking into account the Liberal Democrats and the Nationalist parties (and the prevalence of such overall majorities and the consequent security of single-party government is the central argument deployed for FPTP) they would have required a very much more substantial lead. While there can be no guarantee that the next election will produce precisely the same level of bias, we can say with some certainty that the system will, for a given level of votes, treat Labour better than it will the Conservatives.
43. While systemic bias could, on the record, be argued to display a certain impartiality, running for one long period in favour of one party and then for another period in favour of the other, such irrational alternations must be held as a count against the system. It is moreover a bias which could not by definition occur in a fully proportional system and which would be reduced by any significant move in that direction.
|
