Electoral Reform in the United Kingdom: Lessons From the 2011 Alternative Vote Referendum

Transcription

1 Electoral Reform in the United Kingdom: Lessons From the 2011 Alternative Vote Referendum Abigail L. Heller Advisor: Professor Matthew Schousen GOV490, Government Departmental Honors Thesis Defended April 30, 2012 Graduation May 12, 2012

2 Table of Contents Chapter 1: Introduction 3 Chapter 2: Literature Review...6 Chapter 3: Case Studies..16 Chapter 4: A Theory of Electoral System Reform..26 Chapter 5: History of the British Case 38 Chapter 6: Electoral Reform in the United Kingdom: Why Now?.53 Chapter 7: Why the 2011 Referendum Failed...78 Chapter 8: Conclusion 103 Works Cited 111 2

3 Chapter 1: Introduction Many scholars quite rightly identify 12 May 2010 as an historic day for the United Kingdom. 1 On that day the first coalition government since World War II was formed. This was after the general election of 6 May had returned the first hung parliament (no single party won a majority of seats) since the 1970s. However, one might argue that for the Liberal Democrats at least, 5 May 2011 was equally important. It was then that a referendum on the voting system for the Westminster Parliament was held. Given the troubling economic conditions, rising debt, War on Terror, and other concerns, how did electoral reform become a critical issue put to the British voters? An electoral system is the set of rules that structure how votes are cast at elections for a representative assembly and how these votes are then converted into seats in that assembly (Gallagher and Mitchell 2005a, 3). Different systems convert votes into seats in different ways, some more proportionally than others. Therefore, the type of electoral system can affect representation, accountability, stability, type of government single-party versus coalition the party system, and the experience of democracy more generally (Gallagher and Mitchell 2005a; see also: Duverger 1984; Farrell 2001). The choice of an electoral system involves trade-offs between these key components of democratic governance. Electoral system reforms, therefore, include those changes that alter the way votes are cast and/or how the votes are converted into seats. There is a spectrum of electoral reforms that ranges from the relatively minor, such as the alteration of ballot access requirement or a change in the type of proportional formula, to major 1 In this work, the terms United Kingdom and Britain are used very specifically. The UK refers to the United Kingdom of Great Britain and Northern Ireland, whereas Britain refers only to Great Britain (England, Wales, and Scotland). Britain will be used in direct quotations, when referring to historical Great Britain, or when referring to the party system in Britain (excluding Northern Ireland). 3

4 reforms in which plurality rules are replaced by proportional or mixed systems (Katz 2005, 60). For our purposes the phrase electoral reform refers to the major changes that are less frequent than the minor alterations. Electoral reform is an important topic of inquiry because institutions such as electoral systems matter. Institutions affect which candidates are elected, the parties in power, and the policies adopted by the government. When electoral reform is debated, the future of politics in that nation is being discussed. At the heart of electoral system debates are such democratic principles as accountability and representation. Although the debate may take place within a specific country, the lessons learned, whatever the outcome of the reform attempt, can be applied to other nations. Relatedly, lessons from reform in established nations could be applied to new democracies. Electoral reform is also an important subject of inquiry because reforms have been occurring in a growing number of countries with increasing frequency since the 1990s. As an increasingly pervasive phenomenon, it deserves investigation. The UK is used as the primary case study because it was an instance of a very recent reform attempt that has not yet been widely examined. This work uses the case of the United Kingdom to explore the question: how does institutional reform, and specifically electoral reform, become a salient political issue and what affects its success? In the course of this work, I will develop an original model of electoral system reform. This constitutes a significant contribution, as others have tended to focus on parts of the reform process rather than suggesting a comprehensive model. This theory, if it withstands the test of applying it to multiple countries, could potentially be used by political scientists or even by policy-makers to predict electoral reform. For these reasons, this research question is important and worthy of examination. 4

5 The two-part model of electoral system change is based on the extant literature as well as three brief case studies. These cases are Australia and New Zealand where reform was successful, and Canada where reform attempts to date have failed. Finally, the bulk of the work will test the model against the case of the United Kingdom. This will allow for systematic exploration of the British case in addition to beginning to test the theory. However, it should be noted that this model, while proposed as a potential model for all cases of reform, will not be tested against a wide variety of cases and therefore is limited until such testing is complete. 5

6 Chapter 2: Literature Review Electoral Systems Theory Single-member plurality, proportional representation, alternative vote, single transferable vote, mixed-member proportional, mixed-member majoritarian even from this (incomplete) list, it is evident that there is a wide variety of electoral systems in use today. Electoral systems matter and the differences among them can be substantial and meaningful. To paraphrase Michael Gallagher and Paul Mitchell, electoral systems are crucial because they are the connector between citizens, government, and the policies that governments enact on behalf of the people (2005a, 3). In short, electoral systems are what make representative government possible. There are several in which electoral systems may vary, including district magnitude, the number of votes a voter may cast, the ballot structure, the degree of choice a voter has with regard to the individual candidates, and the levels of seat allocation (Farrell 2001; Gallagher and Mitchell 2005a; Lijphart 1999). These seemingly insignificant or overly technical differences matter because they affect the aforementioned democratic values of representation and accountability. For our purposes, the degree of choice voters have regarding candidates will not be discussed because the countries and systems examined here do not vary on this dimension (in all of our examples, the voter has no choice of the candidates within the party although of course there is a choice of candidates between parties). The first type of variation is the district magnitude (DM) or the number of seats in each constituency (Gallagher and Mitchell 2005a, 6; Lijphart 1999, 150). Not all countries have systems with an equal district magnitude in every constituency. Further, generally speaking, there is huge variation in DM from countries like the UK and the US, which use a district magnitude of one (one seat per constituency) to Israel where 6

7 the district magnitude is 120 (120 seats all elected from one constituency) because the entire country comprises one large district (Gallagher and Mitchell 2005a, 6). The second way in which electoral systems vary is the number of votes citizens cast. In most cases, including the UK, citizens cast only one vote on election day. However, under mixed systems such as those used in New Zealand and Germany, everyone has two votes one for a candidate to represent their constituency and one for a party (Gallagher and Mitchell 2005a, 7). This relates to a third distinction between systems, which is the ballot structure. Douglas Rae distinguished between categorical voting in which citizens vote for only one party and ordinal voting, in which voters rank-order parties or candidates such as under the alternative vote (AV) (Gallagher and Mitchell 2005a, 7). The final variation between electoral systems is the number of levels of seat allocation. When one examines all electoral systems in use, this can become fairly complicated. However, for our purposes, the important distinction is between one and two levels. In the UK, Canada, the US (all single-member plurality or SMP), and Australia (AV), there is only one level of seat allocation. This means that the national total of seats is equal to the sum of the seats [each party] won in each of the constituencies (Gallagher and Mitchell 2005a, 11). In contrast, there are two levels of seat allocation in New Zealand, which uses a mixed electoral system. There the total number of national seats is determined by both the number of MPs elected from individual constituencies and the number of MPs elected from party lists according to proportional representation (Gallagher and Mitchell 2005a, 12). All of these dimensions aid in distinguishing among electoral systems. However, there are also other ways of classifying systems. The preceding distinctions are based on what David Farrell calls component parts (2001, 6). Notably, he also considers the electoral formula in use to 7

8 be a component part. An alternative way to classify electoral systems is by their output, i.e. how votes are translated to seats (Farrell 2001, 5). This leads to three distinct categories proportional, non-proportional, and semi-proportional. However, one of the most commonly used classification systems is based on electoral formula proportional, plurality, or majority. Maurice Duverger (1954), William Riker (1986), and Arend Lijphart (1999, although he condenses it to just majoritarian and consensus democracies) all use this type of classification system. While these are the common distinctions based on electoral formula, in recent years, mixed systems have become more widely used. These systems vary, but are so-called because they combine elements of the others (usually, proportional and plurality) and are generally placed in the category of proportional systems because they are at least partially proportional (Lijphart 1999). For now, I will use electoral formula to classify electoral systems, although the other component parts will also be discussed. Majority systems require candidates to win an absolute majority (50% plus one vote) in order to be seated in parliament. This can take several forms, including the second ballot (essentially a run-off election, although this is commonly used for presidential elections rather than legislative elections) and the alternative vote (AV) (Lijphart 1999, 147). Under AV, voters rank order all the candidates on the ballot in their district. If one person wins a majority on the first count, the election is decided. However, if no candidate receives a majority, the candidate with the fewest votes is dropped and his or her first choice ballots are redistributed to their second choice. This process of dropping the lowest vote-earner and redistributing votes continues until one person receives a majority (Lijphart 1999, 147). Plurality systems on the other hand set a lower bar for winning. Under these types of systems, the winning candidate only needs more votes than any other competitors to win (this 8

9 may be a majority of the votes cast, but it may only be a plurality) (Lijphart 1999, 146). Although plurality can be used with multi-member districts (recall the discussion of district magnitude), single-member districts are much more common (Norris 2004, 42). When combined with single-member districts, this system is called single-member plurality or SMP. In the UK and the US, it is commonly, if rather misleadingly, termed first-past-the-post (FPTP). At their most basic, proportional systems allocate seats in proportion to the number of votes a party receives so that, theoretically, if a party secured 20% of the vote, it would win 20% of the seats in the legislature. However, the formula for converting votes into seats varies so that strict proportionality may not hold. There are three main types of proportional representation or PR list systems (which may be open, closed, or partly open), mixed systems, and single transferable vote systems (STV, often called PR-STV in the UK) (Lijphart 1999). Although there are variations, list systems allow parties to propose lists of candidates and citizens vote for one party list. The seats are then allocated based on the number of votes each party receives. These systems use multi-member districts so that there are two or more members of parliament elected from each constituency (Lijphart 1999, 147). Under mixed systems, some portion of the legislators are elected via SMP and the rest are elected using list PR. Each voter is given two votes one for the individual district candidate and one for the party list. This type of mixed system is used in New Zealand where it is called mixed-member proportional or MMP (Lijphart 1999, 147). The single transferable vote (STV) is the final type of proportional representation. It is not necessary to go into detail here, but essentially under STV voters rank order candidates similar to AV, but the formula for allocating seats is more complicated (Lijphart 1999, 148). Up to this point, the basics of electoral systems have been discussed. Scholars agree on these facts, with the possible exception of preferring one way of categorizing systems to another. 9

10 It is the effects of these systems that are often debated. The debate over how electoral systems affect party systems is one of the most basic and also sets the stage for debates about representation and accountability. Maurice Duverger advanced one of the most important and frequently cited arguments. What has become known as Duverger s Law states, the simplemajority single-ballot system favours [sic] the two-party system (Duverger 1954, 217) (note that some translations replace favours with tends to ). Additionally, the simple-majority system with second ballot and proportional representation favour multi-partism [sic] (Duverger 1954, 239). William Riker, on the other hand, is quite critical of Duverger. He argues that Duverger formulated the law deterministically; Riker takes issue with the notion that SMP causes a twoparty system when he noted several exceptions (including Canada and India) (Riker 1986). Further, Australia, in his view, constitutes a challenge to the hypothesis that a majority system will lead to a multi-party system (Riker 1986). This leads Riker to reformulate Duverger s Law to read plurality election rules bring about and maintain two-party competition, except in countries where third parties nationally are continually one of the two parties locally and except for countries where one party among several is almost always the Condorcet winner in elections (Riker 1986, 32). Riker s main problem with Duverger s Law is that it is formulated in a deterministic manner. Douglas Rae similarly objects to Duverger s argument that plurality formulae cause two-party systems and attempts to reformulate the law (Rae 1971). Further, Giovanni Sartori argues that Duverger incorrectly assumes that a causal relation can be warranted by a correlation (1986, 44). Collectively, the main criticism of Duverger is that he asserts a causal relationship between the electoral system and the party system when none exists. This is, I believe, an overly critical reading of Duverger. Duverger himself admits that he 10

11 originally stated what later became known as Duverger s Law at a 1945 conference and in that instance argued, a particular electoral system would lead to a particular party system (Duverger 1986, 70). However, in every later (published) formulation, Duverger used the phrase tends to lead (Duverger 1986, 70). Duverger recognized that there were exceptions to his law, but he was observing a trend that holds in the vast majority of cases. Pippa Norris uses an analysis of thirty-two countries to show that in its more probabilistic formulation, Duverger s law holds (2004, 94). When stated in this probabilistic manner, Duverger s Law is generally accepted. This leads to more normative arguments about the effects of electoral systems. Since plurality tends to lead to two-party systems, it also tends to be associated with single-party majority government in contrast to multi-party systems (under PR), which tend to result in coalition governments (Blais 1991, 240; Norris 2004, 94). Now that the scholarly literature on the effects of electoral systems is understood, what do scholars argue leads to changes in electoral systems? Electoral Reform Theory Generally speaking, major electoral reform is quite rare, but it does happen (Katz 2005). Under what circumstances should we expect electoral reform to occur? In other words, what factors contribute to reform? Scholarly debate on this topic has increased in recent years. Although each scholar assigns different weight to the various factors, it is possible to observe five major factors that have emerged in the literature alongside several other short-term factors or triggers. The five major factors contributing to reform are: (1) the parties in power wish to reform the system, (2) popular dissatisfaction with the political process, (3) changes in the 11

12 electoral climate, (4) external pressures for change, and (5) the ability of the voters to choose the electoral system against the wishes of parties in power. The first point that almost every electoral reform scholar makes is that reform can rarely occur without the support or even initiation of the parties in power. Since reform, in most cases, must be passed by the legislature, incumbents must agree to the reform (Benoit 2004, 367; Bowler, Donovan, and Karp 2006, 435; Butler 1984, 229; Donovan and Bowler 2011, 240; Katz 2005, 60, 63; Nohlen 1984, ; Shepsle 2001, 322). Many scholars argue this is the reason the world has seen relatively few major reforms the incumbents are, by definition, winning under the existing electoral system and have little incentive to change the system through which they were elected. Therefore, one of the factors contributing to electoral reform is the preferences of the parties in power. However, why would people or parties who are winning under the current electoral system wish to change it? In fact, there are several reasons a party in power may want to change the electoral system or may agree to a change. First, the incumbents may believe that it is in their interest to change the system because the existing system will not continue to benefit them as it has in the past (in other words, they do not think they will continue to win and so wish to change the system) (Katz 2005, 63). Importantly, party members may believe changing the electoral system is in their interest, but may be mistaken or overly optimistic (Katz 2005, 63). The critical point is that the party believes a change will be in its interests and so agrees to or initiates reform. A second reason a governing coalition may agree to reform the system against the wishes of some members is that there is a division of interest among the members (Katz 2005, 62-63). Whether discussing a coalition between parties or a coalition of members within one party, interests will diverge and the party or coalition of parties may adopt a reform that would seem 12

13 contrary to their interests if it were being considered as a unitary actor. Third, a party may wish to change the electoral system because it is willing to trade short-term electoral advantage for some predicted long-term change in the competitive system (Katz 2005, 63). Finally, in a similar fashion, parties may be willing to trade electoral advantage for other goals (Katz 2005, 63). Any of these reasons explains why it may be rational for a party that has won under the existing electoral rules to agree to change. The second major factor that contributes to electoral reform is the public s dissatisfaction with the political process. Scandals, government reversals or failures, and perceived unfairness of the current electoral system may all contribute to popular dissatisfaction (Butler 1984, 226, ; Donovan and Bowler 2011, 246; Dummett 1997, 8; Dunleavy and Margetts 1995, 25; Farrell 2001, 180; Lijphart and Grofman 1984, 11; Norris 1995a, 7). Political scandals such as corruption scandals can seriously damage public trust in the government or political system, and in some cases, can lead to a desire to change the system (Norris 1995a). Similarly, the inability of government to effectively handle the problems facing the nation can lead to popular dissatisfaction. Importantly, the electoral system itself may be seen as unfair. Cases where one party wins a plurality at the election, but another party wins the majority of the seats, make the electoral system controversial (Lijphart and Grofman 1984, 11). Katz explains that public outrage is rarely, if ever, outrage about the electoral system itself, although the electoral system may come to be blamed for, or to be seen as symptomatic of, more deep-seated problems (2005, 69). This public outrage over corruption, election outcomes, the inability of the government to deal with society s problems and the like can lead to electoral reform. A third important factor in electoral reform is natural change in the electoral arena. When the electoral situation changes, parties in power may be willing or feel forced to change 13

14 the electoral system (Boix 1999, 609; Butler 1984, 227; Dunleavy and Margetts 1995; Farrell 2001, 180; Norris 1995a, 7). In particular, the rise of new or previously minor parties may be a catalyst for institutional change. Electoral reform in the direction of a more proportional system is more likely to occur in countries with a larger number of effective parties (Colomer 2004, 62). In other words, an existing multi-party system favors a change to a proportional electoral system. Conversely, fewer effective parties make reform in the direction of plurality or majority more likely to occur (Colomer 2004, 63). This change in the party system or electoral arena can lead to reform in several ways. First, a major party may feel threatened and wish to change the rules to protect its interests. Second, the rise of a minor party can highlight the disproportionality of the electoral system and factor into public dissatisfaction with the electoral system (Boix 1999, 609, 611). Third, new parties may be in a position to directly affect change (i.e. they are now in government). A fourth factor in electoral reform is external pressures or shocks to the existing system that can exert pressure for change or the choice of a particular system (Benoit 2004, 372; Butler 1984, 226). One such shock to the system is the enfranchisement of large numbers of people previously excluded (Katz 2005, 57). Another example of external forces exerting pressure on the choice of a certain electoral system is the inheritance of electoral institutions from colonial rulers (Benoit 2004, 372). The role of international political and financial organizations in shaping electoral institutions would be a further example (Benoit 2004, 372). But this factor is important not only in the choice of an electoral system, but also when a country considers changes to the existing system. Although the United Kingdom uses SMP to elect members of Parliament (MPs), its membership in the European Union means that in order to elect members of the European Parliament (MEPs), a proportional system must be used. This helps bring the 14

15 issue of electoral reform into the public consciousness and constitutes an external (if not particularly strong) pressure for domestic electoral reform (Butler 1984, 226). The final major factor contributing to reform is the voters ability to choose the electoral system. This primarily takes the form of referenda (Benoit 2004, 372; Farrell 2001, ; Norris 1995a, 7). If the voters have the option to initiate binding referenda, it is possible for electoral reform to happen against the wishes of the parties in power. This, as we will see, is similar to what happened in New Zealand in In addition to the major or long-term factors that can contribute to electoral reform, several short-term factors can also trigger reform. These include factors such as the particular leaders in power, pressure groups, reform movements, and crises (Donovan and Bowler 2011, 244; Dunleavy and Margetts 1995, 22; Katz 2005, 74; Norris 1995a, 7). Without some combination of the major factors listed above, these minor factors would most likely not lead to reform. However, they can be contributing factors to a specific reform. As is evident from the preceding discussion, many of the aforementioned factors are related. Popular dissatisfaction may lead to a referendum instituting change or may induce the parties in power to change the electoral system. Changes in the electoral climate can also contribute to popular dissatisfaction. However, in some combination, these factors can lead to and explain electoral system reform. 15

16 Chapter 3: Case Studies One of the best ways to understand how these factors contribute to electoral reform is to examine the cases in which major reforms have been attempted. Analyzing these case studies will demonstrate how these factors contributed to change in a particular setting as well as how they interacted. In other words, the cases will serve to ground the theoretical factors explained above. Three cases will be examined here (in addition to the extended analysis of the British case to follow). Australia and New Zealand are both cases in which the electoral system was successfully reformed. In contrast, Canada and the UK are both places where the electoral system has been criticized and changes proposed, but reform has (so far) not succeeded. There are several reasons these similar cases were chosen for further analysis. First, these are cases of reform that occurred (or did not occur) in established democracies. Second, each country has a common heritage: Australia, New Zealand, and Canada were all once British colonies and remain part of the Commonwealth, with Queen Elizabeth II as the head of state. As far as is possible then, the culture and history of each country is similar, and for the most part, these explanations can be ruled out in terms of reasons for reform (or the lack thereof). Australia Australia is the oldest example of reform included here as the major reforms to the national legislature date back to Established in 1901 with independence, the Australian political system was modeled on the Westminster Parliament, departing mainly in establishing a federal system and a directly elected upper house, the Senate (Farrell and McAllister 2005, 80). 16

17 Between 1903 and 1918, elections to the lower house, the House of Representatives, were conducted using the single-member plurality (SMP) rules inherited from Great Britain. Since 1918, Australia has been one of the few countries to use the majoritarian system of the alternative vote (AV) for elections to the House of Representatives (Farrell and McAllister 2005, 79). Under AV, voters rank the candidates on the ballot in single-member districts. In Australia, voting is compulsory and voters must rank all of the candidates on the ballot for it to be considered valid (Farrell and McAllister 2005, 84-86; Wright 1984, 129). If one candidate receives a majority (50% plus one vote) he or she is elected. However, if no single candidate secures a majority of the vote, the candidate with the fewest votes is dropped and his/her votes redistributed to the second choice listed on those ballots. This process of dropping the lowest vote-earner and redistributing his/her votes based on the voters preferences continues until one candidate has a majority. This candidate is then the winner. There is only one round of voting, but because the candidates are rank ordered, there may be multiple rounds of counting. Australia is also notable for its use of the single transferable vote (STV) for elections to the Senate. While this analysis will focus on the House of Representatives, it is important to understand how STV works. For the voter, STV, or PR-STV as it is called in the UK, works exactly the same way as AV with the exception that instead of single-member districts, STV uses multimember districts (Gallagher and Mitchell 2005b, 593). However, there is a certain number of votes, called the Droop quota, which will always be sufficient for election. This is important because, during counting, any votes a candidate has over and above this are deemed surplus votes, which would be wasted if they were left locked up in the possession of the candidate for whom they were cast (Gallagher and Mitchell 2005b, 594). Therefore, if candidate A s share of the vote exceeds the Droop quota, the surplus votes are transferred to other candidates, in 17

18 accordance with the next preferences marked by the voters (Gallagher and Mitchell 2005b, 594). STV also works like AV in that the candidate with the fewest votes will be dropped from the running and his votes redistributed. How did Australia come to adopt AV for the House of Representatives? Prior to federation in 1901, four of the six colonial parliaments used single-member plurality (SMP); of the remaining two parliaments Queensland used a form of AV (that was referred to then as the contingent vote ) and Tasmania, STV (where it was referred to as Hare-Clark ) (Farrell and McAllister 2005, 82). It was because of this diversity that the first federal election of 1901 was conducted under various voting methods determined by each state. The Commonwealth Electoral Act of 1902 established the use of SMP for federal House of Representatives elections for every state despite some efforts to institute AV for the House and STV for the Senate at that time (Farrell and McAllister 2005, 82-83). According to Farrell and McAllister, the Labour Party was best situated to benefit from the disproportional SMP system because it was better disciplined than other parties (2005, 83). Therefore, it was not until 1917 that a non-labour government was elected. This Nationalist government introduced the Commonwealth Electoral Bill in October of 1918, which passed after little debate in the legislature and established the use of the alternative vote. The Senate has undergone more changes to its electoral system over the years. Since this chamber is not the focus of this analysis, suffice to say that STV was finally adopted in The current system of electing Senators is STV with ticket voting 2 and was adopted in In Australia, several of the theoretical factors explained above contributed to the success of electoral reform. First, there were external pressures for change in the form of debates 2 On the ballot, voters can vote for a party ticket (Farrell and McAllister 2005, 84). 18

19 occurring in Great Britain at the time about preferential voting 3 (of which AV is one variant) (Farrell and McAllister 2003, 288; Farrell and McAllister 2005, 82). A second factor was Australia s history of using a variety of electoral systems for national and state elections (Farrell and McAllister 2003, 288). SMP, AV, the block vote, and the single transferrable vote (STV) have all been used at various times in different locations (Farrell and McAllister 2005, 82). This experience with other electoral systems helped foster a climate suitable for change. Third, despite my attempt to keep political culture constant across each case study, scholars argue the political culture did play a role in the reform of Australia s electoral system. Australia absorbed utilitarian ideas as its core culture as opposed to the more individualistic Lockean view of culture that took root in North America (Farrell and McAllister 2005, 95). This political culture can be seen as an example of popular dissatisfaction with the political system. People both within the legislature and on the streets were dissatisfied with SMP. Fourth, and perhaps most importantly, there was in 1918 a political party (the Nationalists) in power that wished to change the system, which had disadvantaged it in the past. It was this factor that led to actual electoral reform rather than ongoing debates and discussions as in other countries (such as the UK). Finally, what I have called short-term or trigger factors played a role. Activists and individual actors played an important part in Australian reform from as early as 1902 (Farrell and McAllister 2003, 288). New Zealand New Zealand used a system of single-member plurality (SMP) for elections to Parliament (a unicameral body since 1951) between 1914 and 1993 (Gallagher 1998, 205; Vowles 2005, 3 Preferential voting refers to the ability of voters to rank-order candidates (Farrell and McAllister 2005, 79). 19

20 296). In fact, New Zealand has long been considered a prime example of the Westminster [or majoritarian] model (Gallagher 1998, 205; Vowles 2005, 205). However, a mixed-member proportional (MMP) system similar to that used in Germany was adopted in Under MMP, each voter has two votes one vote for the party (closed lists) and one electorate vote (simple plurality). New Zealand elects sixty-nine MPs using plurality and fifty-one using the party lists (Vowles 2005, 297). In order to secure seats, parties must either win at least one single-member district or five percent of the party vote (Vowles 2005, 297). Exactly how did New Zealand move from a prime example of a majoritarian system to a mixed electoral system? Reform was discussed for years prior to In the 1970s, a Select Committee on Electoral Law was established by Parliament. The formation of this committee did not directly lead to change, but it did begin to bring the issue of electoral reform to the fore of the national consciousness (Farrell 2001, 37). Then, in both 1978 and 1981, the SMP system resulted in an incorrect outcome the Labour Party won a plurality of the vote, but the National Party secured a majority of the seats in Parliament (Katz 2005, 65; Nagel 2004, 532). This further highlighted the issues associated with SMP. Partly because of these incorrect election outcomes, Labour included a promise to establish a Royal Commission to conduct an authoritative and exhaustive reappraisal of electoral law (Nagel 2004, 532). The other reason this promise was included was that Geoffrey Palmer, who would become Deputy Prime Minster, was an electoral reform zealot (Katz 2005, 65; Vowles 2005, 296). In 1984, when Labour came to power, it created the apolitical Royal Commission, which suggested that the best alternative for New Zealand was MMP if the people agreed in a referendum (Farrell 2001, 37; Katz 2005, 296; Nagel 2004, 533; Vowles 2005, 297). However, since MMP was likely to produce coalition governments, a majority of political elites never supported this alternative 20

21 (Nagel 2004, 533). After the Commission s report, a grassroots movement called the Electoral Reform Coalition was able to maintain focus on the issue of reform. A minority of members in several parties agreed with the Electoral Reform Coalition that the electoral system should be changed and this combined call for reform became nearly impossible for political elites to ignore (Nagel 2004, 534). During the 1987 election campaign, Labour Prime Minister David Lange made a critical mistake. During a televised leaders debate, he promised to hold a binding referendum on reform of the electoral system despite the fact that his party had not authorized this commitment and strongly objected to holding a referendum (Katz 2005, 66; Nagel 2004, 534). Labour won the election, but had no intention of following through with Lange s rash promise. Interestingly, the opposition National Party was even more opposed to electoral reform than Labour, but could not resist using the broken promise to embarrass the government (Nagel 2004, 534). Thereafter, both parties included promises of referenda in their election manifestos (Nagel 2004, 534). An advisory referendum on electoral reform was held in In this referendum, 45% of the electorate participated, 85% of voters favored a change in the electoral system, and 71% preferred MMP to other alternatives (Katz 2005, 66-67; Nagel 2004, 535). This advisory referendum was followed by a binding referendum in 1993 in which the people voted 54% to 46% percent to change the electoral system to MMP (Katz 2005, 66; Nagel 2004, 535). A variety of factors contributed to the implementation of MMP in New Zealand. First, popular dissatisfaction with the political process was elevated and anti-politician sentiment was particularly high in the 1980s (Vowles 2005, 302). Additionally, the SMP system was returning increasingly disproportionate results (Vowles 2005, 302; Nagel 2004). The perceived unfairness that had led Labour to support reform after winning a plurality of votes, but failing to secure a 21

22 majority or plurality of the seats, contributed to this public dissatisfaction. The political culture that placed a great deal of emphasis on fairness and keeping manifesto promises contributed to the dissatisfaction with SMP and, in this way, to electoral change (Nagel 2004, 536). A second, related factor was change in the electoral arena with the rise of third parties that were treated particularly harshly by the SMP system (Gallagher 1998, ). This contributed to the sense of unfairness, but was also a factor in and of itself. Third, there were several short-term factors that came together to create an atmosphere for reform in the early 1990s. The social movements for reform combined with individual leaders to make reform possible. The Deputy Prime Minster, Geoffrey Palmer, was an active proponent of reform and in a position to further that goal. However, in other cases, particularly that of Prime Minster David Lange, it was a slip of the tongue or rash promise that led him to further the reform effort. The final factor that led to reform was the ability of the voters to force the hand of politicians in the form of two referenda. Politicians mistakenly believed that, if the question were put to the people in a referendum, the people would choose to keep SMP. In this way, electoral change occurred in New Zealand against the wishes of the parties in power. Canada Canada is our first example of a case in which electoral reform has not succeeded at the national level at least not yet. From the start, Canada has used single-member plurality (SMP) to elect representatives to the lower house (House of Commons) of the national legislature (the Senate is an appointed body) as well as to elect members of the unicameral legislatures of each province (Massicotte 2005, 99, 102). However, despite the long history of first-past-the-post, the electoral system has been challenged over the years, and some provinces have experimented with 22

23 other systems for election to the provincial legislatures (Massicotte 2005, 102). In recent years, the topic of electoral reform has garnered even more attention at both the provincial and national levels (Henderson 2006, 41; Kanji and Bilodeau 2006, 829; Massicotte 2005, 99). The main challenges to SMP began in the 1920s. It was then that the alternative vote (AV) was advocated for House of Commons elections, but the change was not adopted (Farrell 2001, 38; Massicotte 2005, 102). At the national level, reformers have historically focused on the issue of malapportionment (i.e. cases in which different regions have widely different ratios of electors) rather than on changing the electoral system itself. Regardless, the topic has come up over the years. In 1979, the Pépin-Robart Task Force on Canadian Unity recommended using PR to elect a portion of MPs (Farrell 2001, 38; Massicotte 2005, 112). The group Fair Votes Canada was organized in 2001 and by 2004, the Law Commission of Canada had advocated a mixedmember proportional system (Massicotte 2005, 112). In 2003, the government defeated a motion in the House of Commons that would have provided for a referendum on PR (Massicotte 2005, 115). Most recently, in 2004 the New Democratic Party pushed the minority Liberal government to consider electoral reform, but the plan was abandoned (Pilet and Boi 2011, 574). In Canada, the greatest leaps and experiments in the area of electoral reform have occurred at the provincial level (Farrell 2001, 38). In 1920, Manitoba introduced the single transferable vote for electing some members. By 1927, the remaining members were elected using AV and this dual electoral system continued until 1957 when they switched back to SMP (Massicotte 2005, 103). Between 1926 and 1955, the province of Alberta utilized a similar combination of STV and AV; British Columbia also experimented with AV around this time (Massicotte 2005, 103). Finally, there was a referendum in British Columbia in 2005 on whether the province should adopt STV for legislative elections. A majority (57%) of the voters favored 23

24 the change, but this was less than the 60% necessary for the referendum to be binding (Henderson 2006, 42; Massicotte 2005, ; Pilet and Boi 2011, 574). In the case of Canada, many of the factors that scholars have argued are crucial for reform or that were present in the successful cases of Australia and New Zealand have simply not been present. There have been no major scandals or government failures leading to popular dissatisfaction with the electoral system. The rise of new parties does not seem to have translated to more pressure for reform and there do not appear to be external pressures for change. When the voters have been given a choice (in British Columbia), they have supported reform, but not in numbers substantial enough to force the hand of the members of Parliament (i.e. the legislature set a 60% threshold for the referendum to be binding and this requirement was not met). However, there is substantial support for change from reform movements and pressure groups, which keep the issue in the national consciousness and, in combination with other factors, could lead to change in the future. Perhaps most importantly, the Canadian political elite has not paid much attention to the issue of the electoral system, although this may well be changing (Farrell 2001, 39). Declining voter turnout is making it increasingly difficult [even for parties that benefit from the system] to simply sideline this issue (Kanji and Bilodeau 2006, 829). The former Prime Minister, Paul Martin, has at times, advocated various elements of electoral reform (Henderson 2006, 42). This could be an indication that support for reform is growing among the political elite, which would increase the likelihood of reform. In 2004, Jack Layton, leader of the New Democratic Party (NDP), made PR a precondition for NDP support for a future minority administration (Massicotte 2005, ). In the end, the NDP was not able to force this issue, as it was not the only party on which the minority Liberal government relied (Henderson 2006, 41; Massicotte 24

25 2005, 112). In future, the NDP could be in a position to press the issue, making reform much more likely. So far, the factors identified as important by scholars have not converged in such a way as to make reform of the electoral system in Canada a reality. This does not mean that these variables are unimportant. Rather, that these factors were not present and reform has not happened may lend credence to the argument that they are critical. Despite the history of failed attempts in Canada, the stage is set for potential reform in the future. 25

26 Chapter 4: A Theory of Electoral System Reform The literature review (Chapter 2) laid out five major factors that scholars have identified as important to understanding electoral reform. Those five factors were (1) the desire of parties in power to change the system, (2) popular dissatisfaction with the political process, (3) change in the electoral climate, (4) external pressures for change, and (5) the ability of the voters to choose the electoral system against the wishes of the parties in power. In the case studies of Australia, New Zealand, and Canada (Chapter 3), we have seen how these factors (and others) played a role in several cases of electoral reform. In this chapter, I will advance a theory of electoral reform based on the extant literature and these case studies. According to this model, reform is a two-part process. The first component consists of putting electoral reform on the agenda. This could be the legislative agenda of a government or, more generally, the national agenda. It is this part of the electoral reform process, and specifically legislative agenda making, that scholars have tended to focus on. But the second component of the theory is equally important what factors contribute to the success of reform efforts? This part of the theory will rely on the scholarly literature, but will also draw more heavily from the examples of reform in the case studies. In explaining these two parts, the variety of factors that interact and combine to affect the process of reform will be explored. It is important to note that, although this theory may well prove adequate to describe electoral reform generally, it is not designed to explain all cases of reform. The theory is developed based on the literature and the experience of several Commonwealth nations and is used here as a framework to explore the British case (which will also serve as a test of the model). In order to generalize the theory beyond the Commonwealth states discussed here, it 26

27 would need to be tested against all (or at least many more) cases of electoral reform. While this systematic testing would be a worthwhile venture in the future, it is critical to recognize that it is not attempted here. Further, this theory is based solely on examples of non-violent change. In other words, in all the cases discussed here, actors work within the current system to create change. Therefore, even if the model should eventually be shown to be more widely applicable, it may still only pertain to cases of non-violent change. Part 1: Putting Reform on the Agenda On the most basic level, there are two ways to make something an agenda item. The first is the most straightforward elected political elites can put reform on the agenda. In other words, elites in parliament can propose legislation that would change the electoral system if passed. The second method is by referendum. Depending on the laws of the country in question, this could conceivably be done in one of two ways. The people could initiate a referendum (if the law allows) and, in effect, force a referendum that the legislature would prefer not take place. Alternatively, the elected political elites could pass legislation that would provide for a referendum to be put to the people, as in New Zealand. These two, potentially related, direct processes of agenda setting are illustrated in Figure 4.1. Figure 4.1. How electoral reform is elevated to the national agenda. 27

28 Now that the two direct ways of agenda setting are clear, what factors influence this part of the electoral reform process? These are factors that have a direct effect on elected political elites, the referendum process, or both, and in this way indirectly affect the agenda. Figure 4.2 shows the process of putting electoral reform on the agenda, including the variety of long- and short-term factors that indirectly influence the agenda. Figure 4.2. Complete process by which electoral reform becomes an agenda item. The top half of Figure 4.2 shows the three factors or categories of factors that influence the agenda only indirectly through elected political elites. These factors do not directly affect the referendum process, but may do so indirectly if the referendum is arranged by elites. The first, interest judgments by the majority party or parties, can be either long- or short-term. As explained in Chapter 2, a significant portion of the literature on electoral reform attempts to explain why parties that have been successful under one electoral system would want to change it (see Chapter 2). This general category of interest judgments is meant to incorporate the four 28

29 reasons incumbents may want to change the system as described in Chapter 2. Those reasons are: (1) incumbents may believe the existing system will not continue to benefit them and so believe it is in their interest to change the system, (2) the interests of all the members of the governing party or coalition of parties are not aligned (this may lead the party or coalition to adopt positions that would appear contrary to their interests if they were considered a unitary actor), (3) the party in power may value long-term change in the competitive system over short-term electoral advantage, and (4) the majority parties may trade electoral advantage for other goals (Katz 2005, 63). In short, the interest judgments of the majority parties may induce them to make changing the electoral system a priority. A second factor that influences this first part of the electoral reform process through political elites is the behavior of the party leaders. This factor is short-term and comes from the lessons of New Zealand. The promise (or gaffe), unsanctioned by his party, by the Labour Prime Minster David Lange during the election campaign of 1987 to hold a referendum on reforming the electoral system was critical in the development of that reform. Without his promise, the elected elites in New Zealand, who for the most part did not favor change, may have successfully resisted calls for reform. In other cases, a leader may conceivably be able to quash a change advocated by some party members or, as in the case of Lange, lead their party towards undesired change. Therefore leaders can be an important factor in agenda setting. Finally, external influences can contribute to the decision of elected political elites to put electoral reform on the national agenda. This is a long-term factor that is discussed in the literature and highlighted by the case of Australia. Generally, external shocks to the system such as colonial rulers (Benoit 2004) of a given country are emphasized by scholars, but there can also be much more subtle influences for change. In Australia, people were influenced by 29

30 concurrent debates taking place in Britain about the possibility of adopting preferential voting for Parliamentary elections there (Farrell and McAllister 2003 and 2005). On the left side of Figure 4.2, the factors that can influence either elected political elites or a referendum are shown. The long-term factors of party system changes and popular dissatisfaction are depicted alongside the short-term factors of crises, pressure groups, and the media. As illustrated in the figure, these factors can interact with one another in addition to directly affecting elites and referenda. The first long-term factor is changes in the party system. Over time, even when the electoral system remains constant, the party system may change. In particular, the rise of new parties or previously (relatively) unimportant parties may influence political elites to change the system either in an attempt to maintain some of their previous influence despite the changes or because new political elites are now influential and able to force the issue. Alternatively, changes in the party system may induce voters to attempt to change the electoral system through a referendum because these party system changes can increase dissatisfaction with the existing system. As new parties become more important, often the disproportionality of the existing system is highlighted, making the public more amenable to change. A related, but also independent, factor is popular dissatisfaction. If people are unhappy with the current electoral system or blame the electoral system for other national problems, the elected politicians may be forced to contemplate changing the system or risk losing the next election. Alternatively, popular dissatisfaction could contribute to a referendum initiated by the people (if this is legally a possibility). In addition, three short-term factors contribute to elevating electoral reform to the agenda through both elected political elites and/or referenda. First, national crises could contribute to a 30

31 desire to change the system. These crises may or may not be a direct result of the existing electoral system, but can nonetheless affect whether reform is an issue that is put on the agenda. A crisis, depending on its nature, could cause reform of the system to become a priority or it could push other issues to the fore so that electoral reform does not become part of the national agenda. Crises create calls for reform if the public blames the existing electoral system for the crisis, for example, if they see corruption as stemming from the electoral arrangement. In this way, crises have the potential to cause a loss of support for the current system. Second, pressure groups such as the Electoral Reform Coalition in New Zealand may lobby for electoral reform. They may approach lawmakers directly and in this way influence the agenda or may influence the public through their publications and other activities and in this way impact the agenda. Finally, the media may influence elected elites and referenda, but only indirectly through popular dissatisfaction. The media is only an indirect influence because the people must accept the arguments of the media for anything to be done. This brings us to the interactions of these factors. As stated, the media is the only factor that does not directly affect elites or referenda, but does so through popular dissatisfaction. Party system changes can directly affect the process, but they can also contribute to popular dissatisfaction, so these factors are not completely independent. For example, changes in the party system may serve to highlight the disproportionality of a plurality system and in this way contribute to popular dissatisfaction. Additionally, crises can contribute to popular dissatisfaction with the political system in general and the electoral system specifically and in this way indirectly affect whether reform becomes an agenda item. Clearly, there are multiple factors affecting whether or not electoral reform becomes an agenda item. Some of these factors can be either positive or negative (meaning that, depending 31

32 on the situation, they could make reform more or less likely). While all of the above factors may not be necessary for electoral reform, they will increase the likelihood of elevating reform on the national agenda. Therefore the conditions most likely to result in reform becoming a salient political issue involve each of these factors. Reform is probable if the parties in power deem reform is in their interest, if the leaders of said parties are supportive of reform, if there are external pressures for reform, if the party system changes and thereby makes the existing system problematic in the eyes of elites or the public, if there is substantial public dissatisfaction with the existing system (and if the media highlights this), if there is one or more crisis that people blame on the electoral system or political system more generally, and if there are pressure groups lobbying for reform. Part 2: Success of reform. So far, I have advanced a theory of how electoral reform becomes an agenda item. However, this does not satisfactorily explain electoral reform because agenda setting and successfully passing reform are two very different things. Up until now, scholars have not necessarily differentiated between these two components of reform (or have equated putting reform on the agenda with success because of the scholarly focus on elite actors), constituting a major weakness in the existing literature. Even if political elites wish to reform the system, that desire does not guarantee the goal of reform will be achieved. Therefore, part two of this model attempts to explain the factors that affect the success of electoral reform attempts. This part of the reform process is shown in Figure

33 Figure 4.3. Factors that influence the success of electoral reform attempts. The first thing that is striking about this part of the process is that the factors influencing the success of a reform attempt are much more interrelated than those in part one (agenda setting). Further, these items affect success of electoral reform both directly and indirectly. One thing that directly affects success is the strategy of the proponents of reform. This factor, labeled strategy in the figure, is meant to encompass the organization and execution of the pro-reform campaign. Interest groups, political parties, specific members of parliament, or some combination could direct this campaign, which may be more or less visible to the public. If the reform is attempted through a referendum, we would expect this campaign to be relatively public and more visible to the casual observer than a lobbying campaign to pass legislation in parliament. As shown in Figure 4.3, one factor that impacts success both directly and indirectly through strategy is which reform is proposed. There are several possibilities for significant reform to the electoral system. For example, if one wanted to change a plurality system, one 33

34 could attempt to institute AV, STV, list PR, or a mixed system, to name a few of the available options. Which reform is being attempted will have a huge impact on the strategy of the proponents. Importantly, the strategy of reformers could also impact which reform is attempted. For example, the reformers may prefer one type of reform, but for strategic considerations may propose an alternative that is not their first choice. This is why there is a double-headed arrow between which reform and strategy in the figure. However, which reform is chosen can also affect success directly. It is conceivable that a reform attempt could fail because the public objects to the option available to it, not because it rejects reform in principle. The remaining five factors can also influence success both directly and indirectly. First, the strength of the opposition can have an immense, direct impact on whether or not a reform attempt is successful. However, the opposition strength can also impact the strategy of the proreform campaigners and in that way indirectly influence the outcome. Included in the concept of opposition strength is opposition funding as well as the strategy and organization of the opposition campaign. Another related factor is the funding of the pro-reform movement (shown as funding in the figure). The amount of funding available may affect the strategy of the proreform campaign, but it could also directly affect the passage of reform. The final three factors that influence reform success are external to the campaigns, but are of course related. Other issues refers to other national issues ongoing at the time of an attempted electoral reform. These could be economic troubles such as a recession, or concerns such as an ongoing war. The presence of these issues will impact how the pro-reform campaign is executed as well as the success of reform, in that citizens assign more or less importance to electoral reform depending on the presence of these other concerns. Another factor influencing success both directly and/or indirectly is national experience with other electoral systems 34

35 ( experience with alternatives in Figure 4.3). If the nation is accustomed to the idea of using alternatives to elect representatives, either because they have used other systems for national legislative elections in the past or because they use other systems to elect officials at different levels of government, reform is more likely to be successful (as illustrated by the case of Australia). Further, the supporters of reform will almost certainly modify their strategy according to whether or not the nation is accustomed to alternatives. Finally, public opinion affects the success of reform movements. Again, the pro-reform campaign should tailor its strategy according to public opinion. However, the strategy of the pro-reform campaign also influences public opinion through its efforts and messages. Clearly, public opinion directly affects success as well. This is especially perceptible if the reform is put to a referendum, but even if political elites are attempting to change the system through the legislature, public opinion must be taken into account or reformers risk losing the next election. These factors, while they affect success both directly and indirectly through the strategy of the reform proponents, also interact with one another. Opposition strength can have an effect on funding (of the pro-reform movement), but funding can also affect the opposition s strength (again, note the double-headed arrow). Recall that opposition strength includes opposition funding and strategy. The campaigns will react to one another and if the opposition is especially strong, the pro-reformers will increase their funding if possible. The reverse is also true the opposition will react to the funding allocated to the pro-reform campaign. Similarly, opposition strength will both affect and be affected by public opinion. That opposition has an effect on public opinion through the opposition campaign is intuitive. But public opinion will also affect the strategy and funding of the opposition. Further, opposition strength may be affected by other national issues at the time (note that this is a unidirectional relationship). The strategy and 35

36 funding of the opposition will be very different if the country is in the midst of a recession than if there are no other major issues at the time of reform. Of course, other national issues do not just impact the strength of the opposition. In the same way that other issues affect opposition funding, proponent funding ( funding in Figure 4.3) will also be influenced. Once again, this is not a reciprocal relationship, as funding will not affect the other national issues that arise. However, these issues will both be affected by and influence public opinion. Other issues will affect public opinion about reform. Unless the public sees the electoral system as the cause of other issues, current problems like an economic recession may take precedence in the eyes of the public. Similarly, public opinion will affect other national issues in that it is public opinion that will (at least in part) dictate the relative importance of the issues facing the nation. Public opinion also impacts and is influenced by which reform is proposed. Proponents of reform may well take public opinion into account when deciding which reform to propose, but the specific system that is suggested will also affect how the public feels about electoral reform. The final factor is experience with alternatives, which is not affected by any of the other factors. However, it does influence both public opinion and which reform is proposed. Experience with other electoral systems is likely to make the public view the possibility of change more favorably. In this way, historical experience has an effect on public opinion. Finally, this experience affects which reform is proposed. Proponents of reform are likely to keep history in mind when proposing a reform. If there is a history of successfully using a particular system, this could strongly influence which reform is proposed in the present. As with part one of the model, not all of the factors explained in this chapter may be necessary for electoral reform to be successful. Additionally, some of the factors are positive and 36

37 some are negative. In short, reform is most likely to be successful when the strategy of the proreform campaign is effective, the reform being attempted is acceptable to the public, 4 the strength of the opposition to reform is particularly weak (either its strategy is poor, its funding is lacking, or both), the funding of the pro-reform campaign is strong, other national issues, if present, increase the desire for reform, there is substantial experience with other electoral systems, and public opinion is supportive of reform. Now that the two-step process of agenda setting and successfully passing a reform has been explained, this theory will be applied to the United Kingdom. The following chapters will test this theory using the case of the UK while exploring why electoral reform became a national agenda item in 2011 and why it failed. 4 When I write that the reform is acceptable to the public, I do not mean to conflate the reform variable and the public opinion or dissatisfaction variable. I mean that some reforms are more likely to pass and I am merely attempting to briefly describe this. 37

38 Chapter 5: History of the British Case The most recent attempt at electoral reform in the United Kingdom in 2011 cannot be properly understood without some understanding of the historical context. It can sometimes seem as if the electoral system in the UK has come down in its existing form through the ages and that change is rarely, if ever, contemplated outside reform society meetings and the academy. Is this true? As it turns out, electoral reform has been a more common topic of discussion in Parliament than one might expect. Likewise, the party system (with the possible exception of the past five to ten years) can, at times, appear very static. This too is a misperception. At one time, Britain 5 was a near-perfect example of a two-party system, but this has changed. It is important to understand the development of the party system because changes in the electoral climate (such as changes in the party system) can contribute to electoral reform. The Electoral System and its Discontents The plurality electoral system used in the UK originated in the Middle Ages (Hart 1992, 5; Mitchell 2005, 158). Interestingly, it was not until 1885 that single-member districts predominated and it was not until 1948 that single-member districts were used exclusively (Mitchell 2005, 158). Prior to 1948, some constituencies (and prior to 1885, most constituencies) had a district magnitude of two, three, or even four (ibid.). Over the years, particularly in the late nineteenth and early twentieth centuries, the issue of electoral reform came up in Parliament 5 Note that any time (such as this) I refer to the party system, I use the term Britain very specifically because Northern Ireland has, since 1970, had a very different party system from that in Great Britain (Kavanagh and Cowley 2010, 413). 38

39 rather frequently. 6 Many of these discussions centered on the issues of suffrage, redistribution, or the administration of voting (Butler 1953, 2). While these discussions and reforms were undoubtedly important, there are several dates that stand out for our purposes. Reform of the actual system of elections was discussed and nearly achieved between 1917 and 1918 and in The final critical year for electoral reform (prior to 2011) was Few people realize how close Parliament has come to electoral reform in the past century. In 1917, proportional representation was nearly adopted for elections (in some constituencies) to the House of Commons. The Wartime Coalition Government composed of the Conservatives, Liberals, and Labour set up a Speaker s Conference in 1916 that was to consider matters of registration, franchise extension, and the electoral system (Mitchell 2005, 159; see also Butler 1953, 7; Hart 1992, 181). In 1917, this group recommended a combination of the single transferable vote (PR-STV), the limited vote, and the alternative vote (AV) (Hart 1992, 181; Mitchell 2005, 159). The government then introduced legislation in 1917 that included the recommendations of the Conference with regards to the electoral system (Hart 1992, 185). However, the government did not fully support proportional representation and the sections of the bill regarding the electoral system were allowed a free vote in the Commons (i.e. members were free to vote as they saw fit rather than instructed to vote with the party) (Hart 1992, 183). Between 1917 and 1918, there were sixteen debates in the House of Commons about electoral reform (Hart 1992, 199). The Bill bounced back and forth between the Commons and the Lords several times. Each time, the Commons voted against PR, but the Lords voted to reinstate it (Hart 1992, 185). The Commons voted on PR six times between 1917 and 1918 and rejected it 6 For a very detailed examination of the history of electoral reform between 1828 and 1951, please see Butler (1953) and Hart (1992). 39

40 each time; however, on the first vote PR was defeated by only eight votes 7 (Hart 1992, 290; Mitchell 2005, 159). At one point in 1918, the Commons narrowly supported the alternative vote, but the Lords insisted on PR (Butler 1984, 225). Eventually a compromise between the Lords and Commons was reached and the 1918 Representation of the People Act provided for proportional representation in some university seats (meaning constituencies representing a university), but did not result in substantial changes to the electoral system overall (Hart 1992, 185). Amazingly, the Commons came within eight votes of adopting a mixed system of proportional representation and the alternative vote. Had the Commons passed the Bill with PR and AV, the Lords would have accepted it. In light of this, it is far from unprecedented for Parliament to consider adopting proportional representation. The next critically important year in the history of electoral reform in Britain was In that year, the Labour government introduced a bill called the Representation of the People (No. 2) Bill. This Bill provided for the use of the alternative vote (AV) for future elections to the House of Commons (Hart 1992, 240). In June of 1931 the Bill was passed in the Commons by a vote of 278 to 228 (Hart 1992, 241). Thereafter, the Bill went to the House of Lords where it was passed by 50 votes to 14 (Hart 1992, 242). Finally, the Bill returned to the Commons in July 1931 with the provision specifying AV still included. However, the Labour government resigned under pressure in August and the Bill was dropped (Hart 1992, 243). We can, of course, never know what might have happened had the Labour government held on for several more months or a year. Regardless, 1931 is an important milestone for electoral reform in that both Houses of Parliament passed a bill stipulating that future elections to Westminster would be conducted using the alternative vote. It is also important to note that this was occurring as the Great 7 Actually, these sources disagree. Mitchell reports that the motion was defeated by seven votes, while Hart reports the eight-vote difference. 40

41 Depression was hitting Britain. Conventional wisdom would suggest that non-essential legislative agenda items like tinkering with the electoral system should not occur during times of crisis. The lesson of the 1931 reform attempt suggests that this is not always true. Political elites may try to change the electoral system even in times of economic crisis. In terms of electoral reform, 1997 may well be the most important milestone. With few exceptions, notably Northern Ireland, prior to 1997 all elections in the UK were conducted using SMP (Mitchell 2005, 157). However, following a veritable burst of devolution and institutional engineering since 1997, the UK now uses a formidable array of different electoral systems (ibid.). Of course, members of the House of Commons are still elected using SMP, but the Welsh and Scottish Parliaments and London Assembly use mixed-member systems, the Northern Ireland Assembly uses PR-STV as do Scottish local government elections, and British members of the European Parliament (MEPs) are elected using closed-list PR (Mitchell 2005, 157). The origin of these changes can be traced back to the 1970s and 1980s, when after a series of defeats the Labour Party appeared to be in jeopardy of never regaining power (Dunleavy and Margetts 2001, 300). It was in the late 1970s that the Scottish Constitutional Convention (SCC) was formed and this body came to be dominated by Labour and the Liberal Democrats (ibid.). However, they did not agree on the way in which the proposed devolved Scottish Parliament should be elected. The compromise devised in the late 1970s was a mixed-member system that influenced the developments in Wales and, eventually, these mixed systems would be adopted (Dunleavy and Margetts 2001, 301). How exactly did the 1997 changes come about? In March 1997, prior to the general election of that year, the Liberal Democrat and Labour Parties agreed on a joint statement on constitutional issues. This concordat included plans for devolution for Scotland and Wales and a 41

42 new London Assembly, all of which would be elected using mixed systems (Dunleavy and Margetts 2001, 301; Mitchell 2005, 175). The agreement further explained plans for the adoption of list-pr for the election of MEPs and promised a referendum on the Westminster electoral system with the alternative to SMP to be decided by a commission (ibid.). In the event, the Labour government (elected in 1997) enacted all of these proposals (devolution was confirmed by referenda in Scotland and Wales) between 1997 and 1999 with the exception of the referendum on the national electoral system (Dunleavy and Margetts 2001, 302; Mitchell 2005, 175). The government did, however, set up the Jenkins Commission, an independent commission chaired by Lord Jenkins to investigate the Westminster electoral system, in December of 1997 (Dunleavy and Margetts 2001, 303; Mitchell 2005, 177). The Jenkins Commission was given two sets of criteria Labour wanted an electoral system that maintained the constituency link (the connection between a member of parliament and the constituency for which he/she is responsible) and ensured stable government, while the Liberal Democrats wanted proportionality and an extension of voter choice (Dunleavy and Margetts 2001, 303). The Commission recommended a mixed system in October 1998 that it called AV-plus (AV+). This is an additional member system where a large proportion of members are elected using AV rules and then there are some top-up seats elected using proportional representation to ensure some measure of proportionality (Dunleavy and Margetts 2001, 303; Mitchell 2005, 178). However, this proposal met with resistance from the Labour cabinet and the issue was delayed, partly because elections were expected in 2001 (Dunleavy and Margetts 2001, 303). This historical context is crucial to understanding the 2011 reform attempt. Clearly, reform has been discussed and very nearly enacted multiple times over the last century. As I have argued, precedent (or experience with alternative systems) can be an important factor in the 42

43 success of reform attempts. The reforms of made SMP the exception rather than the rule in terms of electoral systems in use in the UK (Mitchell 2005, 174). Additionally, as Dunleavy and Margetts point out, the profusion of new electoral systems since 1997: reflects essentially the impact of devolution in creating new electoral bodies in Scotland, Wales, and London. But the shift to fully proportional systems for these bodies (and for the UK s European Parliament elections) is both caused by and has consequential changes for longer-term alterations in British politics towards a multi-party system (2001, 296). As I have argued, changes in the electoral climate can lead to attempts to reform the electoral system, particularly if these changes result in a multi-party system within the framework of a plurality electoral system. How the party system has changed over the last half century will be discussed in Chapter 6. However, it is worth briefly examining the origins of the party system here. The Origins of the British Party System At the turn of the twentieth century, the two major parties in Britain were the Conservatives and Liberals. The Conservative and Liberal Parties had their origins in the Tories and Whigs of the eighteenth and early nineteenth century (Driver 2011, 33). While the Tories and Whigs were less political parties in the modern sense of the word and closer to factions, they were the precursor to the modern parties (Driver 2011, 33). The Liberal and Conservative parties developed in the later half of the nineteenth century as the franchise was extended and the Reform Acts of 1832 and 1867 took hold (Cook 2010, 2; Driver 2011, 33). The Labour Party was, in fact, the latecomer to the British political scene. Officially formed in 1900, the Labour Party is rooted in the socialist activity of the nineteenth century and the growing trade unions (Driver 2011, 90). Between 1914 and the 1920s there were three fairly 43

44 strong parties coexisting in the British system (Duverger 1954, 208; Humes 1990, 229). At the 1922 general election, Labour overtook the Liberals as the primary opposition to the Conservatives and the main center-left party (Cook 2010, 85; Curtice 2010, 626; Driver 2011, 113). After it was replaced by Labour as one of the two main parties, the Liberal Party struggled and came close to disappearing in the 1950s (Curtice 2010, 626). Now that the near extinction of the Liberal party is understood, the rise of the Liberal Democrats out of the ashes should be briefly explored. The Liberal Democrats were originally formed as an alliance between the Liberals and the Social Democratic Party (SDP). The SDP was formed in 1981 when four Labour MPs (the Gang of Four) left Labour because of recent moves to the left (Driver 2011, 115-6). The Liberals and SDP formed an alliance for the 1983 and 1987 elections before finally merging in 1988 into the Social and Liberal Democratic Party that was renamed the Liberal Democrats in 1989 (Driver 2011, 117-8). The final piece of the puzzle is the nationalist parties. The nationalist parties (Plaid Cymru in Wales and the Scottish National Party or SNP in Scotland) emerged in the 1960 s due to ethno-nationalist cleavages (Driver 2011, 175). Without going into any more detail, it is now clear how the two-party (Conservative and Liberal) system of the late nineteenth and early twentieth century developed into the system seen today. But how did this more fragmented party system affect the most recent general election? Five Days in May: The 2010 General Election The general election of 6 May 2010 was a critically important development for the advance of electoral reform in the UK. It resulted in the first hung parliament (no single party won a majority of the seats) since the 1970s and the first coalition government since World War 44

45 II (the first peacetime coalition since the 1930s). It also marked the first time the Liberal Democrats have been in power outright (as opposed to supporting a minority government) since they were formed from the Liberals and SDP in the 1980s. Clearly, this was a remarkable election in a number of ways. So what happened in the wake of this critical election? The general election (which took place on a Thursday) returned a hung parliament in which the Conservatives secured the largest share of the vote and 306 seats (this figure does not include the Speaker, John Bercow), Labour won 258 seats, and the Liberal Democrats, although increasing their share of the vote by 1%, lost seats to a total of 57 (Norris 2010; see Table 5.1). A majority government could only be formed by a coalition of the Conservatives and Liberal Democrats or a rainbow coalition incorporating Labour, the Lib Dems, and a variety of other smaller parties. In other words, based solely on parliamentary arithmetic, a Conservative-Lib Dem coalition would be far easier to form (generally speaking, the more parties included in a coalition, the harder it is to come to a satisfactory agreement) and have a greater majority. This election result set off a harried five days of bargaining among the three main parties. 45

46 Table General Election Results. Party Parliamentary Votes Vote Percentage Seats Conservative ,726, % Labour 258 8,609, % Liberal Democrats 57 6,836, % Democratic Unionist 8 168, % Party* (DUP) Scottish National Party (SNP) 6 491, % or 19.9% in Scotland Sinn Fein* 5 171, % Plaid Cymru 3 165, % or 11.3% in Wales Social Democratic and 3 110, % Labour Party* (SDLP) Green 1 285, % Alliance Party* 1 42, % Others 1 469, % UKIP 0 919, % BNP 0 564, % Notes: Not all parties are listed. Total turnout in the election was 29,687,409 voters or 65.1% of the electorate. For a technical majority in the Commons, 326 seats are needed. However the Speaker and three Deputy Speakers vote only to break ties and Sinn Fein MPs do not take their seats. *Northern Ireland Parties + This figure includes the Speaker Sources: BBC 2010; Bogdanor 2011; Kavanagh and Cowley After the election, there were essentially four options for forming a government. The first was a Conservative minority government that would control 307 out of 650 seats in the House of Commons (although this figure includes the Speaker, so in reality it would control 306 votes). The second was a Conservative-Liberal Democrat coalition, which would control a majority of 364 seats. Third, there was an option to have what some have called a rainbow coalition (Russell 2010). Andrew Russell explains that this coalition would likely have been made up of Labour, Liberal Democrats, Scots and Welsh Nationalists (the Scottish National Party and Plaid Cymru) and the lone Green MP (2010, 511). As explained here, this coalition would 46

47 control 325 seats exactly half of the chamber, but not a technical majority. Finally, there was the option of the traffic-light coalition that would include Labour, Lib Dems and the lone Green MP, but would lack a majority (Cook 2010). All of these options presented difficulties. Although a hung parliament was not wholly unexpected, the parties varied greatly in their preparations for such an event. The Liberal Democrats were perhaps best prepared. Their negotiating team of Danny Alexander, Chris Huhne, Andrew Stunell, and David Laws was formed in the fall of 2009 to begin preparations (Laws 2010, 14; Wilson 2010, 38). The Conservatives formed a negotiating team and began their preparations in earnest about two weeks prior to the election (Wilson 2010, 52). Their team consisted of George Osborne, Oliver Letwin, Ed Llewellyn, and William Hague (Wilson 2010, 106). The Labour team seems to have been by far the least prepared for these negotiations. Indeed, one member of the team, Ed Balls, was not aware he was to be part of the discussions until several hours before the first informal meeting between Labour and the Lib Dems that occurred on Saturday, 8 May (Wilson 2010, 32). The final Labour team was comprised of Peter Mandelson, Andrew Adonis, Harriet Harman, Ed Balls, and Ed Miliband (Laws 2010, 83; Wilson 2010). While the Liberal Democrats and Conservatives both entered the negotiations with fairly detailed policy statements and goals, Labour did not enter the initial phases with a document, a paper, an agreement, [or] a negotiating position, according to Mandelson (Wilson 2010, 32). Prior to the election, Nick Clegg, Liberal Democrat leader, had publicly stated that in the event of a hung parliament he would first negotiate with the party receiving the most votes and seats the party with the moral mandate to govern (Wilson 2010, 71). True to his word, Clegg indicated this priority again on the morning after the election, 7 May (Laws 2010, 40; Wilson 2010, 71). This was followed on the same day by David Cameron s (the Conservative leader) 47

48 big, open and comprehensive offer to the Liberal Democrats (Cameron 2010). Although in this speech Cameron never used the term coalition, it was clear that he was offering a full coalition with the Lib Dems. The Conservative and Liberal Democrat teams met that Friday night in what was described by those involved as a preparation for future meetings (Wilson 2010, 105). Accounts of all the negotiations between the Conservatives and Liberal Democrats are positive both parties evidently came to the table earnestly trying to form a coalition and willing to compromise in order to make that possible (Laws 2010; Wilson 2010). This is contrary to meetings, both informal and formal, between the Labour and Lib Dem negotiating teams in which one or both groups often felt frustrated, unsatisfied, or unclear as to whether the other wanted a coalition at all (Laws 2010; Wilson 2010). Over 6-9 May, the Liberal Democrats met formally and publically with the Conservatives while meeting secretly to have informal talks with Labour. Formal discussions between the Lib Dems and Labour did not begin until 10 May, after Gordon Brown had announced his intention to resign (Laws 2010, ). Even now, it is not entirely clear what a coalition between Labour and the Lib Dems would have looked like. According to Andrew Russell, it would have consisted of Labour, the Lib Dems, the Scottish National Party, Plaid Cymru, and the single Green MP, but it is unclear on what he bases this description (Russell 2010, 511). David Laws of the Lib Dems reports that the likely outcome would have been a minority coalition between Labour and the Lib Dems with a confidence and supply arrangement 8 with the Democratic Unionists, supplemented by further understandings with up to five other groupings of MPs or individual MPs (Laws 2010, 92-93). Since Laws was at the negotiating table, it seems likely 8 A confidence and supply agreement is an arrangement by which the party outside the government would agree not to vote against it [the government] in confidence or supply motions, in exchange for policy concessions and a promise of consultation on legislation [it is] a parliamentary but not a governmental coalition (Bogdanor 2011, 10). 48

49 that this account is more accurate. Laws also reports that Labour negotiators had not spoken with Plaid Cymru or the SNP as of 11 May (2010, 167). Wilson supports Laws account when he reports that on Saturday, 8 May, Labour negotiators indicated they had spoken with the Democratic Unionist Party and that they believed a rainbow coalition was feasible (Wilson 2010, 144). However, DUP MP Sammy Wilson denies that Labour members ever spoke with the DUP members during this time (Wilson 2010, 144). These different accounts illustrate several essential points. First, it is important to notice just how many parties would have been involved in a coalition or supporting a minority coalition government. It is not hard to see that it would have been very difficult to achieve agreement among parties and would likely (although not necessarily) have been an unstable government beholden to many different individuals and parties for its survival. Second, this ambiguity and confusion seems to confirm that Labour was unprepared to deal with the hung parliament. Part of the reason for this is that, by all accounts (including Peter Mandelson s), the Labour party was divided between those who believed a coalition was a good option and those who believed that the party should go into opposition (Laws 2010; Lee and Beech 2011, 47; Mandelson 2010, 549). Each party had its own red lines or critical policies that they wanted in any deal that would prove to be critical in the negotiating process. Labour was in the unfavorable situation of not realistically being able to form a minority government because of the election outcome and needing the Liberal Democrats in order to stay in power. However, they were, according to Lib Dem, David Laws, surprisingly unwilling to make major concessions. Labour was reluctant to speed up deficit reduction, believing that doing so would halt the recovery (Wilson 2010, ). What Labour had not anticipated was that, since the production of the Lib Dem manifesto, the Liberal Democrats had a change of heart and did want to speed up deficit reduction (this was 49

50 a happy surprise for the Conservatives). As Wilson wrote, this left the Labour and Lib Dem teams in the unusual situation of disagreeing on the economy (2010, 183). The other unanticipated issue was over the introduction of the alternative vote for Westminster elections. In its manifesto, Labour had committed to a referendum on AV, which, while not the ideal electoral system for the Lib Dems, in their view, represented a step in the right direction. However, in negotiations, it became clear that a referendum on AV would be more difficult for Labour to deliver than expected (Laws 2010). In addition to a referendum on AV at the minimum, the Liberal Democrats wanted the four main policies they had promised in their manifesto. As indicated, the Lib Dems had already shifted their policy on deficit reduction after watching the Greek crisis unfold (Laws 2010, 109). This meant that in an agreement the Conservatives could achieve their red line policy of deficit reduction. However, AV was a much more difficult issue for the Conservatives. A referendum on AV would be necessary to bring the Liberal Democrats into coalition, but the Conservatives were (and still are) strongly in favor of FPTP. This stalemate had led to the drafting of a confidence and supply agreement between the parties on Monday 10 May (Laws 2010, 293). What finally broke the impasse was the initiation of formal talks between Labour and the Lib Dems. Gordon Brown s resignation on Monday allowed formal talks to begin because the Lib Dems no longer had to fear being seen as propping up an unpopular Prime Minister. Although there is no evidence that Labour offered anything more than a referendum on AV, David Cameron believed that Labour was prepared to offer the Liberal Democrats AV without a referendum (Qvortrup 2012, 109; Wilson 2010, 210). This led on Monday evening to the Conservative commitment to a referendum on AV, against which they would campaign. That and the deterioration of talks with Labour allowed the Lib Dems to enter a coalition with the Conservatives. 50

51 When David Cameron was invited to form a government by the Queen on Tuesday night, the coalition deal was not yet formalized. He was obliged to become the Prime Minister without knowing if he would lead a minority government or a coalition because although the Liberal Democrat negotiating team had drafted an agreement with the Conservatives, the Lib Dems have a rule known as the triple lock. This rule requires that a coalition agreement must be approved by three-quarters of the Lib Dem Parliamentary Party and Federal Executive before it can go forward (Wilson 2010, 277). When this was obtained in the early morning hours of Wednesday, 12 May, the coalition was officially formed. Conspicuously absent from negotiations between the parties was any discussion of the division of ministerial positions in a coalition. This discussion was left entirely to the party leaders. As Clegg told Cameron, this was a policy-based negotiation (Wilson 2010, 103). Most of the discussion about posts came after the coalition was negotiated. As Wilson explains, the number of positions the Liberal Democrats got was done on a pro rata basis in relation to seats won, although they came back to ask for more for example, more whips (2010, 280). Cameron and Clegg agreed to five ministerial portfolios for the Lib Dems including the post of Deputy Prime Minister for Clegg as well as a host of minister of state, whips and parliamentary under-secretary positions. There seem to have been four overriding concerns that guided the negotiators during those five days in May The first was simple parliamentary arithmetic. Some coalitions such as one between Labour and the Lib Dems alone or a coalition of the Conservatives and one or two small parties were simply impractical given the distribution of seats. Whatever government was formed needed to be able to pass the Queen s Speech and withstand motions of no confidence. The second concern was the economy. How to deal with the struggling economy 51

52 going forward was a divisive issue in negotiations. Further, there were great concerns over how the market would react to the election and fears of the pound being devalued. This not only contributed to a sense of urgency, but also impacted calculations about parliamentary arithmetic. While many backbenchers favored a minority Conservative government, it was not considered the best outcome because of the economic situation the party frontbenchers believed a majority government was necessary to deal with the scale of the country s economic problems (Fox 2010, 608). Third, it is important not to discount the role of electoral reform in these negotiations. It is not an overstatement to say that without an agreement on some sort of electoral reform (a referendum on AV at the minimum) the Liberal Democrats would not have entered a coalition. The Lib Dems accepted the referendum despite the fact that Conservatives would campaign against it because it was simply the best deal they could hope for. 9 The Liberal Democrats staying out of a coalition would likely have resulted in a minority Conservative government and another election, possibly within a year. 10 Finally, there were great concerns about the legitimacy of any government. One of the main concerns about some form of Labour-Liberal Democrat deal was its potential lack of legitimacy. There were fears that the public would believe the two losers of the election were desperately clinging to power and maintaining an increasingly unpopular Prime Minister. Given all of these concerns, it seems perfectly logical that the Liberal Democrats and Conservatives formed a coalition and that one of the coalition s commitments was to hold a referendum on the electoral system. 9 Recall that although Cameron believed Labour had offered AV without a referendum, there is no indication that this was true. 10 Malcolm Bruce, personal interview, May 11, 2011, Portcullis House, London, UK; Tim Farron, personal interview, May 11, 2011, Portcullis House, London, UK. 52

53 Chapter 6: Electoral Reform in the United Kingdom: Why Now? On 5 May 2011, a nationwide referendum was held in the United Kingdom on whether or not to change the electoral system from first-past-the-post to the alternative vote (AV). It was the first nationwide referendum since 1975 when the topic of continued membership in the European Community was put to the voters (Whiteley et al. 2011, 1). The referendum failed when a large proportion of Britons voted to keep FPTP. The causes of the failure of the reform will be discussed in the next chapter. This chapter asks: why did this referendum happen in 2011? Why did electoral reform become a priority on the national agenda at this time and not before? To answer these questions, the factors that contribute to reform becoming an agenda item (part one of the theory found in Chapter 4) will each be analyzed in turn. In this way, it will be possible to distinguish which factors played an important part in the UK and which did not. Further, examination of this case will serve as a test of the model. The Parliamentary Voting Systems and Constituencies Bill passed by Parliament made the referendum possible. This is clearly a case in which elected political elites set up the referendum that was then put to the people (as opposed to the people forcing a referendum against the wishes of the elite). The purpose of this legislation was twofold. First, it provided for the next general election to be fought under the alternative vote system, provided that the change was endorsed in a referendum (White 2011, 3; see also Qvortrup 2012). The AV system proposed would allow voters to express a preference for as many or as few candidates as they wish rather than requiring they rank order all candidates according to preference (White 2011, 3). Second, the Bill proposed a reduction in the number of seats in the House of Commons from the current 650 to 600, a change that would take place regardless of the referendum result (Qvortrup 2012, 109; White 2011, 3). The Parliamentary Voting Systems and Constituencies Bill 53

54 was introduced by the government on 22 July 2010, passed by the Commons on 2 November, passed by the Lords on 16 February 2011, and received Royal Assent on the same day (ibid.). As discussed in Chapter 4, there are multiple factors that contribute to whether or not electoral reform becomes an agenda item. The first three factors explored here (interest judgments by parties in power, leaders, and external pressures) are those that affect political elites directly, but do not affect the referendum. The remaining factors (pressure groups, media, party system changes, crises, and popular dissatisfaction) all impact both elites and the referendum process. Interest Judgments by Parties in Power Despite the fact that the Conservative Party opposed replacing FPTP and the Labour Party did not agree internally on the issue, interest judgments by parties in power played an important part in the reform attempt in the UK. Recall that, according to Katz, parties may wish to change the electoral system because: (1) incumbents may believe the existing system will not continue to benefit them, (2) the interests of the members of the governing party or coalition are not aligned, (3) the party in power may value long-term change in the competitive system over short-term electoral advantage, or (4) the majority party may trade electoral advantage for other goals (2005, 63). Historically, the first reason that incumbents do not believe the system will continue to benefit them has been a factor in considerations of reform in the UK. After the October 1974 election in which Labour won a majority of seats with 39% of the vote, some Conservatives thought that a system of proportional representation might act as a safeguard against revolutionary changes being legitimately enacted by a minority party (Butler 1984, 226). This is 54

55 a consideration that seems to have been forgotten by most Conservatives today. For the Conservatives, the electoral system itself does not seem to be perceived as likely to disadvantage them. However, it should be noted that since the 1980s the SMP system is biased in favor of Labour, meaning that if Labour and the Conservatives secure the same percentage of votes, Labour wins many more seats 141 more in 2001 (Mitchell 2005, ). Since the system as currently designed does give a slight advantage to Labour, the consideration that it may not continue to do so does not seem to have played a role in influencing Labour during this round of reforms. Likewise, the idea that the party in power may value change in party competition over the long-term does not seem to have been a consideration for elites in 2010 and 2011 with the significant exception of the Liberal Democrats. Certainly the Lib Dems would prefer a change in the structure of competition they would like to be more competitive, or more specifically, they would like their seat share to more accurately reflect their vote share (having won 23% of the vote in 2010, they control only 8.77% of the seats in the Commons). However, the Conservative and Labour parties would prefer the party competition remain the same. In other words, they would prefer to stay the two largest parties and alternate power. If anything, these parties would like to return to a time when they controlled an even larger percentage of the votes and seats in Parliament, but this type of change unlikely. However, what is quite clear in the case of the reform efforts is that the interests of the members of the governing coalition are not aligned. The governing coalition, formed in May 2010 between the Conservatives and Liberal Democrats, does not agree on the issue of reform at all. The Conservative Party is unified in its desire to keep FPTP while the Liberal Democrats are unified in their aspiration to adopt a proportional system (they prefer the 55

56 single transferable vote or STV). It is, in part, due to this division of interests within the coalition that it is possible for a referendum on electoral reform to be a government priority even though the larger party is firmly against any change. Finally, it is clear from the coalition negations that trad[ing] electoral advantage for other goals was a crucial reason reform was included on the government agenda (Katz 2005, 63). First, it should be stated that the Conservatives certainly hoped that they would not give up any electoral advantage that the system would not change despite a referendum on the issue. However, they were willing to risk the possibility, however remote, in order to go into coalition with the Liberal Democrats. The Conservative party frontbenchers preferred a coalition government that would command a majority of votes in the Commons to a Conservative minority government. In the words of Renwick, Lamb, and Numan, the Conservative leadership had a strong preference for coalition government in the wake of the 2010 election and were willing to swallow a lot including the Liberal Democrats demands on electoral reform to secure it (2011, 39-40). They were particularly concerned with the economy and believed a government would need to take decisive action to confront the crisis, something that would be difficult or impossible with a minority government. In short, the Conservatives were willing to trade being the larger partner in a stable coalition government that would act decisively to cut the deficit and right the economy for the possibility of a change in the electoral system. Clearly, interest judgments by the Conservative, Liberal Democrats, and to a lesser extent Labour, contributed to electoral reform becoming a government agenda item in While they may be overstating the point, Renwick, Lamb, and Numan argue, reform is now high on the government s agenda because of elite bargaining in a hung parliament, not because of public anger over the expenses scandal (2011, 32; see below for further discussion of the 56

57 expenses scandal). In fact, interest judgments by the political parties were not the only factor that influenced when reform became an agenda priority, but they should not be discounted. Leaders As in New Zealand, the party leaders were another important factor explaining why reform became important in the UK. This is seen through the coalition negotiations. Nick Clegg, as leader of the Liberal Democrats, was a crucial actor. Unlike some members of his party and leaders before him, such as Paddy Ashdown, Charles Kennedy or Menzies Campbell, Clegg was willing to work with either the Conservatives or Labour to form a coalition government he genuinely thought his party was equidistant between the other two (Laws 2010, 270). Some believed strongly that an alliance between Labour and the Lib Dems was preferable, but because Clegg was not wedded to the idea, he was able to negotiate with both main parties in the aftermath of the election. The leaders of the other parties were equally important in the negotiations. At the time of the 2010 elections, Gordon Brown was the leader of the Labour Party. This became a critical factor in the immediate aftermath of the election. As discussed in Chapter 5, a Labour-Lib Dem coalition was problematic for a number of reasons, not least of all, the parliamentary arithmetic. However, Brown s presence at the head of Labour became an issue in and of itself. The Lib Dems did not want to be seen as propping up an unpopular Prime Minster whose party had just lost the election. Not only that, but the chemistry between Brown and Clegg was poor (Laws 2010, 274). In other words, the presence of Brown as the Labour leader made a coalition between Labour and the Lib Dems even more difficult than it would have been otherwise. Had 57

58 another leader been in Brown s position, the outcome of coalition negotiations, and indeed electoral reform may have been different. Finally, David Cameron at the helm of the Conservative Party played an important role in the negotiations and in putting reform on the national agenda. Cameron was more disposed to work with the Lib Dems than other members of his party and in fact preferred a partnership government to a minority Conservative government (Wilson 2010, 91; see also Laws 2010, 51 and 269). It was Cameron who believed that Brown offered the Lib Dems AV without a referendum (a conceivable possibility since AV had been included in the Labour manifesto), which motivated him to go back to his party and try to secure such a referendum (Wilson 2010, 210). In this way, Cameron secured a coalition deal with the Lib Dems rather than a confidence and supply arrangement. Had there been a minority Conservative government, which Cameron did not want, or a confidence and supply agreement with the Lib Dems, electoral reform would not have been an agenda item. The leaders of the three major political parties in 2010 were critical in the formation of the coalition government and in this way had an effect on the national agenda. These leaders, particularly Clegg and Cameron, led their parties in directions that were not the natural first choices of many members. Had other leaders been in their position, it is conceivable that electoral reform would not have become a priority in External Pressures In contrast to the previous two factors, external pressures or influences did not have a significant effect on the process of electoral reform in the UK. Given the absence of serious external shocks to the British political system (common examples include decolonization leaving 58

59 a political vacuum), the main external pressure for electoral reform in the UK comes from the European Union. Despite the United Kingdom s membership in the EU, this is a fairly weak influence as many Britons revel in the differences between the UK and the EU and, in some polls, a plurality would support withdrawal from the EU (YouGov 2010). In 1984, Butler predicted that taking part in European Parliament elections that would be universally conducted using a common, proportional, electoral system would exert pressure on the UK to change the electoral system (226). In fact the 1993 EU resolution to adopt a uniform system for elections to the European Parliament has been a source of pressure for change (Norris 1995b, 73). However, this pressure seems to be slight. Elections for other bodies will be discussed in the next chapter. For now, suffice it to say that pressure from other European nations does not seem to have exerted a significant influence on electoral reform in the UK. Pressure Groups While pressure groups played a role in the referendum campaign (discussed in the next chapter), they were not a critical factor in helping to elevate the issue of electoral reform to the national agenda. Groups such as the Electoral Reform Society and Unlock Democracy were important players in the campaign in favor of the referendum (Qvortrup 2012, 112). However, there is no indication that they were responsible for putting reform on the national or government agenda. As an example, the Electoral Reform Society (ERS) was originally established as the Proportional Representation Society in 1884 before adopting its current name in 1958 (Electoral Reform Society 2012). Since its founding, the ERS has made an effort to keep political reform on the agenda in fair years and foul (Electoral Reform Society 2012). However, they have existed and worked for reform since 1884 so the ERS cannot be the answer to the question, 59

60 Why now? Similarly, Unlock Democracy was not a new phenomenon in the months and years immediately preceding the referendum (Unlock Democracy n.d.). Therefore, pressure groups such as these contribute little to explaining why the referendum happened in Media Like pressure groups, the media did not contribute significantly to placing reform on the agenda in 2010 and As explained in Chapter 4, the media is thought to influence the process of agenda setting indirectly through its effect on public dissatisfaction. If the media successfully links, in the eyes of the public, some crisis or problem in the country to the existing political system, public satisfaction with the system will decrease. In this way, the media could contribute to calls for general reforms, or electoral reforms more specifically. One possible way the media could have had this effect is through the expenses scandal (discussed below). The scandal exercised press and public alike and drew attention to the issue of constitutional reform (Whiteley et al. 2011, 9-10). While the idea of constitutional reform was highlighted, the reforms that are clearly linked to the expenses scandal are proposals such as the recall of MPs (Renwick, Lamb, and Numan 2011, 37). The media and the public do not seem to have directly linked the expenses scandal and electoral reform. Therefore, the media is not a factor that explains why electoral reform assumed prominence on the agenda in 2010 and Party System Changes Changes in the electoral climate can be a critical factor that exerts pressure for electoral reform either through elected political elites or through a referendum. Changes in the party system, especially the rise of previously minor parties, can directly lead to elected elites pushing 60

61 for reform as minor parties become important players in the legislature, but can also lead to public dissatisfaction with the existing electoral system. Conventional wisdom says that Britain is a two-party system, in which case there would be little pressure for reform. But is the British party system the near perfect two-party system it was once proclaimed to be? One of the most basic ways to test if Britain is a two-party system is to look at the twoparty vote the proportion of the vote cast for the two main parties (Conservative and Labour) combined. This data is presented in Table 6.1. The 1922 elections are the first listed because that was the first election in which Labour replaced the Liberal Party as the main opposition to the Conservatives. In that first election of its kind, the two-party vote was just over two-thirds of the total vote. From 1931 to 1970, the two-party vote was near 90% and by the early 1950s, the twoparty vote was 96%. In short, between approximately 1931 and 1970, no one would argue that Britain was anything except a two-party system. 61

62 Table 6.1. Trends in party support in the UK (aggregate). Year Conservative and Labour Liberal/Alliance/Lib Dems (%) (%) Others (%) Feb Oct Notes: The 1922 election was the first in which Labour replaced the Liberal Party as the main opposition to the Conservatives. In that year, the figure for the Liberal Party support includes National Liberals and Liberals. The figures between 1922 and 1945 include several university seats in which members were elected by STV. Source: Curtice 2010, 626. It is since 1970 that the two-party support has diminished and things are more interesting. For approximately thirty years, the proportion of the two-party vote was around three-quarters. By 2005, the two-party vote was about two-thirds of the overall vote, and the proportion was even lower by The Liberal/Alliance/Liberal Democrat vote has generally increased since 1970 (with spikes in 1983 and 1987) and the proportion of the vote secured by other third 62

63 parties has also risen steadily since about 1970 reaching a high in 2010 of 11.9%. This means that, in 2010, a full third of the vote was for a party other than Labour or the Conservatives. Again, as Curtice points out, looking at the votes cast for the two main parties, by 2010 it is difficult to argue that Britain is a two-party system any longer (2010, 627). This analysis is supported by an examination of the effective number of parties from 1945 onward. This data is shown in Table 6.2. The effective number of parties at the electoral (vote) level (ENEP) is shown in column two, labeled Eff N V, while the effective number of parties at the Parliamentary (seat) level (ENPP) is shown in column three, labeled Eff N S. 11 Prior to 1970, the ENEP was no larger than 2.72 and usually lower than 2.50, meaning that in terms of votes, the British system before 1970 was between a two-party system and a two and a half-party system. Since 1974, however, the ENEP has been above 3.00 meaning that at the electoral level, Britain is a three-party system and by 2010 is, with an ENEP of 3.71, close to a four-party system. In other words, the ENEP confirms what we saw in Table 6.1 in recent years it is not appropriate to call Britain a two-party system at the electoral level. The effective number of parties at the Parliamentary level is a slightly different story. Once again, since about 1970 a general trend of increasing ENPP is noticeable. This peaks in 2010 at 2.57, meaning that in terms of Parliamentary seats, Britain is appropriately considered a two and a half-party system. In 2001 and earlier years, at the Parliamentary level, Britain was essentially a two-party system. This difference between ENEP and ENPP reflects the disproportionality of the single-member plurality system the way in which votes are converted to seats magnifies the success of the two main parties. Regardless, by 2010, it is clear that Britain is not a true two-party system in terms of votes or seats in Parliament. 11 See Gallagher and Mitchell 2005c for a cogent explanation of the effective number of parties originally presented in Laakso and Taagepera

64 Table 6.2. Effective number of parties-uk. Year Eff N V Eff N S N Seats Feb Oct Eff N V is the effective number of parties at the electoral level. Eff N S is the effective number of parties at the Parliamentary level. The figures for 1983 and 1987 are calculated on the basis of treating the Liberal-SDP Alliance as one party. Source: Gallagher Combined, the data about the two-party vote and the effective number of parties is fairly convincing that at least in 2010 (and 2010 is a continuation of a general trend from previous years), Britain really should not be considered a two-party system at the national or aggregate level. However, some people might argue that this multi-party competition at the national level masks two-party competition at the constituency level. In other words, in each constituency there are two parties competing, but because which two parties compete varies from constituency to constituency, at the national level, a multi-party system exists rather than a two-party system. In fact this is not the case, but in order to test this, I used data from the British General Election book series since The constituency results were coded according to party competition, which is the unique contribution made by this analysis. In this case, party competition means that 64

65 the results were coded according to which parties competed in a given constituency and won 15% of the vote or more. (This should not be confused with competitive districts, which are sometimes defined such that the gap between the two top performing parties is less than 10%.) The reason 15% was chosen as the threshold is that I feared a 20% threshold would be too high, since for a number of years, Liberal Democrat support has been near that level. I did not want to underestimate the number of constituencies in which third parties compete. Party competition at the constituency level in the three regions of Great Britain was then analyzed. 12 Table 6.3 shows the party competition in England from 1987 through In 2010, 54.79% of English constituencies saw multi-party competition 10% more than saw two-party competition. Further, this percentage was even higher in This is very significant. In England at least, the observed multi-party competition at the national level is not due to twoparty competition at the local level with different parties competing in different constituencies. Table 6.3. Party competition in England as percentage of total number of constituencies in England. 13 Year 1-party 2-party Multiparty Other Again, Northern Ireland was excluded because it essentially has a different party system than that of Britain. 13 The other category in this and the subsequent tables was included because the source data in the British General Election series was ambiguous. It was, at times, unclear if a given constituency was more accurately coded as one-party, two-party, or multi-party competition. 65

66 We see an even greater disparity between two-party (35% of constituencies) and multiparty (60% of constituencies) competition in Wales than in England as shown in Table 6.4. Again, this shows that the multi-party system that appears at the national level cannot be explained by a variety of two-party competition at the local level. Finally, the data for Scotland may be found in Table 6.5. While there is less disparity between two-party and multi-party competition in Scotland than in either England or Wales, in 2010 nearly 50% of Scottish constituencies witnessed multi-party competition. When nearly half of party competition is between three or more parties, it is clear that there is not two-party competition at the constituency level. Table 6.4. Party competition in Wales as percentage of total number of constituencies in Wales. Year 1-party 2-party Multiparty Other Table 6.5. Party competition in Scotland as percentage of total number of constituencies in Scotland. Year 1-party 2-party Multiparty Other

67 This data shows that there is multi-party competition in Britain whether one is examining the national level or the constituency level. The data on the two-party competition between Labour and the Conservatives and on the effective number of parties shows that Britain is not a two-party system at the national level. At the very least, there is a two and a half party system at the national level (ENPP) and, in terms of votes, this goes up to nearly a four-party system (ENEP). But the data also show that this multi-party national competition is not masking twoparty competition at the constituency level. In each region, nearly 50% or more of the party competition is among three or more parties. What does this mean for electoral reform? As I argued in Chapter 4, long-term changes in the electoral climate, specifically changes to the party system, can be a contributing factor to attempts at electoral system reform. This discussion shows clearly how the British party system has evolved from near perfect two-party competition to multi-party competition in recent years. The rise of the Liberal Democrats and other minor parties has exerted pressure to reform the electoral system. These changes in the party system can affect whether electoral reform assumes prominence in several ways. First, the rise of the Liberal Democrats, from being a party of protest at Westminster to becoming a party of power, means that Lib Dem MPs and ministers have a direct impact on government initiatives (Driver 2011, 1). A Conservative majority government would certainly never have initiated a referendum on the electoral system, but a Conservative-Lib Dem government did. As third parties, such as the Lib Dems, gain prominence, they are in a better position to force the issue of change. However, this is not the only way changes in the party system can affect reform. 67

68 As the party system changes, the disproportionality of seat allocation in Parliament is highlighted. This can indirectly affect electoral reform through popular dissatisfaction as voters are confronted with the disproportionate results of multiple elections. For example, in 1974, the Liberal Party won approximately 20% of the votes in the general election, but secured only 2% of the seats in the House of Commons (Butler 1984, 226). 14 This was so flagrantly disproportional that the electoral system was questioned (ibid.). Disproportionate results such as this highlight the perceived unfairness of the system in the eyes of the public and can result in decreased public satisfaction. This, in turn, can lead to calls for reform. The British party system has changed substantially in the last fifty years. With the rise of minor parties, and the Liberal Democrats in particular, Britain can no longer be considered a two-party system. These changes in the party system have allowed the Lib Dems to directly influence policy from within the government and have highlighted the disproportionality of the electoral system. For these reasons, changes in the party system have had a significant effect on elevating electoral reform to the national agenda. Crises For our purposes there are essentially two types of crises. The first is a crisis that highlights problems with the political system and contributes to calls for change. The second is a crisis that is not associated with the failings of the political system and therefore does not lead to reform being elevated to the agenda. In fact, this second type may actively work against the process of putting reform on the national agenda in that the public (and/or elites) may view such a crisis as one that must be dealt with promptly and this could lead them to put off reform they 14 Note that there were two general elections in 1974, one in February and one in October. It is unclear to which election Butler was referring in this example. 68

69 might otherwise consider. Each of these types can be seen in the British case. The first type of crisis (one that contributes to reform efforts) came to the UK in the form of the expenses scandal. In contrast, the economic crisis in the UK is better categorized as the second type of crisis. The expenses scandal is the type of crisis that contributes to calls for reform, at least in theory. The scandal erupted in May 2009 after the Daily Telegraph published expense claims by members of Parliament (Freeland 2009). Parliamentary rules allow for MP s [sic] to claim for the cost of maintaining a second home, but the Telegraph showed that some members were using the rule to claim for extravagant (and, it was implied, improper) items (Freeland 2009). The scandal hit all three major parties rather than affecting only one or two MPs and, for the first time in 300 years, resulted in the forced removable of the Speaker of the House of Commons, Michael Martin (Mayer 2009). It could be considered a political crisis in the sense that it captured the attention of the public, media, and politicians. In addition, each of the three main parties felt compelled to propose changes to the political system in their manifestos to combat this problem (Renwick, Lamb, and Numan 2011, 32). In the sense that political parties sought to propose changes to the system to deal with the issue of expenses, the scandal does seem to have galvanized some reform attempts. There are several indications that the expenses scandal did contribute to calls for reform. In a survey conducted between 29 and 31 May 2009 (at the height of the scandal), 80% of respondents agree[d] that it is the system to blame for the expense scandal, and not just politicians (Ipsos MORI 2009). Others report, public debate about issues of electoral and wider political reform in the United Kingdom greatly intensified in the wake of the scandal (Renwick, Lamb, and Numan 2001, 32). Based on evidence such as this, some scholars have argued that the expenses scandal was an important cause of the appearance of electoral reform on the national 69

70 agenda, in the form of manifesto commitments (to electoral reform and/or political reform more generally) by certain parties and of course, the referendum itself in 2011 (Curtice 2011, 14; Whiteley et al. 2011, 9-10). However, despite well-documented public outrage about the expenses scandal, it is not clear that it had a direct effect on successfully putting electoral reform on the national agenda. In this case, timing is incredibly important. In fact, as Renwick, Lamb, and Numan report, the standard indicators of public attitudes towards the political system and political elite have shown remarkably little change. The British Election Study has found no substantial change in the satisfaction with how democracy works or in levels of trust in political parties or politicians (2011, 35). The Hansard Society s Audit of Political Engagement 7 reports that, while there was a clear spike in the percentage of people who believed the system of governing the UK could be improved in May 2009, by the time the Audit survey was done in late 2009, the numbers returned to lower levels consistent with a general upward trend since 2004 (2010, 32). Those respondents who said that the system of governing the UK works extremely well fell to 24% in May 2009 from 33% in December 2008 before recovering to 28% in November 2009 (Hansard Society 2010, 32; Renwick, Lamb, and Numan 2011, 36). Meanwhile, those who said that the system needs a great deal of improvement rose from 24% in 2008 to 37% in May 2009 before falling back to 27% by the end of that year (Hansard Society 2010, 32). What this shows is that, although the public was certainly angered about the expenses scandal, public opinion returned fairly quickly to more normal levels. This means that anger over the expenses scandal alone, while important, cannot fully explain the presence of electoral reform on the national agenda. It should be noted that the expenses scandal may have contributed to calls for other types of reform, such as judicial scrutiny and MP recall (Ipsos MORI 2009). However, the claim that 70

71 AV or PR would have helped prevent the scandal is tenuous; in any case, the scandal is insufficiently linked to people s daily lives to prompt widespread active engagement, so the expenses scandal does not seem to have directly led to electoral reform (Renwick, Lamb, and Numan 2011, 40). Nevertheless, the impact of the expenses scandal should not be discounted. The expenses scandal was clearly a crisis. However, it should be noted that as such it also affected public dissatisfaction (discussed in the next section). 15 It is also through the expenses scandal that we most clearly see the effect of the media. As discussed previously, The Daily Telegraph newspaper first leaked the scandal. By focusing attention on this issue, the media undoubtedly contributed to public anger towards the MPs responsible and the political system in general. The other major crisis of the last several years has been the economic crisis that continues to rock the UK (as well as the United States and rest of the European Union). Unlike with electoral reforms in Japan, New Zealand, and Italy in the early 1990s, the recession in the UK does not seem to have been perceived as the result of a problematic electoral system (Renwick, Lamb, and Numan 2011, 34-35). Rather than rallying people to the cause of reform, the economic troubles have been perceived as the more important problem that deserves attention. In the words of Whiteley et al., the electoral reform debate (and subsequent referendum) in the context of a serious economic crisis risked inviting the query: Why are they fiddling when Rome is burning? (2011, 20). The idea that concern about the economy far outweighed constitutional issues is supported by poll results. Groups such as Ipsos MORI and YouGov regularly conduct polls 15 As the preceding discussion makes clear, the expenses scandal certainly affected public opinion and dissatisfaction. In order to avoid repetition it was included in this section and not the next. 71

72 designed to explore which issues the public considers most pressing. In many of these, the economy is ranked very highly and electoral or constitutional reform does not appear to be mentioned at all (in some cases it is unclear if respondents were not given the option of choosing reform or if it was an open-ended question to which they did not indicate electoral reform). In a poll conducted between 7 and 13 May 2010, Ipsos MORI s Issue Index shows that 71% of Britons said the economy was among the most important issues facing the country and 53% responded, it is the single most important issue (2010). In contrast, only 1% of respondents said that electoral reform was among the most important issues despite the fact that this poll was conducted, for the most part, during the days immediately following the 2010 election (in which the outcome was a hung parliament) and prior to the formation of the coalition government (Ipsos MORI 2010). It is worth noting that this poll was also taken after the expenses scandal and the percentage of people who viewed electoral reform as one of the most important issues was extremely low. The expenses scandal is precisely the type of political crisis that is likely to lead to calls for electoral reform or political reform more generally. However, this analysis has indicated that the expenses scandal may not have had as big an impact as expected. In fact, the economic crisis in the UK seems to have been a much more important issue for voters at precisely the time electoral reform was beginning to be elevated to an agenda item. The recession is the type of crisis that may work in favor of reform if voters see the political system as the root of the economic problems. However this does not seem to have been the case in the UK. Rather, the economic troubles take precedence over many other issues, including reform, in the minds of the people. 72

73 Popular Dissatisfaction A factor closely related to crises is popular dissatisfaction. As discussed in Chapter 4, popular dissatisfaction could be caused in part by changes to the party system or crises and can be augmented by the media. Regardless of its root, if the public perceives the political system is no longer effective, its dissatisfaction could translate into calls for change. In other words, popular dissatisfaction is a major factor that contributes to putting electoral reform on the national agenda. In the UK, there are three main sources of this dissatisfaction election outcomes that are viewed as unfair, the expenses scandal (discussed in the previous section), and finally, general downward trends in the perception of the political system as a whole. Unfair or perverse election outcomes are outcomes that are unexpected or contrary to intuition. The prime example is a case in which the party winning the most votes fails to win the most seats in Parliament. Between 1945 and 2001, this happened in two out of sixteen elections (Mitchell 2005, 173). While this may seem insignificant given the irregularity with which it occurs, the fact that it has happened is important. This type of result in which the winner of the majority of votes does not win in terms of seats weakens the case for SMP. Unfair election outcomes are clearly the product of the electoral system, so not only will supporters of the party that lost in this manner be dissatisfied, but the public in general may begin to see the electoral system as problematic and perhaps worthy of change. The two cases of an unfair or perverse result occurred in 1951 and February 1974 (Mitchell 2005, 173). Given that these elections are distant relative to the 2010 and 2011 attempt at electoral reform, it is unlikely that they had a major direct influence on this round of agenda setting. Nevertheless, these results do have an impact. 73

74 Another situation in which the outcome of an election may be perceived as unfair is the current circumstance in the UK where it takes fewer votes to elect a Labour MP on average than to elect a Conservative MP. Another way to express this is that if Labour and the Conservatives both received the same share of the vote, in 1997 Labour would have won 82 more seats than the Conservatives and in 2001, they would have won 141 more seats (Mitchell 2005, 162). This electoral bias may contribute to public dissatisfaction and efforts to put electoral reform on the agenda. Further, disproportional outcomes for minor parties contribute to the sense that the electoral system is unfair and to popular dissatisfaction more generally. This has been highlighted as the party system has moved away from the once clear two-party competition towards multi-party competition. In the 2010 election, the Liberal Democrats won 23.0% of the vote, but only 8.8% of the seats in the House of Commons (see Chapter 5 for a more detailed discussion of the election outcome). Unfair election outcomes in which the disproportionality of the electoral system is highlighted or the party that should win the election does not secure the most seats are problems that directly implicate the electoral system. As such they lead to popular dissatisfaction with the electoral system specifically as opposed to politics more generally and may lead to calls for reform. The presence of these increasingly visible unfair or even wrong outcomes in the British political system has likely contributed to elevating the issue of electoral reform to the national agenda. Finally, a discussion of public dissatisfaction would not be complete without some analysis of the general trends in the public perception of the political system. As stated, the general trends over the last decade are a decreasing percentage of people responding that the 74

75 present system of governing the UK works well and an increase in the percentage who say the system needs a great deal of improvement (Hansard Society 2010; Ipsos MORI 2009). By June 2009, only 20% of people were satisfied with the way that [the] Westminster Parliament is doing its job and a full 71% were dissatisfied with the way the House of Commons is doing its job (Ipsos MORI 2009). Additionally, while trust in politicians and government ministers has historically been low in the UK, it has decreased further in recent years, as seen in Tables 6.6 and 6.7. Taken together, this data indicates that trust in politicians and satisfaction with Parliament has been waning in recent years. The continuation of these trends contributes to elevating the importance of electoral reform in that the public is clearly dissatisfied with the political system in general. Table 6.6. Percentage of survey respondents who trust them to tell the truth Government ministers Politicians generally Source: Ipsos MORI Table 6.7. Percentage of survey respondents who do not trust them to tell the truth Government ministers Politicians generally Source: Ipsos MORI

76 In short, public dissatisfaction has come from a number of sources and has been increasing in recent years. The unfair or perverse election outcomes combined with disproportionate outcomes that have been highlighted by the changes in the party system can lead to public dissatisfaction. The expenses scandal (discussed in the Crises section), while it did not perhaps have the great effect one might expect, also contributed to public dissatisfaction. Finally, the general downward trends in trust and satisfaction with Parliament indicate some uneasiness in the public. Combined, this indicates that public dissatisfaction has been increasing in recent years and that this likely contributed to successfully putting electoral reform on the national agenda in 2010 and Conclusion This chapter has explored how the factors laid out in my theory played out in the United Kingdom. There, the process of putting reform on the national agenda was very much elitedriven. Without the hung parliament and the coalition negotiations it initiated, there may not have been a referendum. With different leaders at the helm of the three main parties, the referendum may not have happened. Certainly, all the factors included in the model were present in the UK in the months and years preceding May However, in this case, external influences, pressure groups, the media, and crises do not seem to have contributed substantially to elevating electoral reform to the national agenda. Interest judgments by parties in power, the leaders of the three main parties, party system changes (that led to the hung parliament), and popular dissatisfaction (also contributing to the hung parliament), were important factors. Combined, these factors help 76

77 answer the question posed in the title of this chapter, namely, Why now? It was the convergence of these factors between 2009 and 2011 that led to this most recent reform attempt. It is important to note, however, that the factors included in this theory are not just useful in explaining the British case. In fact, the example of the United Kingdom also lends credence to the model. Using only the factors included in Chapter 4, it is possible to explain how electoral reform became an agenda item in This indicates that, when tested against the UK at least, there are no large holes in the theory there does not appear to be anything that was critical in the UK that is not included in the model. Naturally, further testing is necessary, but part one of the model has held up to testing against this case. Will part two also hold up against the test of the UK? 77

78 Chapter 7: Why the 2011 Referendum Failed The AV referendum held on 5 May 2011 presented this prompt to the electorate: At present, the UK uses the first-past-the-post system to elect MPs to the House of Commons. Should the alternative vote system be used instead? (White 2011, 4; Whiteley et al. 2011, 2). With a turnout of 42.2%, the No vote was 67.9% (13,013,123 voters) while 32.1% (6,152,607 voters) voted in favor of AV (Electoral Commission 2011, 16; McGuinness and Hardacre 2011, 16). Out of 440 counting areas, the Yes vote was a majority in only ten (Electoral Commission 2011, 17). This chapter seeks to explain why the referendum failed. To do so, the factors that influence the success of reform attempts (part two of the model in Chapter 4) will be discussed in depth in an effort to determine which factors were important in the United Kingdom. The referendum on the alternative vote was held simultaneously with elections to the National Assembly for Wales, the Scottish Parliament, the Northern Ireland Assembly, and local and/or mayoral elections in parts of England (Electoral Commission 2011, 15). In total, the referendum coincided with other elections in all but 47 (out of 440) counting areas (McGuinness and Hardacre 2011, 16). Part of the reason for holding the referendum on the same day as other elections was to increase the turnout for the referendum while decreasing the overall costs incurred for both the referendum and elections 16 (Electoral Commission 2011, 34). In preparation for the referendum, 27 million information booklets were distributed to approximately 96.1% of households as part of a public awareness campaign by the Electoral Commission that began on 30 March 2011 (Electoral Commission 2011, 5; White 2011, 15). Recall from Chapter 4 that there are multiple factors that affect the outcome of a reform attempt. When present, some of these factors contribute to the success of reform (positive 16 Tristram Hunt, personal interview, May 9, 2011, Portcullis House, London, UK. 78

79 variables), while some make reform less likely to succeed (negative variables). Further, it is possible that not all of these factors are necessary. Regardless, electoral reform is most likely to occur if the reform attempted is acceptable; the opposition to reform is weak, unorganized, or poorly funded; the funding of the pro-reform campaign is strong; other national issues are either not present or increase the public s desire for reform; the public has experience with other electoral systems; and public opinion is in favor of reform. Which Reform? The referendum presented voters with a simple proposition in the sense that there were only two choices on offer retain the existing SMP system or switch to AV. The choice presented to the electorate was primarily the result of elite bargaining during the coalition negotiations, discussed in Chapters 5 and 6. AV is in fact the preference of very few elites and is certainly not favored by either the Conservative or Liberal Democrat Parties that, in coalition, legislated for this referendum. By and large the Conservatives prefer maintaining FPTP (Conservative Party 2010, 67) while the Liberal Democrats have long advocated for a fully proportional system (recall that AV is a majoritarian system) and favor the single transferable vote (STV) 17 (Liberal Democrats 2010, 87-88). In fact, Nick Clegg has referred to AV as a miserable little compromise (Qvortrup 2012, 110). However, during the coalition negotiations it became clear that the Liberal Democrats would not join a coalition without the promise of a referendum on electoral reform and AV was as far as the Conservatives were willing to go (see Chapters 5 and 6). The Coalition Agreement stipulated that the government would pass legislation establishing the referendum ( Conservative Agreements Reached 2010, 3; HM 17 Malcolm Bruce, personal interview, May 11, 2011, Portcullis House, London, UK; Don Foster, personal interview, May 9, 2011, Portcullis House, London, UK. 79

80 Government 2010, 27). Had the referendum succeeded, the type of AV that would have been instituted was the: optional preferential system [which] allows, but does not require, voters to rank the candidates in order of preference by numbering them 1,2,3 etc. Voters may simply vote for one candidate and if they do number their preferences they do not have to do this for all the candidates (White 2011, 12). Recall that which reform is on offer is potentially affected by the nation s experience with alternatives, public opinion, and the strategy of the pro-reform campaigners. In the UK, all of these were less important than elite bargaining. Although the Mayor of London is elected under a system similar to AV, 18 this does not seem to have impacted this particular reform attempt. Public opinion, or more specifically the opinion of each party s supporters, likely played a role in the positions of the respective parties. However, as the reform proposed was neither party s preference, it does not appear that public opinion had a large impact on which reform was on the ballot. Finally, the strategy of those in favor of reform did play a small role in which reform was on the ballot. The Liberal Democrats and other campaigners took a pragmatic stance towards reform. They knew that it would be impossible to secure a referendum on a proportional system given that the support of Conservatives was necessary. However, they viewed AV as a step forward on the path towards more dramatic reforms a step in the right direction. 19 That AV was chosen as the reform option on the referendum certainly impacted the result. Some observers partially attribute the referendum result to the fact that the choice was between FPTP and AV. A portion of the public (and indications are that this was not an insignificant portion) found the alternative vote system confusing or complicated, which played a 18 Tim Farron, personal interview, May 11, 2011, Portcullis House, London, UK. 19 Ibid. 80

81 role in its dramatic defeat. 20 In a British Election Study (BES) poll conducted prior to the referendum, 33.7% of respondents said that AV is difficult to understand (British Election Study 2012). Partially this confusion on the part of voters was a product of the campaigns (more on this shortly), but the fact that the public finds the electoral system confusing does not bode well for the success of a reform attempt. Experience with Alternatives As discussed in Chapter 4, it is plausible that national experience with alternative electoral systems may make the public more receptive to electoral reform. Scholars look to the case of Australia to illustrate this point. Although the UK is sometimes regarded as the bastion of SMP, Britons do in fact use a number of different electoral systems. The Welsh and Scottish Parliaments are elected using mixed-member systems, as is the London Assembly; PR-STV is used for the Northern Ireland Assembly and Scottish local government elections; and Members of the European Parliament (MEPs) are elected via closed-list PR 21 (Mitchell 2005, 157; Clegg and Harper 2011). Further, the supplementary vote is used to elect the Mayor of London, which is essentially a modified form of AV (White 2011, 5). As Mitchell writes, the Westminster SMP system is now the exception rather than the rule in electing UK politicians (2005, 174). One would expect this variety to increase the chances of successful reform, but as it turns out, the evidence is mixed. In their analysis, Whiteley et al. find that, when controlling for other factors, Scots were less likely to vote yes [in the referendum] than were English (this result 20 Tim Farron, personal interview, May 11, 2011, Portcullis House, London, UK; Don Foster, personal interview, May 9, 2011, Portcullis House, London, UK; Sheila Gilmore, personal interview, May 11, 2011, Portcullis House, London, UK. 21 Tim Farron, personal interview, May 11, 2011, Portcullis House, London, UK. 81

82 was statistically significant, p<0.01) (2011, 15). This is surprising given the variety of electoral systems used in Scotland. Sheila Gilmore (Labour MP) reported: I think there was an assumption that the Scottish electorate might be more open to change because it already had experience of different voting systems. I didn t find that, I found if anything the opposite seemed to happen that people felt they already had so many different ways of voting that they didn t, couldn t see the point in yet another one. But I think actually a lot of the Scottish electorate feel they ve already got too many different things and they find that quite confusing in the absence of any powerful motive for change, people tend to go for the status quo. 22 Given this information, it would seem that, in fact, experience with alternatives negatively affected the result of the referendum, contrary to expectation. However, the evidence is ambiguous. In contrast to Whiteley et al., McGuinness and Hardacre report voters in England were more opposed to changing to AV than voters in Scotland, Wales and Northern Ireland (2011, 4). Since voters in those regions have experience with more electoral systems than voters in England, this would seem to indicate that perhaps experience with alternatives did make some portions of the electorate more receptive to the referendum. The other possibility is that some other factor or combination of factors made voters in Northern Ireland, Scotland, and Wales more receptive to reform. Further complicating the analysis, the entire British electorate has at least some experience with alternative systems, as voters in all regions are eligible to vote for MEPs (elected via closed-list PR) at the very least. In light of this, it is nearly impossible to know exactly how big a role experience with alternatives played in this referendum result. Perhaps the result would have been even more lopsided in favor of No if people did not have the experience of voting for MEPs (and other elections for which they were eligible). In short, the evidence with regard to 22 Sheila Gilmore, personal interview, May 11, 2011, Portcullis House, London, UK. 82

83 this factor is ambiguous. It appears, however, that experience with alternative systems of voting was not as critical a factor as expected. Other National Issues In the UK, other national issues seem to have overpowered the referendum on electoral reform. Recall that public opinion can affect other issues. This is because, although the other issues are present regardless of public opinion, the significance of those issues comes from the weight the public attaches to them. The economic crisis in the UK was discussed in the last chapter as the type of problem that did not, in this case, encourage support of electoral reform. Britons were more concerned with the recession than constitutional tinkering so the economic crisis was not a factor in putting reform on the agenda. Similarly, the ongoing economic woes of the UK and the EU contributed to the failure of the referendum because the public did not view electoral reform as important relative to the economic issues. In YouGov s issue survey, there is not even a category for constitutional reform, but in April and May 2011, 75% of respondents said that the economy was one of the most important issues facing the country (YouGov 2012). It is also possible that other government initiatives took precedence over electoral reform and contributed to the failure of the referendum. One such issue that has received wide coverage is the proposed reforms to the National Health Service (NHS). The same YouGov survey in April and May of 2011 found that between 32% and 37% of respondents answered that health was one of the most important issues facing the country (the only issues receiving a higher percentage of responses were the economy and immigration) (YouGov 2012). Given that the NHS is an issue that clearly impacts the daily lives of Britons in a way that the ethereal issue of 83

84 electoral reform may not, it is reasonable that concerns over the future of the NHS contributed slightly to the failure of the referendum by taking precedence in the minds of the voters. Finally, recall that local and devolved elections occurred on the same day as the referendum. All indications are that this contributed to the failure of the referendum in several ways. First, it appears plausible that the electorate viewed the elections as more important than the referendum and voted accordingly. Sheila Gilmore said the fact [that] the timing of the referendum was linked to other elections, particularly for us in Scotland, meant that it [the referendum] took second billing, both for us as party activists, [and] also for the electorate. 23 Second, the elections and referendum were held at a time when support for the Liberal Democrats was particularly low (Curtice 2011, 6; YouGov 2011b). This has led members of all three main parties to argue that having the ballots on the same day meant that people who were angered at the government or the Liberal Democrats in particular not only expressed their dissatisfaction by voting for other parties in the local and devolved elections, but also voted No in the referendum for that reason. 24 Lib Dem MP Don Foster said there was undoubtedly an anti-lib Dem feeling [in the elections], and therefore that helped to engender an anti- AV feeling. 25 Labour MP Tristram Hunt argued they made a strategic mistake by having it on the same day as the elections in Scotland, in Wales, and the council [local] elections in England so that people who wanted to give the government a kick, and give the Liberal Democrats a particular kick, gave them an extra kick by voting against AV. 26 Conservative MP Andrew Turner, when asked what the real message of the referendum result was, answered, Are 23 Sheila Gilmore, personal interview, May 11, 2011, Portcullis House, London, UK. 24 Malcolm Bruce, personal interview, May 11, 2011, Portcullis House, London, UK; Tim Farron, personal interview, May 11, 2011, Portcullis House, London, UK. 25 Don Foster, personal interview, May 9, 2011, Portcullis House, London, UK. 26 Tristram Hunt, personal interview, May 9, 2011, Portcullis House, London, UK. 84

85 they voting against the Liberals? Quite strongly, yes I think they are. 27 It appears that the elections took precedence over the referendum and contributed to the negative result. In short, there were multiple other issues on the national agenda at the time of the referendum. While it is difficult to know for sure, all indications are that the issues of the economy, NHS reform, and the concurrent elections took precedence for at least a portion of the electorate. Of course we can never know what the referendum result would have been if these other issues had not been present, but they certainly seem to have contributed to the substantial No vote. Public Opinion Public opinion is probably the factor most clearly linked to the success or failure of a reform attempt, especially if that reform is put to the voters in a referendum. When considering public opinion in the UK regarding the referendum, it is quite clear from the result that voters opposed changing the voting system. However, it is also possible to begin to tease out why. Other national issues at the time of the referendum have already been discussed and clearly impacted public opinion. The public did not see an urgent need for electoral reform in light of other pressing concerns. Additionally, experience with alternatives and the strength of the opposition (discussed shortly) influenced public opinion. Other major factors that contributed to the negative public opinion were confusion and the low favorability of the Liberal Democrats. Support for AV varied considerably in the lead up to the referendum as seen in Table 7.1 and Figure 7.1. According to Lib Dem MP Tim Farron, the Yes vote was ahead by a very narrow 27 Andrew Turner, personal interview, May 10, 2011, Houses of Parliament, London, UK. 85

86 margin at the end of March 2011, and given the slim lead, he was pretty sure we d lost it back then. 28 Table 7.1. Voting Intention in Referendum 28 February-3 May Yes (%) No (%) Do not Know (%) Would not vote (%) February 28-March March March 31-April 1* April 7-8* April 11-12* April 14-15* April 18-19* April 25-26* April 28-29* May 2-3* Note: The survey stated the referendum question and asked, If the referendum were held tomorrow, how would you vote? *Surveys took into account likelihood to vote. Source: YouGov 2011a. 28 Tim Farron, personal interview, May 11, 2011, Portcullis House, London, UK. 86

87 Figure 7.1. Voting Intentions in Referendum. Source: British Election Study As already discussed, survey respondents found AV difficult to understand even when surveyed very close to the date of the referendum. In this way, which reform was presented to the public affected public opinion. The BES Pre-Referendum Survey found that 33.7% of respondents said that AV was hard to understand (BES 2012). Even more concerning, a majority (51 per cent [sic]) of respondents in a BPIX [British Polling Index] survey conducted just over a week before the referendum indicated that they only partially understood AV or did not understand it at all (Whiteley et al. 2011, 5). Clearly a significant portion of the electorate found AV difficult to understand. Given this, the large No vote is not surprising. However, it is 87