Jump to ContentJump to Main Navigation
The Politics of Electoral Systems$

Michael Gallagher and Paul Mitchell

Print publication date: 2005

Print ISBN-13: 9780199257560

Published to Oxford Scholarship Online: February 2006

DOI: 10.1093/0199257566.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 05 August 2020

Ireland: The Discreet Charm of PR‐STV

Ireland: The Discreet Charm of PR‐STV

Chapter:
(p.511) 25 Ireland: The Discreet Charm of PR‐STV
Source:
The Politics of Electoral Systems
Author(s):

Michael Gallagher (Contributor Webpage)

Publisher:
Oxford University Press
DOI:10.1093/0199257566.003.0025

Abstract and Keywords

Ireland is among the few countries to employ proportional representation by the single transferable vote (STV), an electoral system that is highly rated by many researchers in the field. In the Irish context, STV is used in constituencies of small district magnitude, but it still delivers a high degree of proportionality. The party system has been characterised by moderate multipartism and, unusually for western Europe, by a significant number of independent MPs. Critics maintain that the provision for voter choice leads to intra-party competition with dysfunctional consequences, but the voters wield a veto over electoral system change and are unlikely to approve any new system that reduces their opportunity to participate in choosing their representatives.

Keywords:   Irish politics, proportional representation, single transferable vote, intra-party competition, independent MPs

In a comparative perspective, the mystery of Ireland's party system is one of the two features of the country that attract attention among researchers casting around for something interesting to study. The other is the electoral system. Ireland has proportional representation (PR) and, while this itself may not be cause for excitement, it stands out because its electoral system is not based on party lists, like those of all the other PR countries covered in this book. Instead, it employs the single transferable vote (STV) to elect its lower house of parliament, Dáil Éireann. PR‐STV is not a widely used electoral system. It is employed in only two countries (Ireland and Malta) to elect the national parliament (it was also used in Estonia on one occasion, in 1990).1 In addition, it is or has been employed to elect upper houses or regional assemblies in a number of countries that have come under British influence, including Australia (Chapter 4), India (Chapter 7), South Africa, Northern Ireland, Nepal, Sri Lanka, and Pakistan (Hicken and Kasuya 2003: 139, 144). However, it is an electoral system whose fascination among and appeal to students of electoral systems far exceeds its actual use, and accordingly this chapter will occasionally discuss it in other settings or generically as well as in the Irish context.

Political background

Ireland has a parliamentary system of government and this, like a range of other aspects of Irish cultural, economic, and political life, shows evidence of a clear British imprint. Thus, Ireland exhibits a number of features characteristic of Lijphart's Westminster model (Lijphart 1999: 10–21): cabinet dominance, concentration of legislative power in one chamber (asymmetric bicameralism), and unitary and centralized government. However, it is not always realized how far Ireland deviates from a pure Westminster majoritarian model and instead displays signs of Lijphart's consensus model (Lijphart 1999: 34–41). Its pattern of government is characterized by coalition governments, a multiparty system, a high degree of interest group involvement in policy-making, constitutional rigidity, judicial review carried out by perhaps the most activist court system in western Europe—and (p.512) proportional representation (for an overview of the Irish political system, see Coakley and Gallagher 2005).

The Irish party system, to an outside observer, is a mixture of the familiar and the unique. The familiar elements are rather small: a Labour Party that is archetypally social democratic but atypically weak (averaging only about 11 per cent of the votes at elections since 1945); a Green Party that has been represented in the Dáil since 1989 but has yet to enter office; and the Progressive Democrats (PDs), who combine the traditional European liberal appeals of church–state separation and free‐market economics. The two main parties are less easy to place in comparative terms. Ireland's largest party since 1932, Fianna Fáil, is often described as a ‘centre‐right’ party, yet it has usually enjoyed warm relations with the trade union movement. It has formed coalition governments since 1992 with both the PDs and Labour, with few signs of ideological discomfort in either case (Mitchell 2000). The runner‐up at each election since 1932 has been Fine Gael, a member of the Christian Democratic group at European level.

Having sketched the context, in the rest of this chapter we shall examine why the electoral system was adopted, how it operates, what consequences it has had, and whether it is likely to be replaced.

Origins of the electoral system

The adoption of PR‐STV in Ireland was closely bound up with the genesis of the independent Irish state in the early years of the twentieth century. From 1800 onwards, Ireland was, not entirely willingly, part of the United Kingdom of Great Britain and Ireland. After its foundation in 1884, the PR Society campaigned for the adoption of PR‐STV in all British elections and in the dominions (this account is based on O'Leary 1961: 1–12 and O'Leary 1975: 153–9). By the time the British government had accepted the idea of Irish home rule all shades of nationalist opinion were supportive of the principle of PR. The largest nationalist group, the Irish Parliamentary Party (IPP), was keen for tactical reasons to appear to hold a conciliatory attitude towards the anti-home rule Protestant minority on the island. More significantly in the long run, the leader of the fledgling Sinn Féin party (committed to outright independence for Ireland), Arthur Griffith, became a convert after attending a Dublin meeting organized by the PR Society and addressed rousingly by the Cornish ex-government minister Leonard Courtney (Hart 1992: 170; O'Leary 1961: 3). Provision was made in the 1914 Home Rule Act for the election of a number of Irish MPs at Westminster to be conducted by PR‐STV (O'Leary 1975: 155–6).

This act, though, was overtaken by events, since its implementation was deferred for the duration of the First World War, and by 1918 the Irish political landscape had changed fundamentally. The IPP was now on the verge of being swept aside by Sinn Féin (‘Ourselves’), which had won a series of by‐elections following a militarily unsuccessful but politically transformative rising at Easter 1916. When the next Westminster general election took place in December 1918, Sinn Féin swept the board, winning sixty‐nine of the seventy‐two territorial seats (Coakley 1994: 33). (p.513) Its seats–votes ratio (96 per cent of the seats for 65 per cent of the votes) reflects the fact that the election was fought under the single‐member plurality (SMP) system. However, on the same day, local elections were held in Sligo under PR‐STV, the result of an initiative by the local IPP MP Thomas Scanlan, who had secured the passage at Westminster of a private bill (Hart 1992: 201; O'Leary 1975: 156–7). Here, Sinn Féin was confined to third place, with the seats being distributed among a number of parties, and the lessons of this experiment received national press coverage.

Consequently, during the convoluted run‐up to Irish independence—secured de facto by the Anglo–Irish treaty signed on 6 December 1921—PR‐STV was designated the electoral system for the new state. The British favoured it as a means of securing representation for minorities (specifically for southern unionists, nearly all Protestants, who were opposed to the idea of breaking the union with Britain); Sinn Féin, led by Griffith, supported the system on principle; and other groups in the south also welcomed it, seeing it, as the British did, as less likely than SMP to facilitate a Sinn Féin hegemony. Accordingly, the Government of Ireland Act of 1920 prescribed PR‐STV as the means by which the two future parliaments in Ireland (one in ‘Southern Ireland’ and one in ‘Northern Ireland’) would be elected, and this was one element that did not prove contentious when the details of the treaty were hammered out. Sinn Féin may also have been attracted to PR because it marked a clear departure from British practice, and since it was aware that the adoption of PR was a feature of nearly all constitutions in the new states of post-war Europe (O'Leary 1961: 12).

In the first months of 1922 the pro‐Treaty political elite in the south drew up a constitution for the Irish Free State, which provided that members of parliament be elected ‘upon the principles of Proportional Representation’ (Article 26). The Dáil debates on the subject later in the year featured very little discussion of this, and the subject was wrapped up in 1923, when the Electoral Act specified STV as the method of implementing PR and provided for the division of the country into constituencies.

The adoption of PR, then, caused very little controversy. Why, though, was STV chosen rather than the much more common list system? The explanation seems to be, simply, that few of those making the decision were aware of the range of electoral systems from which they could have chosen. PR‐STV was by now familiar; list systems were not, despite the efforts of an early student of the subject to make the case for them and against STV in Ireland (Meredith 1913). STV was not specified in the 1922 constitution only because TDs (members of the Dáil) did not realize that STV was merely one method, and an unusual one at that, of attaining PR. (Indeed, the electoral system is popularly known in Ireland even today simply as ‘PR’.)

The next significant development occurred in 1937, when Fianna Fáil, now in government, introduced a new constitution (still in force today), which was put to a referendum of the people on 1 July 1937. This contained much more detail than its 1922 predecessor. It specified that (a) the method of election would be PR‐STV (Article 16.2.5); (b) every constituency must return at least 3 TDs (Article 16.2.6); (p.514) and (c) the ratio of population to TDs should ‘so far as it is practicable, be the same throughout the country’ (Article 16.2.3).

Fianna Fáil, and especially its leader Éamon de Valera, had many reservations about the system, and it may be that it was specified in the 1937 constitution (of which de Valera was the author) out of a fear that otherwise the opposition would manage to secure the defeat of the whole constitution in the referendum (O'Leary 1961: 35). In the event, the constitution was approved in the referendum by a margin of 57 per cent to 43 per cent.

Entrenching it in the constitution means that it can be changed only by referendum. Two attempts have been made to alter it, each initiated by Fianna Fáil, and on each occasion the proposal was to replace PR‐STV by SMP (termed ‘first past the post’ or ‘the straight vote’ by Fianna Fáil in the referendum campaigns). The first, in 1959, was defeated only narrowly, with 48 per cent voting to change to SMP and 52 per cent favouring the retention of PR‐STV. The second, in 1968, was much more decisive, only 39 per cent voting for change.2

On each occasion, the proposal was opposed by all parties other than Fianna Fáil. In both 1959 and 1968 the dominant theme of the arguments for change was the claim that PR makes it difficult to achieve stable government and thus weakens democracy, a perspective inspired by the work of the US‐based professor Ferdinand Hermens, who indeed played a minor part in the 1959 debates (for accounts of the two campaigns see O'Leary 1961: 58–83; 1979: 66–70). Opponents of PR‐STV occasionally threw a little nationalism into the mix, alleging that the system had in effect been foisted onto Ireland by the British, and especially by Britain's PR Society, which ‘wanted to try out its nostrum on the dog’ (O'Leary 1961: 32). During the Dáil debates of 1968 one TD suggested, as a compromise, that the alternative vote rather than SMP should replace PR‐STV, but his proposal found no support. Other than this, there was very little discussion or even awareness of the potential range of options, and academics remarked mournfully on the generally low and ill‐informed standard of debate (Chubb 1982: 163). Since 1968, while there has been discussion of changing the electoral system, as we outline later, there have been no further firm proposals to do this.

How the electoral system works

The electoral system has remained unaltered since it was first employed in 1922. The basics of STV are straightforward enough, and are explained in detail in Appendix A, but, as has been noted (Farrell et al. 1996: 30–3), certain aspects of the system such as ballot paper design, district magnitude, counting methods, and the filling of casual vacancies, vary from case to case. First, we will look at the way seats are filled.

(p.515) Converting votes into seats

STV is very much a candidate‐centred method of election, with parties mattering as much or little as the voters feel them to matter. Unlike other methods of PR, PR‐STV does not convert party shares of votes into a more or less ‘fair’ number of seats. Instead, candidates come to be elected by receiving sufficient support from the voters, as described in detail in Appendix A (pp. 593–6). Because of this, all seats are allocated at constituency level.

Table 25.1 shows the result of the 2002 election. In broad terms it shows that even though PR‐STV does not guarantee any precise relationship between votes and seats, in practice it produces outcomes that are typical of PR rather than plurality or majority systems and facilitates a multiparty system. However, Table 25.1 shows that the largest party secured a marked over-representation, and indeed disproportionality reached an all‐time high in 2002, something that we return to later.

The details of counting votes, as explained in Appendix A, can give rise to the charge that PR‐STV is unduly ‘complex’.3 However, there are different levels of ‘understanding’ involved, and voters need only know how to cast a vote and (in general terms) what effect this has (Meredith 1913: 36; Sinnott 2005: 109–10). Taagepera (1996: 31) reports that though many voters in Estonia were troubled by

Table 25.1 Result of 2002 general election, Ireland

Party/group

Votes

% votes

Seats

% seats

Fianna Fáil

770,748

41.5

81

49.1

Fine Gael

417,619

22.5

31

18.8

Labour

200,130

10.8

21

12.1

Prog Democrats

73,628

4.0

8

4.8

Greens

71,470

3.8

6

3.6

Sinn Féin

121,020

6.5

5

3.0

Socialist Party

14,896

0.8

1

0.6

Christian Solidarity

4,741

0.3

0

0

Workers' Party

4,012

0.2

0

0

Socialist Workers

3,333

0.2

0

0

Independents

176,305

9.5

13

7.9

Total

1,857,902

100.0

166

100.0

Source: Weeks 2003: 248–9.

Note: Only 165 of the 166 seats were contested, since the outgoing Ceann Comhairle (Speaker) of the Dáil, on this occasion a Labour TD, is automatically re‐elected. The ‘seats (%)’ column refers to the percentage of contested seats won by each party.

(p.516) their inability to figure out exactly how the votes were converted into seats when PR‐STV was introduced there, they had no difficulty in handling the system. Of course, by offering voters a choice among the candidates of each party, PR‐STV, like open‐list PR systems, could be seen as adding to the complexity of the decision to be made by inviting voters to acquire information about the candidates on a number of criteria. Thus, in focus group analysis in Britain, one respondent indicated concern that being given a choice among a party's candidates would impose some kind of responsibility to find out about them all (Farrell and Gallagher 1999: 307–8).

Suggestions that PR‐STV would be so complicated that voters would have to place themselves in the hands of their parties for guidance (Meredith 1913: 76–8) have proved wide of the mark. It is true that parties do frequently try to ‘manage’ their votes by asking certain voters to rank their candidates in a particular way, in order to maximize the number of seats they can win with a given level of support.4 However, this is not really party control; it is simply advice that party‐oriented voters may or may not choose to accept.

Critics point out that STV is vulnerable to non‐monotonicity and is known to have other theoretical deficiencies, as do all voting systems (Nurmi 1996), though others dismiss the possibility of non‐monotonicity as ‘a nonissue’ (Austen‐Smith and Banks 1991: 531). Dummett's claims that PR‐STV is ‘quasi‐chaotic’ and ‘exceptionally erratic in its operation, producing results that are virtually random’ is simply hyperbole, and even Dummett acknowledges that when voting patterns are structured, for example by party allegiances, the outcome ‘is likely to be to a great extent fair’ (Dummett 1997: 142, 151, 161). The conclusion of the editors of a collection of chapters on PR‐STV as to whether the system can be manipulated is robust:

STV generally presents such difficult calculations to voters seeking to behave tactically that it seems to make little sense to do anything other than register a sincere preference for the party that they would most like to see win. (Bowler and Grofman 2000: 268)

Ballot paper design

In other contexts where PR‐STV is used, candidates are grouped on the ballot paper by party, but in Ireland they are listed alphabetically. The legal fiction that elections were about selecting individual MPs died hard. The Irish constitution makes no mention of parties, and electoral law did not recognize them until 1963—only since that date have candidates' party affiliations, if any, appeared on the ballot paper.

(p.517) In recent years ballot paper design has evolved in two ways. First, as a means of trying to make the act of casting a ballot more ‘voter‐friendly’ at a time of declining turnout, candidate photographs and party logos are included on the ballot paper (see Figure 25.1 for an example). Casting a valid vote entails, at a minimum, writing the number ‘1’ beside the voter's favourite candidate, indicating a first preference. Voters are free, but not obliged, also to rank some or all of the remaining candidates in order of their preference. Second, proposals have been made to switch to electronic voting and counting, and three constituencies voted this way at the 2002 election (see Weeks 2003: 265–7). However, plans to move to fully electronic voting by the following general election were thrown off course when a commission advised against its adoption for the 2004 local and European parliament elections, warning that it was unable to satisfy itself as to the accuracy and secrecy of the proposed system (Sinnott 2005: 131).

District magnitude

One of the first scholars to appreciate the important role of district magnitude in determining the impact of any electoral system was the Irish academic James Hogan (1945: 13), who observed: ‘the decisive point in P.R. is the size of the constituencies: the larger the constituency, that is, the greater the number of members which it elects, the more closely will the result approximate to proportionality’. Hogan suggested further that when district magnitude falls below 5, ‘there ceases to be a real proportionality between votes cast and seats obtained’, and although he offered no firm evidence for this, subsequent research backs it up. Taagepera and Shugart (1989: 114) observe that with a typical multiparty constellation, five or six ‘is the lowest district magnitude that can be counted on to provide relatively proportional outcomes’.

In that context, Ireland's average district magnitude is strikingly low. Not since 1933, in fact, has it reached the supposedly minimum level needed to deliver proportional outcomes (see Table 25.2). Hogan was alive to the implications of this. Writing under the influence of the work of Ferdinand Hermens, according to whom PR led more or less inevitably to the collapse of democracy, he felt obliged to explain why stable democracy had survived in Ireland. His explanation was that small district magnitude had prevented genuine proportionality and the effect was ‘to tilt the balance definitely in favour of the large parties’ (Hogan 1945: 15). Thus, he argued, PR had ‘succeeded’ in Ireland precisely because it had been so attenuated by small district magnitude that it was not ‘full P.R.’ at all (Hogan 1945: 24). The system in operation, he argued, was ‘a compromise between P.R. and the plurality method of voting, with P.R. still the dominant element in the combination’ (Hogan 1945: 15). Since this statement is sometimes quoted in support of particular classifications of PR‐STV, incidentally, we should be clear that Hogan was not making a judgement on PR‐STV per se; rather, he was referring to that system as applied in Ireland, with its small district magnitudes (for assessments of Hogan's work, see Garvin 2001; O'Leary 2001). (p.518)

Ireland: The Discreet Charm of PR‐STVIreland: The Discreet Charm of PR‐STV

Figure 25.1 Ballot paper from Wicklow constituency, Irish election, 2002

(p.519)

Table 25.2 District magnitude, and distribution of constituencies, at Irish elections 1922–2002

Elections

Seats

Constituencies

Average district magnitude

Number of TDs returned

3

4

5

6

7

8

9

1922

128

28

4.6

3

16

4

2

1

2

1923–33

153

30

5.1

8

4

9

5

3

1

1937–44

138

34

4.1

15

8

8

3

1948–57

147

40

3.7

22

9

9

1961–5

144

38

3.8

17

12

9

1969–73

144

42

3.4

26

14

2

1977

148

42

3.5

26

10

6

1981–9

166

41

4.0

13

13

15

1992–7

166

41

4.0

12

15

14

2002

166

42

4.0

16

12

14

(p.520) Counting methods and casual vacancies

Although the basic principles of counting votes under STV are constant, some of the details vary. In particular, the decision as to how to transfer ‘surplus’ votes (those votes over and above the quota) from an elected candidate is made differently in different countries (Farrell et al. 1996: 32–3). In Ireland, as explained in Appendix A (pp. 595–6), the votes that are physically transferred are an accurate sample with regard to their next preference, but may not be with regard to later preferences. Whether this has ever made any difference to an actual constituency count is not known, but one attempt to estimate the effect concluded that it certainly arose very rarely. Examining the 800 individual constituency results over the 1922–82 period, it found that in only fourteen of these was there, even on the ‘worst’ assumptions, as much as one chance in a hundred that a different selection of votes would have made a difference to the outcome (Gallagher and Unwin 1986: 251). The problem can be solved by the ‘Gregory method’ (see Appendix A), although this adds to the time and complexity of the count when votes are counted by hand.5 It may be that if in the future Ireland moves to fully electronic voting and counting, the Gregory method will be implemented.

The filling of vacancies caused by the resignation or death of an incumbent also varies across STV systems. In Ireland, casual vacancies are filled by by‐elections employing the alternative vote (which is also used for presidential elections). Even though this should in theory discriminate against small parties, in practice this factor is countered by the swing against large parties, especially if in government, that is characteristic of second‐order elections (Gallagher 1996a). In Tasmania and Malta, such vacancies are filled by the ‘countback’ method, examining the ballot papers of the vacating MP and performing a fresh count on these to ‘elect’ a replacement.

Political consequences of the electoral system

Impact on the party system

PR electoral systems are associated with multiparty systems, and Ireland does not flout this pattern. There have never been fewer than three parties represented in the Dáil, and there have been as many as eight. However, fractionalization has been relatively low, with the two major parties, Fianna Fáil and Fine Gael, often dominating the picture. These parties together have averaged 75 per cent of the votes at elections since 1932, and have twice (1938 and February 1982) won 85 per cent. In the 1980s, indeed, Ireland was cited by both Riker and (more cautiously) Duverger as constituting a ‘devastating counterexample’ to Duverger's law by displaying something close to a two‐party system despite having PR, something Duverger (p.521) attempted to explain away with the far‐fetched suggestion that the direct election of the president might have led to ‘the polarization of the citizens around the two large parties’ (Riker 1982: 758; Duverger 1986: 72, 74–5). However, the need to explain anything away soon disappeared: the two major parties together had exceeded 80 per cent of the votes at each of the five elections before Duverger's 1986 chapter, but at the following five elections (1987–2002) their average combined vote was only 68 per cent.

Still, even allowing for this, the Irish party system has never approached the levels of fractionalization that characterize, say, Belgium, Finland, or Switzerland. The effective number of parties in parliament (see Appendix B for the meaning of this concept) after the elections over the 1923–2002 period has averaged 3.0, which, as Table 26.1 shows, represents a relatively low level of fragmentation. Only once has the number come close to 5, and this was at a very early election when the party system was still settling down. Table 25.3 (effective number of elective parties column) also makes it clear why seats in parliament have been relatively ‘concentrated’—it is because votes, too, have this pattern.

The reduction in fractionalization—the ‘mechanical’ effect, as Duverger (1964: 224–6) terms it—under PR‐STV has clearly been only mild, as is characteristic of PR systems generally. Thus, it is reasonable to conclude that ‘PR‐STV in Ireland has…facilitated moderate multipartyism when other factors were leading in that direction’ (Sinnott 2005: 119). It is also worth noting that Malta, the only other country to use PR‐STV to elect its national parliament, has a virtually pure two‐party system: almost all the votes at elections are won by either the Nationalist Party or the Maltese Labour Party, and no other party has won a seat since independence in 1964.

Defractionalization means disproportionality—that is to say, the reason seat shares are less fragmented than vote shares is because some parties, usually the larger ones, get a ‘bonus’ from the system while others are under-represented. Sure enough, in Ireland it is the largest party, Fianna Fáil, that does best; with an average first preference vote of 43 per cent at elections from 1923 to 2002, its average share of the seats has been 47 per cent. This, though, is quite a modest bonus, and as a comparison between Tables 25.3 and 26.4 shows, disproportionality has been relatively low at Irish elections over the whole period since independence.

Table 25.3 Fractionalization and disproportionality at Irish elections, 1923–2002

Effective number of elective parties

Effective number of legislative parties

Disproportionality

Average

3.4

3.0

3.9

Maximum

5.7 (June 1927)

4.9 (June 1927)

6.6 (2002)

Minimum

2.6 (1938)

2.4 (1977)

1.7 (Feb 1982)

Note: N = 26. Dates in parentheses show the election concerned; there were two elections in 1927 and in 1982. Disproportionality is measured by the least squares index. The 1922 election is excluded because in a high proportion of constituencies there was no contest.

(p.522) This is very surprising given the widespread agreement, as already mentioned, that disproportionality increases as district magnitude decreases (Lijphart 1994: 117). Since district magnitude in Ireland is very small by the standards of PR elections, only about four, how does it come about that disproportionality has not been larger?

In order to understand this, we need to look at the way in which the electoral system affects interparty relations. Under PR‐STV, the tendency of the largest party to win a seat bonus creates its own countervailing force. The transferability of votes means that it is possible, and may be strategically sensible, for supporters of smaller parties to use their votes in such a way as to help each other and thereby prevent the larger ones reaping a sizeable bonus. In Ireland, this helps to explain the relatively low level of disproportionality. At all elections since 1932 the largest party, Fianna Fáil, has been perceived to have had some kind of chance of winning an overall majority of seats, and this gives supporters of other parties an incentive to deploy their lower preference votes against it, regardless of their respective policy positions. Indeed, even supporters of parties that might see Fianna Fáil as a potential coalition partner have an incentive to do this, in order to prevent Fianna Fáil securing an overall majority and thereby rendering their own party redundant in the government formation process (Laver 2000). This can be achieved under PR‐STV by ranking candidates of all other parties above the Fianna Fáil candidates on the ballot paper, and there is no doubt that many supporters of other parties have done precisely this at many elections.

In this way, the electoral system has helped to give a pronounced ‘Fianna Fáil against the rest’ shape to the Irish party system (Mair 1987: 36). At most elections from the late 1940s to the late 1980s there was clear evidence of this, with supporters of other parties using their lower preferences against Fianna Fáil even in the absence of explicit agreements between parties—the example of an STV count given in Appendix A is a good illustration of this. The effect of this has been to dampen the seat bonus that Fianna Fáil, as the largest party, could otherwise have expected.

It follows that if Fianna Fáil could transform the main fault line of the party system in such a way that it was no longer regarded as everyone's opponent, it would stand to extract much more advantage from the electoral system. In 1989 it did just this. For the first time it took part in a coalition government, having previously made a virtue of its opposition to the notion of coalitions per se, and indeed it also entered coalition governments after each of the following three elections. As a result, competition has become somewhat unstructured (Mair and Weeks 2005: 149). One party, the PDs, had an all‐but‐explicit transfer pact with Fianna Fáil before the 1997 election, and in 2002 virtually every party other than Fine Gael considered itself to be a potential coalition partner of Fianna Fáil (Mitchell 2003). Consequently, with the impact of vote transfers no longer working against Fianna Fáil, both overall disproportionality and Fianna Fáil's bonus have been rising. The figures for both reached all‐time highs in 1997, and these were themselves surpassed in 2002, when Fianna Fáil achieved a bonus of nearly 8 per cent (49 per cent of the seats for 41 per cent of the votes) and disproportionality reached 6.6 (least squares index).

(p.523) Another aspect of the party system that may be attributed partly to the electoral system is the persistence of independent TDs. The Dáil usually contains more independents than all other west European parliaments put together; in 2002 no fewer than thirteen were elected in the 166‐member legislature, the highest number for over fifty years. This illustrates the very low barriers to entry provided by PR‐STV; a party or a candidate that can muster perhaps 10 per cent of the votes in a single constituency can earn Dáil representation. In 2002, an independent whose first preference support amounted to just 0.2 per cent of the national vote total was elected. The reason why independents fare well under PR‐STV was identified nearly a century ago by Meredith (1913: 79–80), who foresaw that the independents elected might not be the talented and principled individuals hoped for by proponents of the system. Instead, the independents to prosper would be

those of far less ability who in a small way have moved in a somewhat local ‘limelight’. They form good compromise candidates for the last seat. They have a small band of admirers that would give them sufficient first preferences to enable them to keep their flags flying until late preferences were reached. They would begin to accumulate the votes of electors of each party who prefer them to members of the opposite party. Now, if the tendency of the single transferable vote is to place the balance of power in the hands of such candidates, is it really an argument in its favour?

Some argue that these concerns have been borne out. For example, the 1997–2002 government was a minority coalition sustained by the support of four independent TDs who demanded ‘pork’ (public spending) in their constituencies as the price of their continued support (Mitchell 2001: 205). Sinnott (2005: 120) concludes that PR‐STV ‘does increase the probability of government reliance on independent deputies whose support may be delivered only at a disproportionate price and even then may not be durable’.

Impact on the parties

Intraparty electoral competition is inherent in PR‐STV. The two major parties both run, on average, two or three candidates per constituency, and these candidates are competing for votes with each other as much as, perhaps more than, with candidates from other parties. They are competing with each other for the first preferences of voters who are committed to the party; for the first preferences of floating voters; and for the lower preferences of voters who support other parties. Thus, unlike candidates under an open list system such as those discussed in Chapters 19–24, they are not competing only for support from the pool of party voters; they have an incentive to try to be liked by everyone. Thus there is definitely such a thing as a ‘personal vote’: even if most of the support for a TD comes from within the party fold, TDs also draw on support from voters who do not support the party per se. The smaller parties rarely run more than one candidate per constituency, so they do not experience internal electoral competition, but their sole candidates, too, aim to attract lower preferences from as many voters as possible.

(p.524) PR‐STV cannot be said to have led to disunited parties in Ireland. In theory, it might seem that major party TDs, secure in their electoral base, would respond to local pressures as much as to the party line in parliamentary votes. In practice, though, the Irish parliamentary parties are highly disciplined and by any conventional test are among Europe's most cohesive parties, despite suggestions from some writers that this apparent unity is only ‘superficial’ (for discussion see Gallagher 2000: 108–10; Sinnott 2005: 119–20).

This is not to suggest that internal party life is all sweetness. At local level, candidates and TDs of the two main parties know that they have rivals within their own party as well as in other parties. Over the years, around 56 per cent of Fianna Fáil TDs, and 37 per cent of Fine Gael TDs, who suffer defeat at an election lose to a running mate rather than to a candidate of another party (Gallagher 2000: 97). This internal rivalry expresses itself not in candidates' staking out distinct policy positions but in their seeking to earn reputations as assiduous and effective constituency workers when it comes both to casework for individual constituents and to activities as a ‘local promoter’ to secure resources for the constituency (Gallagher and Komito 2005). Consequently, constituency election campaigns are often characterized by ‘turf wars’, whereby each candidate stakes a claim to the territory around their home base and the neutral territory is fought over (for examples at the 2002 election, see Gallagher 2003: 108–10). Intraparty competition is openly acknowledged and discussed by politicians in Ireland, which is not the case in all preferential‐list PR countries (see Chapters 19–24).

The personal vote built up by TDs over the years gives them a powerful position in the candidate selection process (Galligan 2003; Marsh 2005: 172–5), one cornerstone of which is that incumbents very rarely fail to secure reselection. In part this is because other actors in the process do not wish to jeopardize the personal votes that an established TD will attract but, beyond this, TDs like to minimize any risk of an upset by building up networks of personal loyalty within the local party organization. A study of Fine Gael found that around 45 per cent of its members regarded themselves as a strong supporter of one particular local Fine Gael politician rather than as supporters of all the local Fine Gael politicians (Gallagher and Marsh 2002: 106–13).

However, this rivalry must be, and is, accommodated within the framework of the party (Marsh 2000). Whereas under the former Japanese electoral system, the single non‐transferable vote, LDP candidates could openly regard each other with animosity, under PR‐STV party candidates must appeal for personal support while simultaneously behaving as a team. The reason is that when one candidate is eliminated from the count, or has a surplus distributed, it is important that as many as possible of his or her votes transfer to the party's other candidates. Thus candidates usually urge their supporters to ‘Vote No 1 for me and continue your preferences for my running mates’, and most votes, when transferred, do remain within the party fold. In the past this tendency was strong; prior to 1992, Fianna Fáil's internal transfer solidarity rate was over 80 per cent. It has been dropping recently, though, and in 2002 only 63 per cent of votes from one Fianna Fáil candidate passed to another (p.525) Fianna Fáil candidate where this was possible; the solidarity figure for Fine Gael was 64 per cent (Gallagher 2003: 105).6 Still, the salience of government formation as an issue at general elections, together with the widespread assumption that behaviour by a TD that earns the party an extra seat is more likely to earn promotion to office than candidate‐centred behaviour that costs the party a seat, combine to ensure that candidates operate as team players, albeit personally motivated ones. When these factors are not present—for example, at second‐order elections—entirely individualistic behaviour becomes more likely, and examples are regularly to be found at Ireland's European Parliament elections.

Impact on parliament

The backgrounds of members of the Dáil are not especially different from those of other west European parliaments. Most TDs have very strong local connections: they were born and grew up in their constituencies, and before being elected to the Dáil they were county councillors. Impeccable local roots are a strong resource given the high level of constituency attention expected from TDs. Strikingly, a high proportion (usually 20–25 per cent) of TDs are following in the footsteps of a close relative; in the Dáil elected in 2002, for example, thirty-seven of the 166 TDs (nineteen of whom were sons of previous TDs) were closely related to a former TD, usually denoting the ‘inheriting’ of a political base. The proportion of women is relatively low (only 13 per cent in 2002). There are various reasons for this (Galligan 2005), and while it might be argued that, if conservative public attitudes in a predominantly Catholic country are a main cause, then PR‐STV is unhelpful by giving more weight to these attitudes than, say, a closed list system would, there is in fact no reason to suppose that matters would be much different under a different electoral system.

The behaviour of TDs is often perceived as involving a primary focus on constituency affairs with only a secondary interest in the national political issues discussed in parliament. While this is an over-generalization, and parliamentary committees in particular have become more active in recent years (Gallagher 2005), this has emerged as the main line of criticism of PR‐STV in Ireland in recent years. The gravamen is

  1. • first, that PR‐STV imposes patterns of behaviour that TDs must follow for their own electoral survival, promoting a focus on constituency work and close attention to intraparty manoeuvring (‘a time‐consuming and unproductive drudgery’, in the words of one former minister—Hussey 1993: 57–61);

  2. (p.526)
  3. • second, that these patterns weaken the national parliament and, because under the Irish constitution all ministers are required to be members of parliament, they also interfere with the business of government; and

  4. • third, that knowledge of the workload involved in being a TD, as well as hostility from established politicians who see talented newcomers as a threat, discourages many able people from entering politics, resulting in a low quality of political leadership.7

On the other hand, others take the view that

  1. • close links between MPs and constituents are not necessarily bad and that, anyway, these links are demanded by constituents and cannot be attributed in significant measure to the electoral system;

  2. • ministers have large staffs to do their constituency work and are thus not personally greatly discommoded by the personal demands of their constituents;

  3. • there is no evidence that Irish politicians are of a ‘lower standard’ than their counterparts anywhere else, and that, while intraparty competition constitutes a threat to incumbents, it also offers newcomers an opportunity to break through.8

Even though it has been suggested that ‘the very high turnover of members also discourages potential candidates’ (FitzGerald 2003: 93), turnover is not particularly high. At the elections of the 1927–97 period, around 75 per cent of outgoing TDs, and 82 per cent of non‐retiring TDs, secured re‐election; the figures for Malta are very similar (Gallagher 2000: 94). By comparative standards, this does not amount to high turnover (Matland and Studlar 2004: 93). As already noted, around a third of all defeats, and a half of major party defeats, in Ireland come at the hands of running mates; in Malta the figure is around two‐thirds (Gallagher 2000: 97–8). It cannot, then, be argued that high turnover is a significant factor in political recruitment.

Government formation

As a result of its electoral system, Ireland has a multiparty system and much experience of coalition government. Although all governments up to 1948 were single‐party—a characteristic of the Westminster model—between 1948 and June 2002 coalitions were in office for 52 per cent of the time. No party has won an overall majority since 1977 or been able to form a single‐party government since 1987 (Mitchell 2000). Moreover, minority government is increasingly common, occupying office for 34 per cent of the time from 1948 to 2002 (Mitchell 2001).

PR‐STV provides an incentive for the parties to identify their chosen coalition allies in advance of the election, so that party leaders can urge their supporters to award lower preferences to allied parties. However, this does not always happen. (p.527) Sometimes two or more parties explicitly form a pre‐election partnership, as occurred in 1973, 1989, and 1997, for example—but only in 1997 did voters have a choice between competing identifiable coalition alternatives. On other occasions, even though party leaders act coyly, it is clear to voters which party is the closest ally of their own, and vote transfers flow accordingly, as in 1981 and 1982. But at some elections there are no alliances, even implicit ones, between parties, and voters cannot be sure what coalition combinations are possible. This happened most strikingly in 1992, when Labour criticized the outgoing Fianna Fáil government vehemently but after the election went into coalition with it, to the horror of many of its voters. It might also have happened in 2002, when the outgoing Fianna Fáil–PD government was the only identifiable government, and had these two parties not won enough seats to continue in office, the government to emerge would have been one not offered to the people in advance. Thus, parties have become both more coalitionable and also more ‘promiscuous’ (Mair and Weeks 2005: 155). Parties have an incentive to form pre‐election alliances in order to attract vote transfers from supporters of their allies, but at the same time the disincentive is that this is likely to reduce the flow of transfers from other parties, and there is also a risk that in the post‐election situation the distribution of seats might mean that a party might want to form a coalition with one of its erstwhile opponents. Consequently, PR‐STV does not in practice always help identify the governmental options in advance of an election in the way it might be expected to, and the identifiability of future governments varies more than in most countries (Powell 2000: 73–6).

The politics of electoral reform

Because the electoral system is entrenched in the Irish constitution, and because this constitution cannot be amended except by referendum, PR‐STV enjoys a level of protection against its critics. Without this, it might well not have survived. As we have seen, the largest party in the state, Fianna Fáil, twice initiated referendums designed to replace it by SMP, failing only narrowly on the first occasion. These moves were primarily partisan in motivation—since Fianna Fáil, as the largest party, stood to benefit greatly from the adoption of a plurality system—and as such generated opposition from all other parties and interests, which sufficed to defeat them.

In recent years, the main focus of discussion has been an aspect that did not feature at all in the earlier referendums: the supposed impact of PR‐STV on the constituency work and legislative duties of TDs. As we have seen, several critics of the electoral system argue that it has a number of dysfunctional effects and should be replaced by one that does not include intraparty electoral competition. In the mid‐1990s a heavyweight committee that assessed the entire constitution reviewed the question. It advised that no change should be made without ‘careful advance assessment of the possible effects’ and that, if a new system were adopted, the introduction of a list or mixed system would be better than a non‐PR system (Constitution Review Group 1996: 60). A couple of years later, an academic analysis (p.528) argued that a mixed system was the only feasible alternative and suggested that one side‐effect would be that Fianna Fáil would probably win the great majority of the constituency seats (thus perhaps being left with most of the constituency work) while the TDs of other parties, being mostly elected from lists, would be free to concentrate on parliamentary work (Laver 1998).

The Fianna Fáil Minister for the Environment (whose brief covers elections) appointed in 1997, Noel Dempsey, was a long‐standing critic of PR‐STV, and at his instigation the government asked the parliamentary All‐Party Committee on the Constitution to consider reform of the electoral system ‘as a matter of urgency’ early in 2000. Dempsey, together with former Fine Gael prime minister Garret FitzGerald, addressed the committee to argue the case for a move towards the German system, or something very much like it. The committee had already conducted a survey of members of the Oireachtas (TDs and senators), and this found that most did not favour change. Of the 38 per cent who replied to a questionnaire, 69 per cent wanted to retain PR‐STV with only 26 per cent wanting to replace it (the others expressed no opinion) (All‐Party Oireachtas Committee 2002: 18). The TDs who favoured the status quo emphasised the value of close contact between deputies and their constituents and of the voters' right to choose who their representatives should be. Critics (some of whom favoured SMP or the alternative vote rather than a mixed system) wanted to eliminate intraparty electoral competition and argued that this would improve the calibre of parliamentary work and of politicians (All‐Party Oireachtas Committee 2002: 18–19). The only parties to express a clear view were the PDs, the junior coalition partner at the time (in favour of change) and Labour (in favour of retaining PR‐STV). Having reviewed the various submissions to it, the committee concluded:

The committee…is not convinced that the weaknesses of PR‐STV are as considerable as might be claimed, or, put otherwise, that PR‐STV is itself responsible for all of the failings that have been laid at its door.…Finally, and decisively, there is no evidence of serious or widespread public discontent with the existing system: on the contrary, there is in our view a strong and enduring attachment to it. The fundamental and insurmountable argument against change is that the current Irish electoral system provides the greatest degree of voter choice of any available option. A switch to any other system would reduce the power of the individual voter. For all of these reasons, we recommend against any change in this aspect of the Constitution. (All‐Party Oireachtas Committee 2002: 29)

This, coupled with the formation of a new government a few months later in which Dempsey was moved to a different portfolio, defused the issue, and while it will no doubt be raised again, the committee's last point is telling. Any other electoral system would reduce the power of Irish voters, and increase the power of party hierarchs, to determine which individuals should represent them in parliament, and the current system cannot be changed without the express agreement of those voters in a referendum. The likelihood of the electorate's voting to give itself less power seems slim. Recent survey evidence on public attitudes is confined to a poll in January 2000 that bore out the committee's view that the electorate is attached to the (p.529) present system; it found that only 24 per cent wanted a new electoral system while 52 per cent favoured the retention of PR‐STV.9

Conclusion

The distinctiveness of Ireland's electoral system makes it a fascinating object of study but, at the same time, opens up the danger of attributing far too much causal power to it. Reformers in various countries, faced with evidence that the political system is not working well, have found the ‘change the electoral system’ button a tempting one to press in the hope of putting things right. Sometimes, though, the problems that electoral system reform were supposed to eliminate persist, sometimes they re‐emerge in a new form, and sometimes entirely new problems appear. If there is any lesson to be learned, it is that many features of political behaviour have roots that run far deeper than a single institution such as the electoral system.

Under PR‐STV, Ireland has a multiparty system and coalition government has become the norm. The proportion of women in parliament is among the lowest in Europe. MPs are extremely loyal to the party line in parliament and spend a high proportion of their time engaged in constituency work rather than in the scrutiny of legislation or in parliamentary committee work.

Clearly, we have no way of knowing how the Irish political process would now look if the country had had a non‐PR system or a mixed system. Perhaps things would be very different. More probably, though, some of those features most highlighted by critics of PR‐STV would be just as pronounced under any electoral system. In particular, the emphasis placed by MPs upon constituency work has many causes besides the electoral system, not least the expectations of the voters; nineteenth‐century Irish MPs were assiduous in their attention to local grievances, long before PR‐STV was even invented. TDs would be placed under intense pressure to discharge constituency duties whatever the electoral system, and if PR‐STV were replaced by an electoral system that did not permit intraparty electoral competition, this intraparty rivalry would simply be shifted to the candidate selection stage, at which a strong local reputation acquired through casework might well be a powerful resource. Whether things really would be different under another electoral system will not be known unless PR‐STV is abandoned. Since the approval of the voters would be needed for this, however, the likelihood of Ireland replacing its electoral system in the near future seems low. PR‐STV is looked upon warmly not only by electoral systems specialists; the Irish electorate too seems to be persuaded of its enduring charm.

(p.530) REFERENCES

The Politics of Electoral Systems Ireland

Bibliography references:

All‐Party Oireachtas Committee on the Constitution (2002). Seventh Progress Report: Parliament. Dublin: All‐Party Oireachtas Committee on the Constitution.

Austen‐Smith, D. and Banks, J. (1991). ‘Monotonicity in Electoral Systems’, American Political Science Review, 85/2: 531–7.

Boland, J. (1991). ‘Dáil Can Only be Reformed if TDs are Liberated from Multi‐seat Constituencies’, Representation, 30/111: 42–3.

Bowler, S. and Grofman, B. (2000). ‘Conclusion: STV's Place in the Family of Electoral Systems’, in S. Bowler and B. Grofman (eds.), Elections in Australia, Ireland, and Malta under the Single Transferable Vote: Reflections on an Embedded Institution. Ann Arbor: University of Michigan Press, 265–70.

Carty, R. K. (1981). Party and Parish Pump: Electoral Politics in Ireland. Waterloo, ON: Wilfrid Laurier University Press.

Chubb, B. (1982). The Government and Politics of Ireland, 2nd edn. London: Longman.

Coakley, J. (1994). ‘The Election that Made the First Dáil’, in B. Farrell (ed.), The Creation of the Dáil. Dublin: Blackwater Press, 31–46.

—— and Gallagher, M. (eds.) (2005). Politics in the Republic of Ireland, 4th edn. Abingdon: Routledge and PSAI Press.

Constitution Review Group (1996). Report of the Constitution Review Group. Dublin: Stationery Office.

Dummett, M. (1997). Principles of Electoral Reform. Oxford: Oxford University Press.

Duverger, M. (1964). Political Parties, 3rd edn. London: Methuen.

—— (1986). ‘Duverger's Law: Forty Years Later’, in B. Grofman and A. Lijphart (eds.), Electoral Laws and their Political Consequences. New York: Agathon Press, 69–84.

Farrell, D. M. and Gallagher, M. (1999). ‘British Voters and Their Criteria for Evaluating Electoral Systems’, British Journal of Politics and International Relations, 1/3: 293–316.

—— and McAllister, I. (2003). ‘The 1983 Change in Surplus Vote Transfer Procedures for the Australian Senate and its Consequences for STV’, Australian Journal of Political Science, 38/3: 479–91.

——, Mackerras, M., and McAllister, I. (1996). ‘Designing Electoral Institutions: STV Systems and Their Consequences’, Political Studies, 44/1: 24–43.

FitzGerald, G. (2003). Reflections on the Irish State. Dublin: Irish Academic Press.

Gallagher, M. (1987). ‘Does Ireland Need a New Electoral System?’, Irish Political Studies, 2: 27–48

—— (1996a). ‘By‐Elections to Dáil Éireann: The Anomaly that Conforms’, Irish Political Studies, 11: 33–60.

—— (1996b). ‘Electoral Systems’, in Constitution Review Group, Report of the Constitution Review Group. Dublin: Stationery Office, 499–520.

—— (2000). ‘The (Relatively) Victorious Incumbent under PR‐STV: Legislative Turnover in Ireland and Malta’, in S. Bowler and B. Grofman (eds.), Elections in Australia, Ireland, and Malta under the Single Transferable Vote: Reflections on an Embedded Institution. Ann Arbor: University of Michigan Press, 81–113.

—— (2003). ‘Stability and Turmoil: Analysis of the Results’, in M. Gallagher, M. Marsh, and P. Mitchell (eds.), How Ireland Voted 2002. Basingstoke: Palgrave Macmillan, 88–118.

—— (2005). ‘Parliament’, in J. Coakley and M. Gallagher (eds.), Politics in the Republic of Ireland, 4th edn. Abingdon: Routledge and PSAI Press, 211–41.

(p.531) —— and Komito, L. (2005). ‘The Constituency Role of Dáil Deputies’, in J. Coakley and M. Gallagher (eds.), Politics in the Republic of Ireland, 4th edn. Abingdon: Routledge and PSAI Press, 242–71.

—— and M. Marsh (2002). Days of Blue Loyalty: The Politics of Membership of the Fine Gael Party. Dublin: PSAI Press.

—— and A. R. Unwin (1986). ‘Electoral Distortion under STV Random Sampling Procedures’, British Journal of Political Science, 16/2: 243–53.

Galligan, Y. (2005). ‘Women in Politics’, in J. Coakley and M. Gallagher (eds.), Politics in the Republic of Ireland, 4th edn. Abingdon: Routledge and PSAI Press, 272–301.

—— (2003). ‘Candidate Selection: More Democratic or More Centrally Controlled?’, in M. Gallagher, M. Marsh, and P. Mitchell (eds.), How Ireland Voted 2002. Basingstoke: Palgrave Macmillan, 37–56.

Garvin, T. (2001). ‘Hogan as Political Scientist: Representative Government in Ireland’, in D. Ó Corráin (ed.), James Hogan: Revolutionary, Historian and Political Scientist. Dublin: Four Courts Press, 177–84.

Hart, J. (1992). Proportional Representation: Critics of the British Electoral System 1820–1945. Oxford: Oxford University Press.

Hicken, A. and Kasuya, Y. (2003). ‘A Guide to the Constitutional Structures and Electoral Systems of East, South and Southeast Asia’, Electoral Studies, 22/1: 121–51.

Hogan, J. (1945). Election and Representation. Cork: Cork University Press and Oxford: B. H. Blackwell.

Hussey, G. (1993). Ireland Today: Anatomy of a Changing State. London: Viking.

Ishiyama, J. T. (1994). ‘Electoral Rules and Party Nomination Strategies in Ethnically Cleaved Societies’, Journal of Communist and Postcommunist Studies, 27/2: 177–92.

Katz, R. S. (1984). ‘The Single Transferable Vote and Proportional Representation’, in A. Lijphart and B. Grofman (eds.), Choosing an Electoral System. New York: Praeger, 135–45.

Laakso, M. and Taagepera, R. (1979). ‘ “Effective” Number of Parties: A Measure with Application to West Europe’, Comparative Political Studies, 12/1: 3–27.

Laver, M. (1998). A New Electoral System for Ireland? Dublin: All‐Party Oireachtas Committee on the Constitution.

—— (2000). ‘STV and the Politics of Coalition’, in S. Bowler and B. Grofman (eds.), Elections in Australia, Ireland, and Malta under the Single Transferable Vote: Reflections on an Embedded Institution. Ann Arbor: University of Michigan Press, 131–52.

—— (2004). ‘Analysing Structures of Party Preference in Electronic Voting Data’, Party Politics, 10/5: 521–41.

Lijphart, A. (1994). Electoral Systems and Party Systems: A Study of Twenty‐Seven Democracies, 1945–1990. Oxford: Oxford University Press.

—— (1999). Patterns of Democracy: Government Forms and Performances in Thirty‐Six Countries. New Haven, CT and London: Yale University Press.

Mair, P. (1987). The Changing Irish Party System: Organisation, Ideology and Electoral Competition. London: Frances Pinter.

—— and Weeks, Liam (2005). ‘The Party System’, in J. Coakley and M. Gallagher (eds.), Politics in the Republic of Ireland, 4th edn. Abingdon: Routledge and PSAI Press, 135–59.

Marsh, M. (2000). ‘Candidate Centered but Party Wrapped: Campaigning in Ireland under STV’, in S. Bowler and B. Grofman (eds.), Elections in Australia, Ireland, and Malta under the Single Transferable Vote: Reflections on an Embedded Institution. Ann Arbor: University of Michigan Press, 114–30.

(p.532) —— (2005). ‘Parties and Society’, in J. Coakley and M. Gallagher (eds.), Politics in the Republic of Ireland, 4th edn. Abingdon: Routledge and PSAI Press, 160–82.

Matland, R. E. and Studlar, D. T. (2004). ‘Determinants of Legislative Turnover: A Cross‐National Analysis’, British Journal of Political Science, 34/1: 87–108.

Meredith, J. C. (1913). Proportional Representation in Ireland. Dublin: University Press, Trinity College.

Mitchell, P. (2000). ‘Ireland: From Single‐Party to Coalition Rule’, in W. C. Müller and K. Strøm (eds.), Coalition Governments in Western Europe. Oxford: Oxford University Press, 126–57.

Mitchell, P. (2001). ‘Divided Government in Ireland’, in R. Elgie (ed.), Divided Government in Comparative Perspective. Oxford: Oxford University Press, 182–208.

—— (2003). ‘Government Formation in 2002: “You Can Have Any Kind of Government as long as it's Fianna Fáil” ’, in M. Gallagher, M. Marsh, and P. Mitchell (eds.), How Ireland Voted 2002. Basingstoke: Palgrave Macmillan, 214–29.

Nurmi, H. (1996). ‘It's not Just the Lack of Monotonicity’, Representation, 34/1: 48–52.

O'Leary, C. (1961). The Irish Republic and its Experiment with Proportional Representation. Notre Dame, IN: University of Notre Dame Press.

—— (1975). ‘Ireland: The North and South’, in S. E. Finer (ed.), Adversary Politics and Electoral Reform. London: Anthony Wigram, 153–83.

—— (1979). Irish Elections 1918–1977: Parties, Voters and Proportional Representation. Dublin: Gill and Macmillan.

—— (2001). ‘Hogan: The Development of a Political Scientist’, in D. Ó Corráin (ed.), James Hogan: Revolutionary, Historian and Political Scientist. Dublin: Four Courts Press, 163–76.

Powell, G. B. Jr (2000). Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven, CT and London: Yale University Press.

Riker, W. H. (1982). ‘The Two‐Party System and Duverger's Law: An Essay on the History of Political Science’, American Political Science Review, 76/4: 753–66.

Rose, R. and Munro, N. (2003). Elections and Parties in New European Democracies. Washington, DC: CQ Press.

Sinnott, R. (1995). Irish Voters Decide: Voting Behaviour in Elections and Referendums since 1918. Manchester: Manchester University Press.

—— (2005). ‘The Electoral System’, in J. Coakley and M. Gallagher (eds.), Politics in the Republic of Ireland, 4th edn. Abingdon: Routledge and PSAI Press, 105–34.

Taagepera, R. (1996). ‘STV in Transitional Estonia’, Representation, 34/1: 29–36.

—— (2001). ‘Party Size Baselines Imposed by Institutional Constraints: Theory for Simple Electoral Systems’, Journal of Theoretical Politics, 13/4: 331–54.

—— and Shugart, M. S. (1989). Seats and Votes: The Effects and Determinants of Electoral Systems. New Haven, CT and London: Yale University Press.

Tideman, N. and Richardson, D. (2000). ‘A Comparison of Improved STV Methods’, in S. Bowler and B. Grofman (eds.), Elections in Australia, Ireland, and Malta under the Single Transferable Vote: Reflections on an Embedded Institution. Ann Arbor: University of Michigan Press, 248–64.

Weeks, L. (2003). ‘Appendices’, in M. Gallagher, M. Marsh, and P. Mitchell (eds.), How Ireland Voted 2002. Basingstoke: Palgrave Macmillan, 247–67.

Notes:

(1) For the Estonian experience see Taagepera 1996; Rose and Munro 2003: 167–8; Ishiyama 1994.

(2) In a second referendum held on the same day in 1968, a proposal to relax the constitutional requirement that the ratio of population to seats be the same across the country was defeated by the same margin.

(3) Indeed, quite apart from the rules followed in practice, aficionados have devised and vigorously promote a range of far more complex methods of vote‐counting, particularly when it comes to distributing surpluses (Tideman and Richardson 2000).

(4) Space does not allow a full explanation of the circumstances in which vote management might be beneficial (for some discussion and examples see Gallagher 2003: 108–10). Vote management schemes have been employed by the Irish parties virtually since the state was founded, and indeed were reported as early as the Johannesburg municipal elections of 1909 (Meredith 1913: 77).

(5) To add a further complication, there are several variants of the Gregory method, each of which has its advocates, and these variants can lead to different outcomes—see Farrell and McAllister (2003).

(6) In Malta, internal transfer solidarity for both main parties is extremely high (almost 100 per cent), and the electoral system was modified in the 1980s in a way that assumes that all votes are party votes. Thus, if the party with a plurality of the votes fails to win a majority of the seats, it is awarded additional seats so as to give it an overall majority (see Chapter 3, p. 65). These additional seats are sometimes seen as constituting a ‘higher tier’, ‘putting right’ an anomaly caused by the second largest party winning more seats than it was entitled to (i.e. obtaining überhangmandate) in the constituencies.

(7) Some or all of these criticisms can be found in, for example, Boland (1991); Carty (1981: 109–39); FitzGerald (2003: 92–3).

(8) Examples of such cautious conclusions can be found in Sinnott (2005: 117–25); Gallagher (1987); Gallagher (1996b).

(9) MRBI/Irish Times poll, MRBI 5122/00. The other 24 per cent of respondents said that they had no opinion or that they felt they did not know enough to make a decision. There was no significant subgroup variation.