Research Paper 4 2002-03

Research Paper no. 4 2002-03

Politician Overboard: Jumping the Party Ship

Sarah Miskin
Politics and Public Administration Group
24 March 2003

Contents

Executive Summary
Introduction
Section 1: Defections

Executive Summary

'Give back our seat' is a common cry in Australia when a Member or Senator chooses to quit the party on whose platform s/he was elected while retaining her or his seat in parliament. To hold on to the seat while joining another party or sitting alone as an independent is seen as 'little short of a fraud upon the voters'.⁽¹⁾ That is, such an action is seen as impairing democracy in that the decision of voters as expressed at a democratic election has been overturned.

In discussions of what should be done about party jumping, some argue that the Australian party system has developed to the point where voters no longer vote for candidates personally; rather, they vote for parties. In addition, an individual is unlikely to be elected without the party label, the party money, the party workers and the party preference deals. Therefore, an elected Member or Senator has no right to a seat because it 'belongs' to the party.

However, others argue that the seat rightly belongs to the voters, who want their representative to exercise judgment on their behalf-and that means defecting if the representative believes this is in the voters' best interests.

The question of seat 'ownership' is important because party jumping can have significant consequences-Senator Meg Lees's decision to quit the Australian Democrats to sit as an independent in July 2002 changed the political dynamics of the Senate and may yet allow the Howard Government to pass legislation that it could not pass before the defection.

Given the potential consequences of defecting, should there be legislation to control it as has been suggested in editorials commenting on recent Australian defections and as New Zealand decided to do in December 2001?⁽²⁾ Both arguments outlined above-party ownership versus representative ownership-do not preclude the option of forcing a defecting politician to surrender the seat and put her or his decision and the justifications for it to an immediate test in a by-election. In fact, going back to the electorate is a common demand in the wake of a defection. Most recently, South Australian MP Kris Hanna's party jump from the Australian Labor Party to the Greens in February 2003 prompted a demand that 'Australia's electoral laws ... be changed to compel politicians to stand aside from parliaments in such circumstances'.⁽³⁾

But is it possible-and more importantly, is it desirable-to control such party jumping? Does legislation to force party jumpers to relinquish their seats when they quit their parties contravene basic freedoms, such as those of conscience and association, or does it ensure that the voters' will as expressed at an election is maintained throughout the legislative term? Would such a law resolve the alleged democratic deficit or would it create new problems?

At its most basic, an anti-defection law would force those politicians who quit their parties to quit parliament. But an immediate problem arises: if an MP knows that the price of defecting is leaving the legislature and in all likelihood not returning-because, historically, few defectors are re-elected-then why would s/he resign?⁽⁴⁾ Such a law may provide the incentive to stay within the party and agitate-to be a 'politician-in-exile', so to speak.⁽⁵⁾ Of course, a party can expel a dissenting, disloyal politician from its ranks, but would this count as a defection in terms of the legislation?

To be effective, the law would have to allow for expulsion from the party to count as a defection, but this opens a new question about the power of parties. If voters elect their favoured candidate to a seat, is it democratic to allow a party to expel that person and force a by-election if s/he dissents from the party line? Does this give a party a level of power that would stop politicians from exercising their judgment and speaking against bad policy? What if the politician is simply following the will of her or his voters, who may disagree with a specific national policy? And if dissent is acceptable in this case, then when might it not be acceptable and who is to judge?

Before looking at the viability of legislation to prevent party defections, it is worth examining some more general issues. Why, for example, are there not more defections? One reason is that parties, not individual politicians, have become the central organisational pillars of today's democracies. Politics without established parties can descend into chaos, as has been the experience in Papua New Guinea. However, another reason that more politicians do not defect is that there are considerable tangible and intangible benefits to party membership, including electoral, collective, institutional and distributive advantages. That said, on the other side of the ledger, politicians must pay for these advantages by relinquishing some of their freedom to make independent choices on policy issues. Rather, in most cases, they must obey the dictates of the party.

Members of the legislature carry the burden of multiple responsibilities, being responsible to their voters, their party, the 'greater good' and their own consciences. Generally, they are able to balance the oft-competing demands that arise from these different responsibilities. However, occasionally one of these areas may come to outweigh the others, and when that occurs, a politician may choose to defect to another party or to independence while retaining the seat in the legislature.

Defectors advance in the public arena a number of common reasons to justify their action, especially their 'right' to keep the seat. These reasons include: that their party has changed while the defector has not, that they are obeying the voters' will, or that they are exercising judgment on behalf of the greater good. These reasons are not mutually exclusive, nor are they exhaustive. In fact, cynical observers would suggest that there are often unspoken self-interested motivations behind the defection.

As noted above, the question of seat 'ownership' has convincing arguments that can be advanced on both sides. In the absence of legislation, there appears to be no correct answer-it may be that, like the children's game of musical chairs, the seat 'belongs' to whoever is sitting in it at the time.

Evidence suggests that the consequences of party jumping for politicians are two-fold: they substantially change their behaviour in order to fit with the demands of the new party (if switching to another party), but ultimately their level of support usually falls. In short, defection brings 'short-term fame but long-term obscurity'. (There are, of course, exceptions to this generalisation as will be canvassed in the paper.)

Although defections in Australia are usually accompanied by a hue and cry about the party jumper's right to retain the seat, there is nothing beyond this expression of outrage that can be done-Australia has no law to control what should happen when a defection occurs.

It is rare for a country to have legislation to control party jumping. As this paper reveals, while anti-defection law may suit the circumstances of countries, such as Papua New Guinea where it may help to impose order on a chaotic party system in order to stabilise government, its usefulness in other countries is less clear. Elsewhere, the legislation is problematic at best and unworkable at worst. India has experienced more party jumping since introducing anti-defection legislation than it encountered before; New Zealand's legislation, implemented in December 2001, has been a dismal failure, unable to cope with even the most obvious cases of 'waka [canoe]-hopping'.

In short, a law to cover such instances is not a sensible option in Australia, especially given the low numbers concerned, and is likely to create more problems than it resolves. Such a law would not necessarily enhance democracy or lead to more democratic outcomes.

Introduction

'I've had enough': with these words, Senator Meg Lees quit the Australian Democrats on 26 July 2002, adding that it was in 'the best interests of the Australian Democrats that I leave and sit on the cross benches as an independent senator for South Australia'.⁽⁶⁾

Her decision brought immediate calls for her to resign from the Senate and 'give back' her seat to the Democrats. The argument was that South Australian voters had not voted for Lees personally, but for the Democrats, which meant that Lees had no right to her Senate seat because it 'belonged' to the party. The party also emphasised that Lees had signed a pledge to resign from the seat if she ever quit the party.

However, Lees claimed an equal right to keep the seat on the basis that it was the party that had shifted and that her principles had not changed.⁽⁷⁾ Hence, any pledge that she may have signed was not binding because the party had broken the contract first. Building on this argument, Lees contended that she more accurately reflected the views of South Australian voters and '[a] solid base of 40 per cent [of the party] are in tune with my way of thinking'.⁽⁸⁾

With no legal way to force Lees to surrender the seat, the Democrats have been able to do nothing but grumble. South Australian voters will not get to express their own views on the seat until a few months before Lees's six-year term expires on 30 June 2005 (unless there is a double dissolution).

Quitting the party under whose banner a politician has been elected to federal parliament while retaining the seat is not new. Since federation, 114 Members and Senators have left their parties but remained in parliament and 13 of these have done so more than once.⁽⁹⁾ (These figures will be explored in a separate, forthcoming Parliamentary Library publication on the history of party defections in Australia. The purpose of this paper is to discuss party jumping in general, using Australian and overseas examples. While defection also occurs at the state level, these cases will not be examined here because the focus is on federal politics. That said, the general points are applicable to both levels of politics.)

Nor is quitting the party while retaining the seat a phenomenon confined to Australia. In some countries, party switching is a common political behaviour. Well-known politicians who have quit their party to join the ranks of another include Winston Churchill, who entered Parliament in 1901 as a Conservative, but crossed to the Liberal benches in 1904-a move that led some to refer to him as 'the Blenheim rat'.⁽¹⁰⁾ (Churchill later 'reratted' and returned to the Conservatives.)

What makes the Lees defection worthy of particular note is its potential impact on Australian politics. In a single move-quitting her party to sit as an independent-Lees changed the political dynamics of the Senate, wresting away from the Democrats the king-making ability the party had had for 19 of the past 22 years.⁽¹¹⁾ Today, if the Howard Government can attract the support of all three independents (Lees, Brian Harradine and Shayne Murphy) and One Nation Senator Les Harris, then it has more votes than Labor, the Democrats and the Greens combined. This makes it possible for the Coalition Government to proceed with legislation that previously it may have had little chance of having passed.

Again, the potential for defections from political parties to have significant consequences is not peculiar to Australia. In the United States in May 2001, Vermont Senator Jim Jeffords changed the balance of power in the Senate when he abandoned the Republicans to sit as an independent, albeit one who caucuses with the Democrats.⁽¹²⁾ In New Zealand in 1997 and 1998, party jumpers kept the minority National government in power for several months, a move that angered the Opposition parties so much that they pledged to introduce legislation to ban what became known by the derogatory term 'waka [canoe] hopping'.⁽¹³⁾

Thus, the issue of a politician jumping overboard from the party ship while keeping the seat as a life-buoy for her or his political career is not simply a matter of vexing the party to which the politician formerly belonged, but of potentially altering the political horizon.

This paper is not concerned with those politicians who move to another party and surrender their seats when doing so (for example, Cheryl Kernot). Nor is it concerned with those who are in parliament under one party's banner and are re-elected at some later date under another party's banner. That is, those who change affiliation outside parliament (for example, Peter Slipper, who was with the National Party for a term from 1984-87, but returned to parliament in 1993 under the Liberal Party's banner).

In addition, this paper is not concerned with 'crossing the floor'; a separate, forthcoming Department of the Parliamentary Library publication will deal with crossing the floor and its consequences for a politician. Rather, the focus is on the instances in which politicians quit their parties to sit as independents or join another party while retaining the seat to which they were elected. This is known in much of the literature and media commentary as 'jumping the party ship' or party jumping.

Section 1 examines defections in general, discussing the following questions:

• why do so few politicians defect?
• why do politicians choose to abandon their parties and how do they justify keeping their seats when doing so?
• who has the 'right' to the seat: the party or the politician?
• what are the consequences of jumping for politicians in terms of both their post-defection behaviour and their chances of re-election?

Section 2 discusses attempts to legislate against party switching in Papua New Guinea, India, South Africa and New Zealand. It argues that such legislation may be effective in resolving Papua New Guinea's debilitating problem with defections, but elsewhere such law has proved to be problematic at best and unworkable at worst.

The paper concludes by suggesting that, although politicians quitting the party while keeping their seats may raise the ire of both the party and voters, attempting to control such defections through legislation is not a sensible option.

Section 1: Defections

An important initial issue to settle before proceeding is what should be counted as a defection or a party switch. Jumping to another party or to sit alone is obviously a defection, but is the politician overboard when s/he votes against the party in the legislature? Counting this as a defection would prevent a politician from ever crossing the floor to vote against the party, unless the party had given permission for the dissenting vote or there was a free (or conscience) vote.⁽¹⁴⁾ Yet, as will be discussed below, this is the case in some countries that have legislation prohibiting party jumping (see Section 2, pp. 23-33). The aim of such a strict interpretation of what constitutes a defection may be to prevent a politician from staying with a party while continually opposing its decisions (a 'rat in the ranks' or a 'politician-in-exile' within the party). However, its effect may be to curtail a politician's freedom to exercise her or his judgment and act against party policies and procedures. As noted above, this paper is not concerned with this interpretation of 'defection'.

Disgruntled party members, furious with the decisions and actions of their party, have long had the option of cutting up their party membership cards and either joining the other side or staying out of active political involvement.⁽¹⁵⁾ However, the focus here is not on party members but on sitting politicians. What happens when politicians decide to quit the parties with which they were associated when running for office but retain their seats in the legislature? Are they under any obligation-legal or moral-to forfeit the seat when they jump off the party ship?

Given that political defections are common in some countries, it could be assumed that such questions had been explored in depth over time and could easily be answered. For example, in several countries party defections are part of the political culture and party jumping is an expected-and, to an extent, accepted-behaviour. One author notes that 'Japan, Bolivia, Ecuador, Nepal, Russia, the Philippines, France, Italy, Brazil ... suffer, or have suffered bouts of frequent [party] switching'.⁽¹⁶⁾

In Brazil, for example, party switching in the Congress of Deputies is routine and growing. Since Brazil returned to democracy in 1985, party defections have increased from 25 per cent in the period 1987-90 to about 40 per cent in the period 1999-2002.⁽¹⁷⁾ All of the switchers have kept their seats when jumping and the 'lack of party loyalty is accepted by most Brazilians who expect nothing better from politicians'.⁽¹⁸⁾ Ecuador, too, has experienced high levels of party switching, with about 12 per cent of congressmen joining another party or becoming independent in the period 1979-96.⁽¹⁹⁾ Writing on this 'common phenomenon in Ecuadorian politics', political scientist Andres Mejia-Acosta observes that politicians and voters alike perceive party switching, also known as 'camisetazo' or 'change of shirt', as 'an expected feature of legislative behavior that does not undermine the individual reputation of the switcher'.⁽²⁰⁾

In other countries, party switching may not be common, but nevertheless it does occur. For example, in the United States, since 1990 12 members of the House of Representatives and the Senate have either switched to another party or become independents. In the United Kingdom, there have been eight such jumps since 1990.⁽²¹⁾ In Australia, the number for the same period is 17: eight Senators and nine members of the House of Representatives have defected.

The differences in the level of party defections across countries may be attributable to a number of factors, including a country's political culture or electoral and party systems. However, the concern of this paper is not with the level of jumping or why there are more defections in some countries than in others. While these are interesting questions, worthy of further investigation, the focus of this paper is more general, looking at issues to do with party defections that are common to all countries, regardless of culture or system.

Despite the (overall) considerable number of occurrences of party jumping around the world, and the media coverage that such defections often attract, there is little in-depth analysis of the phenomenon. The academic literature contains only a few country-specific discussions-mostly of party switching in the United States, India and Latin America-and most of the focus is on party cohesion and party discipline rather than the rules-or lack of them-governing defections. There appears to be no cross-country comparison of party jumping, and no consolidated discussion of the issues surrounding those sitting members of legislatures who decide to quit their parties but keep their seats. Thus, before exploring the effects of political culture or electoral and party systems, it is worth probing some general issues surrounding party jumping that are applicable across countries and that have, thus far, been overlooked, namely:

• to whom is the politician chiefly responsible: herself (or himself), the voters or the party? (That is, is a politician obliged to exercise judgment on behalf of voters, obey the voters' will or follow the party line?)
• to whom does the seat 'belong': the politician or the party?

Comments on party jumping both in Australia and overseas show that, in the main, defection is regarded negatively; quitting the party while keeping the seat is regarded as breaking faith with voters, who elected the politician on a particular platform that is associated with a particular party. While, as noted, voters in some countries have come to accept defections, generally those who switch allegiance are condemned as 'unscrupulous' and 'opportunistic', 'unethical', and 'self-serving'.⁽²²⁾ Their behaviour is 'unconscionable', a 'breach of contract', and an act of 'treachery to the party'.⁽²³⁾ They are considered to have 'defrauded constituents', pandered to their own 'large egos', engaged in 'corporate theft', and to be in possession of 'stolen property'.⁽²⁴⁾

There is, however, an opposing view (to be explored in detail later in the paper) that appeals to the theory of political philosopher Edmund Burke (1729-97), who argued that elected representatives do not owe voters their blind obedience, but rather their judgment.⁽²⁵⁾ That is, according to Burke, politicians best serve their constituents when they exercise their judgment on voters' behalf and do not simply follow the demands of voters' or their parties. Today, this can be interpreted to mean that it is acceptable for a politician to, at times, go against the specific wishes of voters and leave the party if s/he thinks that such an action is in the voters' best interests.⁽²⁶⁾

A party's attitude to party jumpers appears to depend on where it stands in relation to the defector. In Australia, for example, the Australian Labor Party traditionally has condemned in scathing terms any 'rats' who desert the party-recall Senator Robert Ray's caustic description of Senator Mal Colston as the 'quisling Quasimodo from Queensland' who was treated as 'anyone who rats on the Labor Party' deserved.⁽²⁷⁾ Yet, just over a year later, this same party welcomed Senator Cheryl Kernot-until then the leader of the Australian Democrats-into its ranks, albeit under different circumstances in that Kernot surrendered her Senate seat.⁽²⁸⁾

Generally, Australian parties have an overwhelmingly negative view of party defectors who retain their seats, with one journalist observing after Bob Katter's defection from the National Party in July 2001 that Katter was now 'as popular as a brown snake in a sleeping bag with some of his Canberra colleagues'.⁽²⁹⁾ That said, it has been the case that the Liberal Party has not been as harsh over the years as the Australian Labor Party. The same cannot be said of the British Conservative Party, which launched a bitter smear campaign against defector Shaun Woodward in 1999.⁽³⁰⁾

In the United States, both major parties encourage their opponents to jump to their side, with one journalist describing the observations of these efforts as a 'biennial Washington parlor game'.⁽³¹⁾ However, when a defection occurs, the losing party generally is furious.⁽³²⁾

How voters feel about defecting politicians is less clear. In the long term, the most obvious test is whether or not a party-jumping politician is re-elected at the next election and the evidence here is not good for defectors, with many of them failing to retain their seats in open contest. (Election results for defecting politicians will be discussed further below.) In the short term, however, a defecting politician may attract some support for demonstrating independent judgment, 'rather than slavishly implementing party directives'.⁽³³⁾ Such support may be considerable at a time when increasing numbers of voters are not aligned with either of the major parties, respecting instead those who stand apart from parties and the party discipline that may cause politicians to vote against the interests of their local constituents.⁽³⁴⁾

The Party Line

Before looking at why politicians quit their parties, it is worth canvassing why politicians belong to parties in the first place. That is, a different approach to the central concern of this paper may be the following question: why don't more MPs defect? The answer can be divided into two main sections:

1. the role and strength of parties in a democracy, and
2. the benefits that a politician obtains from party membership

The Role and Strength of Parties

Parties have become a central organisational pillar of today's democracies. It is no longer possible to have the direct democracy of 5^th century Athens, in which all concerned citizens were required to participate directly in legislating because such tasks could not be delegated to others. Instead, modern societies have representative democracy in which voters elect representatives to make political decisions on their behalf. Parties arise out of this system; as one author notes, 'parties are inevitable. No one has shown how representative government could be worked without them'.⁽³⁵⁾

Parties are devices that organise collectively to place their representatives in the legislature and thus achieve their particular policy goals. This is not to say that parties are always formally recognised as part of a country's political system. While some countries have parties institutionalised in their constitutions (for example, France), others do not (for example, until relatively recently, Australia).⁽³⁶⁾

However, whether formally recognised or not, parties have become entwined with contemporary democracies through the tasks that they perform. Generally, parties:

• recruit and nominate candidates for elective office
• mobilise electoral support for their candidates and stimulate electoral participation
• generate symbols of identification and loyalty and simplify choices for voters
• aggregate specific interests into broader electoral and governing coalitions and articulate those interests
• contribute to the formulation of policy alternatives
• form and sustain governments, and
• integrate citizens more broadly into the nation-state and its political process.⁽³⁷⁾

Thus, one answer to the question-why don't more MPs defect?-is that parties, not individuals, are the organising focus of a democratic electoral system. (Whether or not voters back the party or candidates personally will be canvassed below.)

Benefits of Party Membership

It is not simply the role of parties that keeps elected politicians within their party's ranks. Undoubtedly, politicians also derive significant benefits from belonging to a political party that they would not receive were they standing alone. Some of these are intangible, such as the degree of personal satisfaction a politician may derive from membership in a party whose ideological viewpoint is the same as the politician's own.⁽³⁸⁾ However, other benefits of party membership for politicians are distinctly tangible and include electoral, collective, institutional and distributive advantages.

Electoral Advantages

Parties can enhance a politician's chances of being elected and re-elected.⁽³⁹⁾ This is not to say that independents cannot be elected; rather, that it is more likely for a candidate to be elected if she or he is associated with a political party.⁽⁴⁰⁾ Parties supply a 'brand name' that allows voters to recognise easily a politician's basic policy positions and match them with their own preferences. Political scientist Scott Desposato notes that parties can also facilitate voters' assignment of blame or reward for a failed or successful administration: 'Voters do not have to research all candidates' platforms or ties to the current administration-they just use a party label to summarize that information'.⁽⁴¹⁾

Politicians themselves use party labels to provide credibility to their claims to represent various segments of society, and to support their claims to be part of, or oppose, the current administration.⁽⁴²⁾

In terms of material electoral advantages, parties provide considerable practical assistance to their politicians, supplying electorate volunteers, finance for campaigns, resources to mobilise and organise voters, and general advice. As political researcher Michel Rossignol observes, 'only the party has the resources to mount the costly campaigns in the mass media which are an essential ingredient of modern day elections'.⁽⁴³⁾

Collective Advantages

Politicians benefit from belonging to a group that collectively has power to implement its objectives. Parties act as like-minded voting coalitions, working to advance common policy agendas and to defeat their opponents' agendas.⁽⁴⁴⁾ Thus, politicians benefit from being part of a larger collective when trying to implement policies that voters demand, both those that apply across all constituencies and those in a particular constituency. That is, politicians are able to seek-and, to an extent, expect-support from their colleagues for policies that may relate only to their own constituents in return for reciprocal support for fellow party members in the future.

Institutional Advantages

Parties may hold the keys to a politician's career in that they have a considerable amount of control over institutional factors that govern both the chances of a politician being elected and the career opportunities available once a politician is in the legislature.

In an election, parties may control such factors as the order in which candidates appear on the ballot paper and on a party's list. In the legislature, parties may control access to institutional resources, including the nominations for desirable political offices, committee assignments and leadership posts. These positions may enhance a politician's career by raising the legislator's profile within both the party and the electorate. While parties in opposition may be confined to internal promotions rather than external posts, they can offer longer-term advantages, rewarding those who are loyal while in opposition once they are in power.⁽⁴⁵⁾

Such is the power of parties to control these institutional resources that, even if a non-party-aligned politician is elected to a legislature, s/he may find herself or himself without the institutional capacity to achieve policy and career goals.

Distributive Advantages

Parties may also be able to offer politicians the opportunity to distribute resources to political advantage. As Desposato observes, parties aligned with, or controlling, government are privileged in access to government resources, including such things as 'road construction, military procurement contracts, social security pensions, and even cash that can be distributed directly to constituents'.⁽⁴⁶⁾ These resources allow politicians to deliver promised policy outcomes. Politicians who are members of ruling parties can garner significant credit from voters, and thus enhance their political careers, if they can fulfil their election promises as well as distribute extra goods to their constituents.

The opportunity for access to such resources is a powerful incentive for politicians to stay attached to their parties. Again, independent politicians are unlikely to have the same access to these distributional advantages.

Paying for the Party Favours

In short, then, the role of parties within a democratic political system and the benefits that politicians obtain from party membership are strong inducements for politicians interested in a long-term political career to both join parties and stay within their ranks.

However, there are costs to party membership, with one of the most important being that a politician is not free to make an independent choice on most policy issues. Joining a political party implies a voluntary contract to support that party and its views. In some parties, the contract is formalised through a signed pledge in which politicians promise to uphold the party's aims and constitution.⁽⁴⁷⁾ In Australia, both the ALP and the Democrats have such pledges. (The effectiveness of such pledges will be discussed below.)

To an extent, the contract with the party is a two-way obligation: as the politician agrees to support the party, so the party agrees to remain true to its manifesto-unless it can justify its deviations. It is the alleged breach of the party's side of this contract that many party jumpers use to vindicate their decisions to defect. That is, a common explanation for leaving the party is that the politician's principles remain unchanged (and aligned with the voters' will) whereas the party has abandoned core principles and is heading in a different direction. This will be discussed further below.

Politicians do have the option of going against the party while remaining within its ranks. On some policy issues, parties allow a free (conscience) vote in which their members do not have to adhere to the party's preference-or the party may have no preference-but may vote according to their own beliefs.⁽⁴⁸⁾ At other times, politicians may choose to defy party instructions and 'cross the floor' to vote with the opposition. However, consistently voting against one's party carries the risk of punishment, which may include reduced access to party resources, less influence on the party's legislative agenda or-in the worst case-expulsion from the party.⁽⁴⁹⁾ In some systems, of course-including Australia's-a single act of defiance may be enough to bring such punishment, as when Labor Postmaster-General Joe Lyons and his supporters voted in favour of a motion of no-confidence in the Labor government in 1931.

Party Discipline and Defections

Parties cannot achieve their goals if politicians continually choose to pursue their own individual agendas. A constant stream of floor-crossings undermines the ability of party leaders to maintain coherent policy platforms and may damage the party's reputation with voters, who may withdraw their support from a fractious and divided party. As political scientist Timothy Nokken observes, 'collective benefits are only enjoyed if members do not defect, if the party maintains cohesion'.⁽⁵⁰⁾

Academics Shaun Bowler, David Farrell and Richard Katz stress the importance of party cohesion to parliamentary systems, arguing that:

A difficulty for parties in maintaining cohesion is that parties are comprised of individuals, each of whom is capable of independent action and each of whom has her own set of priorities that may differ from that of the party.⁽⁵²⁾

One way that parties maintain cohesion is through party discipline or, in colloquial parlance, ensuring that its members 'toe the party line'. Some methods of discipline that parties have at their disposal are obvious. Parties in parliamentary systems, for example, have 'whips' or officers responsible for both keeping members informed of the party line on various issues and issuing any necessary directives (or whips) for how the party wants members to vote on specific issues on the floor of the House.⁽⁵³⁾ Sanctions that may be imposed on disobedient MPs include expulsion from the caucus or the denial of the opportunity to ask questions at Question Time.⁽⁵⁴⁾

Less obvious means of discipline at a party's disposal include the ability to offer or withhold desirable promotions and positions from recalcitrant members. Thus, 'loyal MPs may be promoted and disloyal MPs demoted'.⁽⁵⁵⁾ Dissenters may also lose office accommodation and staff or assistance with constituency work. Some of the ultimate sanctions at a party's disposal include withdrawing its support for the politician in the pre-selection for the seat at the next election or expelling the member from the party, both of which may potentially end a politician's political career.⁽⁵⁶⁾

These forms of discipline mean that, even in systems where discipline is characterised as 'weak', such as that of the United States, because there are few overt sanctions for those going against the party, politicians cannot afford to defy the party on every vote. This casts doubt on the claim that defections are less likely in countries with weak party discipline because such weakness allows politicians to vote however they please, meaning that a politician has no reason to defect.

Even where party discipline is 'weak' and politicians have 'relatively free rein to cast ... votes in nearly any manner they see fit', the underlying contract with the party governs-to a degree-a politician's behaviour.⁽⁵⁷⁾ Thus, even in the United States, politicians cannot always vote against the party line because doing so 'may alienate their core electoral constituencies and thus increase the likelihood of a strong challenge in primary elections'.⁽⁵⁸⁾

In addition to these 'negative' reasons for conforming to party discipline, there are several 'positive' ones, most notably those discussed above in the section on benefits of party membership. Political scientists Michael Laver and Kenneth Shepsle also argue that politicians submit to party discipline because it creates 'more powerful bargaining units' that are able to achieve policy goals that are close to the 'ideal points' of those submitting to the discipline.⁽⁵⁹⁾ That is, politicians recognise that they are more likely to achieve their aims when they work with others.

Last, but not least, politicians who owe their seats in parliament to their position on a party's list may feel a particular obligation to adhere to the line of the party that put them in power.⁽⁶⁰⁾ In summary, then, politicians are tied to their parties and the policies of those parties in a number of often overlapping ways. Generally, politicians gain significant benefits from party membership in terms of brand recognition, administrative and financial support and so on. However, on the cost side, the voluntary contract that binds politicians to a party also obliges them to support that party. They may also be bound by a pledge, owe their seat to their position on a party's list, or face disciplinary action up to and including loss of endorsement for their electorate if they fail to obey the party's rules and decisions.

Thus, the role of the party in democratic systems, the benefits of party membership and the costs of disobedience are among the reasons why politicians choose to stay aligned to their parties rather than jump overboard from the party ship.

After the Party's Over: I've Had Enough

Given the reasons discussed in the previous section for why politicians choose to remain with their parties, how can we account for those who choose to defect?

Justifying the Jump: Reasons for Defecting

According to a common understanding of representative democracy, politicians are responsible to those who elect them, either the people of a particular constituency or the people in general if the politician's seat has been allocated from a party list. However, as flagged earlier, another notion comes from Edmund Burke, who claimed that MPs were responsible to their consciences rather than voters. In an oft-cited speech to the electors of Bristol in 1774, he said that while voters' wishes and opinions should have 'great weight' with an MP:

In this view, politicians do not always have to vote as the electorate desires. Yet both of these notions-responsibility to voters and responsibility to conscience-may conflict with that of politicians owing their loyalty to their parties.

Generally, politicians are able to keep the demands of their voters, their party and their consciences in balance. However, they may feel they have no option but to defect if one of these competing demands comes to outweigh the others. It may also be that simply dissenting or abstaining on a vote 'would not indicate the full measure of their disenchantment'.⁽⁶²⁾

Academics Timothy Nokken and Keith Poole note that studies of party defections over a 'wide swath' of American history (1789-1984) have found that party switching usually coincides with important political events, such as military conflict or changes in partisan control of legislatures or key economic indicators.⁽⁶³⁾ This suggests that a politician may jump when her or his view on a major policy position differs from that of the party.

In Australia, the periods with the largest number of defections support the contention that most party jumping coincides with important political events. For example, large numbers defected in 1931 and 1936 over policies to deal with the Depression.⁽⁶⁴⁾ (Further details of party defections since federation will be supplied in a separate, forthcoming Parliamentary Library publication.)

Looking at more recent party switching in the United States, Nokken and Poole observe that defections over the past two decades are associated more with ideologically cross-pressured members or those members whose views do not sit obviously on one side of the left-right political spectrum and who, therefore, may feel at home in more than one party. The increase in defections for this reason fits with the idea that defections are less likely in highly polarised systems; that is, those systems in which parties have distinctly different ideological views that manifest in opposing views on policy issues.⁽⁶⁵⁾ In a highly polarised system, it is unlikely that a left-wing politician would jump to a right-wing party or vice versa.

However, today's major parties are more likely to appear towards the centre of the left-right ideological spectrum, which itself has become a less useful descriptive tool in a complicated world in which party views no longer fit so neatly into a single position on the left-right scale. As parties move to 'centre' positions, and even the same position on some issues, then there is an overlap that may lead to those in a 'right' faction of a 'centre-left' party jumping to the 'left' faction of a 'centre-right' party and vice versa. The jumps are justified with the claim that the new party offers a better match with the views of the politicians and their constituents. This helps to explain Nokken and Poole's observation on recent defections in the United States, where politicians are more likely to switch to another party than to sit as independents.

The following sub-sections discuss some of the most common reasons that politicians advance to justify their defections. Two points must be made with regard to these reasons:

1. they are not mutually exclusive nor exhaustive; party jumpers may use more than one justification when explaining their decisions and may offer other reasons not explored here
2. they are the publicly given reasons for the defections; there may, of course, be other underlying factors and circumstances that are not made public.

Deserting the Deviating Ship

A common explanation for leaving a party is that the party has changed while the defecting politician has not. The claim here is usually that the defector is still 'true' to the party's principles whereas the party has either abandoned those principles or twisted them beyond recognition.

Much of the justification that Meg Lees offered for her defection from the Australian Democrats in July 2002 was based on the claim that the party had shifted and her principles had not changed:

Senator Shayne Murphy used similar arguments when he resigned from the Australian Labor Party in October 2001 to sit as an independent. He responded to demands that he resign his seat with the following statement:

Further afield, in the United States, when Vermont Senator Jim Jeffords resigned from the Republican Party in May 2001 over differences on education spending, taxes and environment issues, he noted that he was increasingly finding himself in disagreement with the party on whose ticket he was first elected to Congress in 1974. Justifying his decision to jump from the Republicans to sit as an independent aligned with the Democrats, he said:

In a similar vein, in the United Kingdom, MP Shaun Woodward, who quit the Conservative Party in December 1999 to join Labour, argued that he had not 'ratted' on his party; rather, the party had 'ratted' on him:

When politicians believe that their parties are deviating from agreed core principles such that they find they can no longer vote along party lines, then they may feel they have no option other than to quit their parties. However, they retain their seats because they believe that they still represent the principles that attracted the support of voters.

Obeying the Voters' Will

Another explanation for party jumping is that the politician is obeying the will of the voters, who either no longer have faith in the party or have changed their political viewpoint. For example, in the United States, Jeffords's defection from the Republicans was seen to reflect that Vermont had changed in the past forty years 'from one of the nation's most reliably Republican states into one of the most strongly Democratic'.⁽⁷⁰⁾

A similar argument can be made in Australia to account for Bob Katter's decision to quit the National Party in July 2001 to sit as an independent. Katter claimed that the party's policies were crippling people in his constituency and:

That is, Katter believed the party's policies were no longer of benefit to Katter's rural Queensland constituents. Country residents themselves argued that the party had 'become city orientated and forgotten the country'.⁽⁷²⁾

When politicians consider that their parties no longer reflect the interests of their local constituents, then they may feel that they have no option but to quit their parties. Again, they retain the seats because they believe that this is what the voters want.

Exercising Judgment

Falling somewhat between the two previous justifications for why politicians choose to jump from the party ship is another explanation: that a politician has exercised her or his judgment and decided that it is in the voters' best interests for the politician to defect. This justification clearly reflects Burke's position, as outlined above, that representatives must use their own critical faculties when assessing policy options, taking into account party and electorate views but ultimately making what they believe is the best decision for voters and 'the country'.

There are many examples of politicians using this reasoning to justify quitting their parties. In the United States, Democrat Matthew Martinez (a member of the House of Representatives) jumped to the Republican Party in July 2000 on the grounds that his Democratic colleagues had:

Jeffords claimed in May 2001 that he had no other option than to quit the Republicans, even though it would give control of the Senate to the Democrats: 'I knew it was my responsibility to do what I truly believed was in the best interest of our country'.⁽⁷⁴⁾

In the United Kingdom, Paul Marsden, who quit the Labour Party to join the Liberal Democrats in December 2001, argued that the electorate was suffering as a result of the Government failing to deliver on its promises. In his view, the Liberal Democrats were the better alternative: 'Having thought about this over many months, I am convinced that the Liberal Democrats stand for honest and credible policies which can change this country for the better'. ⁽⁷⁵⁾Two years earlier, Woodward had justified his defection from the Tories on the grounds that the party's policies towards Europe were 'foolish and irresponsible' and had 'consequences that are anything but in Britain's national interest'.⁽⁷⁶⁾

In Australia, when MP Andrew Theophanous quit the Labor Party to sit as an independent in April 2000, he blamed-among other things-the party's 'unacceptable policies' on immigration, stating: 'My conscience can no longer allow me to pursue the party line in relation to these policy matters'.⁽⁷⁷⁾

Of course, the stories behind the defections are usually much more complex than this justification suggests, but the appeal to acting for the greater good is common. When politicians believe that, in their view, the voters are best served by moving away from the party, then this may justify the decision to move to another party or to become an independent. As before, the justification also supports the retention of the seat.

Self-interest

A final explanation for defection to be explored in this paper is one to which politicians would seldom-if ever-admit, yet is one of which they are often accused: acting in their own self-interest. No matter what justification politicians offer for their actions, it often appears to outside observers that a substantial amount of self-interest is involved.

If we assume that politicians are interested in being re-elected, then self-interest in achieving such a goal can be applied to explain everything from the varied timing of defections to the appointment of politicians to desirable posts in the aftermath of defections.

Timing: Politicians who defect early in their elected term often have a considerable amount of time before they have to contest an election and face voters' judgment of their actions. Jumping soon after their election (or mid-way through their term) may give politicians time to explain to voters the reasons for the defection, to win over voters and to prove their ability to independently achieve positive outcomes for constituents. Examples of politicians who quit early in their terms include Lees in Australia, who quit in July 2002 and whose term expires in June 2005, and Jeffords in the United States, who quit in May 2001 and whose term expires in 2006.

At the other end of the time scale are those who quit their parties shortly before an election, perhaps in order to distance themselves from unpopular policies that they fear may cost their parties-and therefore themselves-the chance of success at the election. Desposato notes that, in such instances, the benefit of a 'brand name' that party membership supplies may prove to be a disadvantage (or an advantage if the opposing party is in power). That is, he writes:

This reasoning may partially explain the rash of defections from the two major parties in New Zealand before the 1996 election. Both the Labour and the National parties had lost support as a result of introducing a range of unpopular policies when in government in the 1980s (Labour) and the early 1990s (National) such that minor parties attracted 30 per cent of the vote in the 1993 election. Before the 1996 election, thirteen Labour and National MPs quit to form their own parties or join existing minor-party alternatives.⁽⁷⁹⁾

American commentators anticipate this type of party-switching occurring in the United States before the next congressional elections in 2004, given the Republican Party's success at the November 2002 mid-term elections and the belief that Republican control over both chambers of Congress will continue for several election cycles.⁽⁸⁰⁾

Losing pre-selection: Politicians who lose pre-selection for their seats may feel betrayed by their parties such that they decide to quit and stand as independents. This occurred in Australia in 1996 when two sitting Liberal Party MPs, Alan Rocher and Paul Filing, 'fell foul of the swirling factional politics within the West Australian Liberal Party' and lost party pre-selection.⁽⁸¹⁾ Both stood successfully as independents at the 1996 election (but were defeated at the following election).

Desirable posts: Politicians may change their party affiliation in order to obtain access to desirable political posts, such as committee or leadership positions, or to retain their current assignments. Desposato argues that preserving committee assignments was the motivation behind several American legislators switching from the Democrat Party to the Republican Party in early 1995.⁽⁸²⁾ The formal 'courting' of politicians, where one party tries to coax members of another party to swap sides, appears to be an accepted practice in the United States.⁽⁸³⁾ In the United Kingdom, Conservative Party leaders responded to Woodward's 1999 defection with allegations that he had made a deal with Labour to get a ministerial job, a claim Woodward vehemently denied.⁽⁸⁴⁾ However, in what could be interpreted as a 'reward' for jumping, Woodward was later selected for the safe Labour seat of St Helens at the 2001 election.⁽⁸⁵⁾

Where the offer of desirable position induces politicians to jump to another party, the system may be open to allegations of corruption. In India, for example, which-as will be discussed below-has a particular problem with defections, a review of electoral processes argued that rewarding defectors with political positions 'and other such perquisites' encouraged corruption at the highest level.⁽⁸⁶⁾

While it may be tempting to assume that such 'rewards' are not offered in Australia, this is not necessarily the case. The Coalition Government elected Mal Colston as Deputy President of the Senate after his defection from the Labor Party in August 1996 in what many commentators saw as a reward for quitting Labor and supporting the Government on the sale of Telstra. Political commentator Alan Ramsey subsequently claimed that Colston had 'ratted on his party in 1996 for John Howard's irresistible 30 pieces of silver (the Senate deputy presidency)'.⁽⁸⁷⁾

Desirable resources: If parties in power have access to desirable resources, then politicians in opposition parties may decide to jump to (or support as an independent) the resource-rich parties in order to obtain these goods for their voters. In New Zealand, for example, defecting politicians who supported the minority National government in the 1996-99 parliamentary term were allegedly rewarded with leadership budgets and special treatment as well as ministerial portfolios.⁽⁸⁸⁾

'Give Back Our Seat': Rights and Ownership

Politicians who decide to quit the party on whose platform they were elected while retaining their seats in the legislature usually face immediate demands from the original party to 'give back' the party's seat. The claim is that the seat belongs to the party and not the individual politician because people voted for the former and not the latter.⁽⁸⁹⁾ In response, defectors usually present one of the justifications outlined above and can also argue that the seat is 'theirs' because the electorate voted for them personally as much as they voted for the party.⁽⁹⁰⁾

Most countries do not have legislation to dictate what should happen in the case of defections; those that do can find such law troublesome. This will be discussed further in Section 2, pp. 23-33.

In the absence of legislation forcing a politician to relinquish the seat, some parties ask their candidates to sign a party pledge that they will give up the seat if they quit the party. However, such pledges are not a watertight guarantee, as the Australian Democrats found with Meg Lees in July 2002.

The Democrats alleged that Lees had signed a party pledge according to which Democrat candidates agree to resign from parliament if they quit the party. Initially, Lees was reported as saying that she had not signed any document.⁽⁹¹⁾ Two days later, however, she was reported as saying that she might have signed something agreeing to abide by the party's constitution, which contains the following clause regarding the obligations of candidates:

The problems the Democrats encountered in seeking to enforce such a pledge are universal to all such party allegiance pledges:

• once the politician quits the party, s/he is outside its rules and cannot therefore be disciplined for any breaches of contract
• the pledge is not legally supported and it is unlikely that a court would 'countenance compacts of a political kind that would impinge on a member's freedom to act in good conscience, or on a member's right to hold and perform the functions of office'⁽⁹³⁾
• the pledge is not legally binding; rather, as Democrat leaders noted at the time, it is only 'morally' enforceable.⁽⁹⁴⁾

In the absence of an anti-defection law, then, the question of the 'right' to the seat becomes a moral, rather than a legal, issue and the arguments can be divided into two major streams of thought:

1. that parties are the modern units of politics and that politicians elected on party platforms are bound to follow the dictates of the parties-and the voters-who put them in parliament. Therefore, politicians should surrender their seats if they quit their parties.
2. that politicians are obliged to exercise their judgment on behalf of their constituents rather than their parties (Burke's notion, outlined above). Therefore, they should have the right to quit their parties without sanction if, in their judgment, the voters are best served by this course of action.

As can be seen from these arguments, a key question is whether parties or individuals are considered to be the central elements of the political system.

Once again, the answer to this question is complex. Some claim that, even when they stand for election on a party platform, constituency politicians are primarily responsible to their electorates, rather than their parties, and this responsibility entitles them to retain their seats should they defect. In addition, it may be that parties are not as much a focus of the system as is suggested. In Australia, for example, parties were not listed against a candidate's name in Australia's official election statistics until 1975.⁽⁹⁵⁾

Political scientist Dean Jaensch counters the primacy of parties in Australia with the following arguments:

• political parties are not mentioned in the electoral parts of the 'sovereign law' of Australia, the Constitution
• members of the House of Representatives are referred to as the member for [the electorate], not the Labor, Liberal or National member for [the electorate]; senators are referred to as Senator [name] and not Senator [name] of the Democrats or the Greens
• political parties do not elect Members of Parliament; voters elect Members and Senators on the basis of personal choice
• more independents are being elected, which weakens the party argument, as independents 'depend for their election on a personal vote, personal appeal and personal support'.⁽⁹⁶⁾

Evidence from New Zealand-where voters have both an electorate and a party vote-supports the claim that party is not the only consideration for voters. Political scientists making a submission to a New Zealand parliamentary committee inquiry into anti-defection law observed that about 35 per cent of voters split their votes in the 1999 election; that is, they did not give their party vote to the same party that their preferred candidate represented.⁽⁹⁷⁾ Such 'split voting', which grew to 39 per cent in the 2002 election, could reflect strategic voting to prevent one party having too great a majority, but it could also reflect that people vote for the candidate they like, even if that candidate represents a party they dislike. That is, they vote for the person and not the party.

In addition, one of New Zealand's most notorious party jumpers-Jim Anderton, who has changed parties several times in the past 20 years-was re-elected in the July 2002 election, despite the controversy surrounding his latest jump earlier that year (see the discussion on New Zealand, pp. 30-33). The outcome indicates that votes for particular personalities may still be an important factor in the New Zealand electoral system, even with the alleged primacy of parties.

Countering this, of course, is the example of elections for the Australian Senate, where most of the votes are cast 'above-the-line' for the party and not against the individual candidates' names.

Jaensch notes that an election is a viable democratic process without parties, but that 'party domination is a "tier" which has grown on the top of the election process like a fungus'.⁽⁹⁸⁾ He argues that it would be insulting to suggest that voters do not care who the candidates are as long as the party label is there and concludes that it is not the politician or the party that owns the seat. Rather:

Those who want defectors to surrender their seats offer a straightforward counter-claim to these arguments. That is: if personally popular constituency MPs are responsible to their electorates before parties, then they should have no objection to putting their popularity-and approval of their decision to defect-to the test by agreeing to surrender their seats and submit to an immediate by-election.

But Jaensch offers a feasible counter-argument:

Which returns us to the argument about the need for politicians to be able to exercise judgment on behalf of constituents, even when that judgment results in a decision to defect.

In summary, in the absence of legislation, there is no easy answer to the question of who has the moral right to a politician's seat. It may be that, like the children's game of musical chairs, the seat 'belongs' to whoever is sitting in it at the time.

The Consequences of Jumping Overboard

Although many of the reasons for defecting outlined above appear to be perfectly acceptable justifications for the jump, politicians who quit the party while retaining the seat pay a hefty price for their actions. Among the most obvious costs to those who choose to sit as independents are the loss of the benefits of party membership, from material resources through to the deprivation of the company of one's former colleagues.⁽¹⁰¹⁾ Those joining another party may face hostility from their new colleagues, who may be reluctant to trust those who are considered to have 'betrayed' another party. As one journalist observes, defection 'sends a signal of unreliability when crunches come'.⁽¹⁰²⁾ Such hostility and lingering doubts over a defector's motivations may have a negative impact on the politician's career and the positions and promotions that s/he would ordinarily expect.⁽¹⁰³⁾

The ultimate price for both new independents and party switchers, of course, is that they may lose the support of voters and thus lose their seats the next time they stand for election. As journalist Steve Richards observes: '... the trauma of being abused by former colleagues and viewed with suspicion by new ones is exacerbated by the prospect of a political void after the election'.⁽¹⁰⁴⁾

So, if the costs of defecting include hostility and suspicion on the part of new colleagues and voters, how does this affect the behaviour of defectors and what are the outcomes for party jumpers at the next election?

Changes in behaviour

Research into the consequences of party switching in the United States suggests that, because defectors feel they have to prove themselves to the new party and to voters, they significantly modify the way they vote in order to fit the line of the new party. Thus, even in a system with weak party discipline, in which there are no overt sanctions to voting against the party, party-switchers are more likely to vote in accordance with their new party's wishes. As Nokken and Poole observe, 'Some legislators may "switch with a vengeance" in order to show their new colleagues in the new party that they are "real" Republicans or "true" Democrats'.⁽¹⁰⁵⁾

Nokken and Poole count those who quit their parties to sit as independents as party switchers, even when these politicians label themselves as 'independent' members of their original party. Thus, Nokken and Poole's conclusions on the significant shifts in voting patterns can be taken to apply to these defectors as well; that is, both those who join another party and those who become independent change their votes after their defection.

However, while this may be true for the United States, there appears to be little (or no) research outside the United States on the voting patterns of those who quit the party while claiming to remain true to the party's ideals. In Australia, for example, it is not known whether a disaffected Liberal, Labor or Democrat politician continues to vote as if s/he still belonged to the party.

Political journalist Michelle Grattan, writing about Meg Lees in the wake of her defection, noted that Lees still slipped into using 'we' when she spoke of the Democrats. To be fair to Lees, Grattan was writing only a few days after Lees quit. However, Grattan's subsequent point is valid:

Given the number of independents in the Australian Parliament (currently six-three in each chamber), it would be useful for future research to examine whether there is a difference in voting patterns between those 'bona fide' independents-those who have never been affiliated to a party-and those independents who were once faithful party members.

Another interesting question is whether or not defectors vote as though they retain a mandate from their electorate for their actions; that is, do party jumpers argue vehemently for particular policies or do they abstain more often than they did when they still belonged to a party?

Of course, the new 'loyalty' need not be confined to voting: reports in the wake of Woodward's defection from the Tories to Labour in 1999 noted that Woodward had become 'one of Labour's most active backbenchers', visiting far-flung constituencies to 'offer his inside knowledge to highlight the dangers posed by the Conservatives'.⁽¹⁰⁷⁾

'Short-term fame and long-term obscurity'

Evidence suggests that defections are not popular with voters in that few party jumpers retain their seats in the longer term. This is not to say that defectors are never re-elected; rather, the number who lose their seats is vastly greater than the number who retain them. In the United Kingdom, for example, defections are seen as 'rare and risky' because quitting the party 'tends to bring short-term fame and long-term obscurity'.⁽¹⁰⁸⁾

Whether or not a party-jumping politician is re-elected appears to depend on how believable voters find the justification for the defection. It appears that the chances of re-election decline if voters perceive the action to be based on self-interest. In New Zealand, for example, none of the controversial party jumpers in the 1996-99 parliamentary term were re-elected at the 1999 election, but those who defected at other times have been re-elected and some who defected years ago remain in parliament today.⁽¹⁰⁹⁾

For those politicians who become independents, their re-election chances may depend not only on voters' judgment of the justification for the jump, but also on voters' attitudes towards independents. In Australia, for example, Bob Katter as an independent attracted 47 per cent of the primary vote in the 2001 election, up from the 44 per cent he received as a National Party member in 1998. The result may partially reflect an increased willingness on the part of rural voters to support independent candidates over party candidates. Journalist Cameron Forbes reported on this phenomenon before the election, noting that 'Rural rage is translating into growing support for unaffiliated politicians who will stand up for the home side'.⁽¹¹⁰⁾

However, even when defecting politicians are re-elected, research in the United States suggests that their level of support may be lower than before they quit their parties. Political scientists Christian Grose and Antoine Yoshinaka find that, on average, incumbent politicians who change parties have poorer showings after their switch in both general and primary elections and that these electoral consequences occur in every election after their switch.⁽¹¹¹⁾

In summary, then, defection may have several significant costs and consequences for a politician, including:

• hostility from former colleagues and suspicion from new ones
• suspicion and anger from voters that may manifest in an electoral backlash against the politician at the next election
• a more limited ability to undertake independent action because of the burden of proving one's allegiance to the new party
• higher costs at the next election if a politician standing as an independent has to fund and direct a campaign without party support

Given these costs and consequences, a reasonable assumption would be that politicians decide to quit their parties only if-in each particular circumstance-the benefits of doing so outweigh the costs. Such an equation may be a matter for each individual to judge as it is unlikely that all would assign the same weight to the various psychological and material benefits. That is, some politicians may give greater weight to psychological benefits, such as the personal satisfaction of standing up for one's principles, while others give greater weight to practical considerations, such as the positions or resources they may obtain through the defection.

Section 2: Legislation

Although defections in Australia are usually accompanied by a hue and cry about the party jumper's right to retain the seat, there is nothing beyond this expression of outrage that can be done-Australia has no law to control what should happen when a defection occurs. As noted above, the losing party usually demands the return of the seat to its rightful owner (the party) while the defector claims the moral right to retain the seat for any one of the reasons outlined above. Ultimately, the jump itself generally sinks almost without trace-the only ripple on the political surface being the continued grumbling of the defector's original party.⁽¹¹²⁾

In fact, most countries do not have legislation controlling party jumping. Neither the United States nor the United Kingdom (from which Australia's political system is drawn) has anti-defection legislation. Even those countries in Western Europe that are (or have been) plagued by party jumping-France and Italy among them-have refrained from introducing laws to curb the practice.⁽¹¹³⁾ As noted above, often the only attempt at control is through conditions in a party's constitution.⁽¹¹⁴⁾ Those countries that do have legislation to control defections, including Bolivia, Fiji, India, Namibia, Papua New Guinea, South Africa, Trinidad and Tobago, Vanuatu, Zambia and Zimbabwe, have been dismissed by some as needing the legislation because they are 'dictatorships and fragile democracies'.⁽¹¹⁵⁾

Should there be legislation in place to force party jumpers out of parliament? The question of legal control is vexed because sound arguments can be advanced on both sides. In effect, there are two levels of argument: 1) whether or not the law has a role in controlling political defections and 2) the impact of such anti-defection law on democracy.

The Role of the Law

There are two views on whether or not party jumping is a matter for the law to control. On the one hand, 'the law has a role to play in establishing a moral position, in declaring a standard of behaviour which is acceptable'.⁽¹¹⁶⁾ On the other hand, the law can be a heavy-handed instrument and, because it applies to all cases, cannot distinguish between 'creditable' and 'discreditable' party hopping.⁽¹¹⁷⁾

Journalist Gerard Henderson demonstrates the problem with the latter in a different context when he queries the blanket application of the term 'rat' to all ALP defectors, arguing that some had good reason to break from the party.⁽¹¹⁸⁾ Then-Labor leader Kim Beazley had made the same point in 1998 when describing Mal Colston's defection from Labor:

Thus, even a party can identify times when party jumps occur for what are deemed to be acceptable reasons. The question is whether it is possible to design a law that could make the same distinction.

Democracy and Anti-Defection Law

Again, there are two views on the impact of the law on democracy. On the one hand, it is argued that anti-defection law ensures 'genuine democracy' is maintained throughout the legislature's term because it:

• ensures that the will of the people as expressed in a democratic election is upheld
• maintains (in proportional electoral systems) the proportionality of the elected parliament
• enforces stability in the legislature by preventing defections that may alter the balance of power between parties
• strengthens party leadership and organisation by preventing politicians from quitting their parties in order to reject party decisions, and
• allows parties to enforce disciplined voting and maintain coherent policy platforms in the legislature.⁽¹²⁰⁾

On the other hand, anti-defection law is said to undermine many of the principles of democracy because it:

• prevents politicians from following their consciences and binds them to a party view, whether or not they believe that view is right or wrong
• stifles free speech and freedom of association (and, by implication, dissociation)
• elevates parties to a position in the political system that may not be justified by a reading of a country's constitution, and
• tightens party control of its elected members to the point that it weakens a key principle of parliamentary democracy, 'namely that a government is kept under control by the possibility of its supporters crossing the floor'.⁽¹²¹⁾

Attempts to Control Defections

As noted above, some countries have implemented legislation demanding that politicians who quit their parties also quit their seats in parliament. In some, the law goes so far as to count as defectors those who vote contrary to party directions, who abstain from voting to avoid following a party directive, or who simply dissociate themselves from their parties.⁽¹²²⁾

How successful are these attempts to control party jumping? In order to examine this question, four case studies are offered:

• Papua New Guinea, which passed legislation in December 2000 in a bid to control the rampant party hopping that had undermined successive governments
• India, which is cited in many of the discussions so far of party jumping because of its staggering number of defections⁽¹²³⁾
• South Africa, which tried to amend its anti-defection law in June 2002, and
• New Zealand, which passed the Electoral (Integrity) Amendment Act in December 2001 in a direct attempt to stop the party hopping that had plagued the 1996-99 parliamentary term.

Papua New Guinea demonstrates that anti-defection legislation may have a valuable role in resolving destabilising problems within the political system. That said, it may be too early to judge whether such legislation has been-or will be-successful. The other three examples reveal many of the difficulties associated with anti-defection legislation and suggest that such laws are problematic at best and unworkable at worst. In India, for example, more defections occurred each year, on average, after an anti-defection law was introduced than occurred before.⁽¹²⁴⁾ In New Zealand, one of the main proponents of the anti-defection legislation found a way to obey it in practice but disobey it in spirit before the July 2002 election.

Papua New Guinea: an Essential Law

Background

Although Papua New Guinea 'boasts one of the developing world's most impressive records of democratic longevity', having held elections for almost 40 years, democracy in practice has been difficult-no government since independence in 1975 has survived a full five-year parliamentary term.⁽¹²⁵⁾ A major problem has been the fragmented and unstable party system, which suffers from rampant party jumping. MPs have regularly switched allegiance in return for 'ministerial posts, perks, and other financial inducements'.⁽¹²⁶⁾ As political scientist Benjamin Reilly observes:

Underlying the fractured party system is the ethnic diversity of the country, which Reilly describes as being 'both a blessing and a curse'. That is, it is a 'blessing' in that it makes it unlikely for one group to be able to control power alone, but it is a 'curse' in that such cleavages make cohesion difficult, with governments comprising 'impermanent coalitions of various parties, groups and individuals'.⁽¹²⁸⁾

Papua New Guinea's electoral system, first-past-the-post, exacerbates the difficulties arising from the clan and tribal divisions in that it allows candidates to be elected with a very low level of support.⁽¹²⁹⁾ Reilly notes, for example, that 15 Members of Parliament were elected in 1997 with less than 10 per cent of the vote each; three had attracted less than 7 per cent of the vote. Thus, 'for many MPs, their constituency is not the electorate which they supposedly represent, but the much smaller subgroup within their electorate to which they owe their allegiance, and their parliamentary positions'.⁽¹³⁰⁾ With such low levels of support, politicians are vulnerable to losing their seats at the next election if there is even a small shift in vote share. The result is extreme electoral volatility: 'over half the parliament regularly lose their seats at each election'.⁽¹³¹⁾

The Integrity of Political Parties and Candidates Law

Anti-defection legislation was introduced in December 2000 as part of a package of reforms aimed at improving the country's electoral system, parliament and party system. The law, an Organic Law called the Integrity of Political Parties and Candidates Law, came into force for the 2002 election.⁽¹³²⁾ Its aim is to ensure politicians cannot 'hop around for reasons of political opportunism'.⁽¹³³⁾

The Integrity Law restricts the freedom of Members of Parliament to change parties, and imposes penalties if they leave the party with which they were aligned when first elected and join another party or become independent. Those who choose to leave their party must face a 'leadership tribunal' (the Ombudsman Commission), which decides whether their grounds for resignation are valid. Under the legislation, valid resignations are possible only when the party has breached its own constitution or when the party has been declared insolvent.⁽¹³⁴⁾ If the tribunal rules against the politician, a by-election must be held.

In addition, MPs elected with party endorsement must vote in accordance with their party's position on key issues including the election of a prime minister, the Budget, votes of no-confidence, and constitutional amendments. Reilly notes that MPs can abstain, 'but if they vote against their party's position, they face a range of possible penalties, up to and including dismissal from the parliament'.⁽¹³⁵⁾ The law also imposes a range of restrictions on politicians who are elected as independents but who then join a party.

Engineering to achieve objectives

Reilly notes that the Integrity Law is an attempt at 'political engineering' or 'crafting the institutional "rules of the game" to achieve certain objectives'-in this case, party stability and therefore government stability:

Reilly acknowledges that the net effects of the laws are unlikely to be seen for some time as such institutions must be kept constant for several years and through several elections 'before their strategic impacts stabilize and become clear to political actors and voters alike'.⁽¹³⁷⁾ But others are sceptical about whether it is possible to change Papua New Guinea's entrenched political culture 'from above-in effect, by constitutional fiat'.⁽¹³⁸⁾

In addition, there are fears that the Integrity Law will impinge on the rights of individual Members of Parliament to 'represent their constituents as independents' and that it may result in entrenching 'a poor or even dangerous' prime minister in office for five years.⁽¹³⁹⁾ One commentator pointed to the increase in parties-from the usual 20 to 42 in the 2002 elections-to argue that candidates already were attempting to find a way around the laws.⁽¹⁴⁰⁾ That is, candidates preferred to keep their options open, either avoiding party memberships or leading their own minor parties, so that they were not locked in to supporting a particular party for five years.⁽¹⁴¹⁾

It is too early to judge how successful the law has been in achieving its aims. In the 2002 elections, independents won only 17 of parliament's 109 seats, but more parties (24) won seats. Whether or not the law will guarantee stability with such results is not yet clear. However, Reilly argues that, at the very least, the law is an 'innovative institutional response' to the problems that have plagued the country's politics.

India: a Problematic Law

In January 2001, a consultation paper on India's electoral law discussed the problem of party jumping, noting that defections had 'haunted the Indian polity' for more than 30 years.⁽¹⁴²⁾ The numbers given for defections were staggering: between 1967 and 1972, nearly 2000 of the roughly 4000 members of the Lok Sabha and Legislative Assemblies in the states and union territories defected and counter-defected.⁽¹⁴³⁾

In a bid to control party jumping, in 1985 India introduced anti-defection law-the Tenth Schedule-which disqualified from parliament any politician who voluntarily quit his or her party or who voted against it without first obtaining permission from the party or without such a vote being a party-sanctioned free (conscience) vote. However, the schedule exempted from disqualification those politicians who left their parties through a party split or joined another party through a party merger. The question of when a politician could be said to have defected was to be judged by the chairman or the Speaker of the House.

However, the anti-defection law has proved to be problematic. Critics argue that the clauses allowing party jumping through party splits and mergers have made the legislation ineffective. The 2001 consultation paper noted that the law had had little effect on the number of jumpers, serving to decrease individual defections but increase en masse defections. It argued that much of the problem lay with the implementation of the Schedule, with Speakers tending to act in a partisan manner and without due regard to the law's provisions.

India's Chief Election Commissioner, Dr Manohar Singh Gill, reinforced this interpretation of the root of the problem in June 2001 when he blamed the law's failure on the 'enforcing authority, the Speaker of the Legislature', and not on the law itself.⁽¹⁴⁴⁾ He argued that, because the Speaker was dependent on a political party for his position, the 'ineffective' law could work only if responsibility for enforcing it was taken from the Speaker and given to an independent body such as the Election Commission. In the absence of such a move, 'the malaise of defection' would continue to 'eat ... into the political system like a cancer'.

South Africa: an Out-Dated Law

South Africa's post-apartheid Constitution included an anti-defection clause modelled on that contained in India's Tenth Schedule. The clause provides that politicians lose membership of a legislature-national or provincial-if they cease to be a member of the party that nominated them. The aim was to ensure political stability in what was expected to be a volatile transition from apartheid to non-apartheid politics by 1) protecting the proportionality of parliament and 2) protecting multi-party democracy.

1. The anti-defection clause was seen as a means of protecting the proportionality of the legislature as decided at the election. That is, it would ensure that an elected member's decision to quit a party after the election would not take away one of the seats to which the party was entitled on the basis of the proportion of the vote it received at the election.

A key argument here is that it is 'the party, and not the members, which is entitled to the seats'.⁽¹⁴⁵⁾ The logic of the argument is based on South Africa's electoral system-list proportional representation-under which the names of parties, not individual candidates, appear on the ballot paper. Thus, voters decide how many seats a party should receive in parliament and politicians gain their seats in the legislature through their position on a party's list. Permitting a member to quit the party but retain the seat would distort the proportionality between parties decided by voters at the election. The anti-defection law ensures that if a politician chooses to leave his or her party, then the party has the right to take back its seat and give it to another person on its list.

2. The anti-defection clause was also seen as a means to protect multi-party democracy by reinforcing the position of smaller parties in the proportional system. Smaller parties were seen to be under threat because the African National Congress, which held two-thirds of the seats in the National Assembly, was in a position to offer inducements to members of smaller parties in order to get them to join its ranks. If this were allowed to occur, then the number of parties ultimately would decline and multi-party democracy could fail.⁽¹⁴⁶⁾

The anti-defection clause was tested in the country's constitutional court in 1996 and 1997 and was ruled to be 'consistent with the core policies and principles of democracy, including representative government, freedom of speech and freedom of association'.⁽¹⁴⁸⁾ Again, the court's ruling was based on the primacy of parties in the electoral system. It found that parties, not individuals, were accountable to the electorate. Under an anti-defection clause, individual members were still free to follow the dictates of personal conscience and could resign if they felt the party had abandoned core principles. However, voters held the right to judge whether or not it was true that the party had deviated from its policy platform, and could exercise this judgment at the next election. That is, the court noted: 'A party which abandons its manifesto in a way not accepted by the electorate would probably lose at the next election'.⁽¹⁴⁹⁾

The positive view of the effect of the anti-defection regulation is not universal. Political consultant David Welsh, writing for a South African research institute, notes that the ANC has found the combination of list-system proportional representation and anti-defection legislation highly effective in controlling its politicians. He argues that the law 'gives the leadership a tight grip on the compilation of lists and ensures that, once elected, MPs toe the party line or face expulsion from Parliament'.⁽¹⁵⁰⁾

Changing circumstances

With several years having passed since the transition to a non-apartheid system, the circumstances under which the anti-defection clause was introduced can be argued to have changed. In June 2002, the South African government enacted two constitutional amendments aimed at relaxing the anti-defection controls. The ANC fought for the legislation on the grounds that it strengthened democracy by giving 'elected public representatives the right to make choices within the confines of the Constitution'.⁽¹⁵¹⁾

An opposition party, the United Democratic Movement, immediately launched legal proceedings to have the amendments struck down as unconstitutional. However, in October 2002, the Constitutional Court upheld the legislation, ruling that the anti-defection provision was not 'so fundamental to our constitutional order as to preclude any amendment of their provisions'.⁽¹⁵²⁾ Moreover, defections in a proportional representation system were not inconsistent with democracy, nor was an anti-defection provision 'an essential adjunct to the proportional representation system'.⁽¹⁵³⁾

In an interesting difference to its earlier view, the court appeared to rule in favour of individuals over parties, arguing that:

Nonetheless, while the court found that legislation allowing party jumping (called 'floor crossing' in South Africa) was not inconsistent with the constitution, it did not approve of the way the government had introduced the amendments allowing floor-crossing at the national and provincial level. While it allowed the new legislation to apply at local government level, it insisted that the government introduce a further constitutional amendment for national and provincial changes.⁽¹⁵⁵⁾ As a result of this decision, it ruled in November 2002 that five KwaZulu-Natal legislators who defected to the ANC would not be protected and would lose their seats. However, in a further complexity, these legislators may be reinstated this year (2003) if parliament passes the floor-crossing legislation because the proposed law allows for retrospective protection.⁽¹⁵⁶⁾

New Zealand: an Unworkable Law?

New Zealand offers an example of anti-defection law worthy of examination in Australia because it so clearly demonstrates that a law implemented with the best of intentions-preventing self-interested party-hopping-may not work as well as its instigators desire. In fact, the New Zealand case proves that there can be many pitfalls between a law's aims and its outcomes.

Background

In 1996, New Zealand changed its electoral system from first-past-the-post (FPP) to mixed-member proportional (MMP), and divided the 120 seats in parliament into electorate seats and list seats. The latter are apportioned according to the share of the party vote that parties receive in the election.⁽¹⁵⁷⁾ The list system means that some politicians owe their seats in parliament to their position on a party's list, not to a personal vote.⁽¹⁵⁸⁾ Thus, list MPs are in parliament by virtue of their party affiliation and therefore 'have a particular responsibility to represent their party'.⁽¹⁵⁹⁾

Politician(s) Overboard

In July 1997, only a few months after the October 1996 election, in an act seemingly unforseen in the legislation outlining the intricacies of MMP, an Alliance list MP, Alamein Kopu, decided to quit the party and sit as an independent. (She later claimed to represent the parliamentary wing of a new Maori party, Mana Wahine.) Although she had signed a party pledge to resign from parliament if she left the Alliance, she refused to surrender her seat. The Alliance challenged her decision, but the Privileges Committee ruled against the party because there was nothing in the MMP electoral legislation that said an MP-whether an electorate MP or a list MP-must quit the seat when s/he left the party.⁽¹⁶⁰⁾

Kopu's successful leap prompted several other politicians to quit their parties, with another Alliance member and at least five members of New Zealand First defecting. The latter quit their party when its coalition agreement with the National government collapsed. The su