See Thompson 1980.

Parker argued that the successful transplanting of the Westminster system was one of “four overt aims” of Australian policy in PNG (1971, 348). Following the 1972 elections, the forms of the House of Assembly increasingly resembled a genuinely sovereign parliament, with the Westminster format of government and opposition being institutionalised by the appointment of Matthias ToLiman as leader of the opposition to Michael Somare’s National Coalition government (Waddell 1973, 28fn).

The ‘British model ’ o f constitutional inculcation

The way the institutions of elected parliamentary government were developed was also of great importance to the consolidation of democracy in PNG. In Weiner’s study of elections in developing countries cited earlier, the six successful developing-world democracies (PNG, India, Sri Lanka, Malaysia, Jamaica, and Trinidad and Tobago) were assessed as having little close relationship on almost any social or economic indicator, but “one common feature ... all are former British colonies” (1987, 19). While PNG’s major colonial influence came from Australia (itself an artefact of British colonialism), the significance of Weiner’s argument is based on the common democratic heritage afforded to many former members of the British Empire as part of the ‘British model’ of democratic tutelage and decolonisation, a model followed in form — and largely in function — by the Australian government in relation to PNG. Weiner’s conclusion about the importance of British colonialism has been reinforced by S.M. Lipset, who has found that past experience with British rule emerges as one of the most powerful correlates of democracy in the developing world (1996, 153).

The ‘British model’ (and the Australian model) of democratisation and decolonisation was typified by two basic components: the creation of bureaucratic structures which stressed the legitimate role of state authority (e.g. the police and the judiciary; functions exercised in colonial PNG largely the Australian patrol officers known as kiapsf and the establishment and institutionalisation of the principles of representative democracy and free and fair elections. By institutionalising rules and procedures as a starting point for democratic politics, and by emphasising the primacy of the rule of law and constitutionalism as guiding and binding forces upon political competition, the ‘British model’ created a well-defined arena for political competition before such competition was effectively unleashed in the form of mass suffrage elections and full indigenous participation in politics. Eckstein has argued that this process of building consent and

commitment to democratic principles is the crucial element in making democracies work. The British, he maintains, “invest with very high affect the procedural aspects of their government and with very low affect its substantive aspects ... Procedures, to them, are not merely procedures, but sacred rituals” (quoted in Weiner 1987, 14).

This commitment to procedures is more significant than it perhaps first appears. Weiner argues that the tradition of imposing limits on government, establishing norms for the conduct of those in power and creating procedures for the management of conflict has been hugely influential in the establishment and consolidation of democracy in developing countries (1987, 20). He concludes that the ‘British model’ of tutelary democracy is the single most effective colonial model of democratisation, due to its creation of centralising institutions that can gradually be made indigenous, and the establishment of democratic institutions and free and fair elections. This institutionalisation of political conflict into an adversarial system of political competition bounded by established rules leads to an acceptance of these rules and procedures amongst members of the political elite. In this way, the institutional framework for democracy “helps create the conditions of its own persistence” by enabling the emergence of classes which have an interest in the maintenance of these institutions, and by nurturing popular attitudes supportive of democratic procedures (Weiner 1987, 32).

Regan has described constitutionalism in PNG as being characterised by an acceptance amongst elites that the constitution is the appropriate instrument for managing political secession and mediating elite competition (Regan 1995, 10). A number of incidents have demonstrated the success of this acceptance of the legitimacy of the constitutional arena, and of what Saffu typifies as the “strong hold” of constitutionalism on PNG’s political actors (1985, 531). On 11 March 1980, Michael Somare’s government, which had been in power since the introduction of self-government in 1973, was defeated on the floor of parliament by a vote of no-confidence. This was the first time that PNG’s governing elite had faced the prospect of an alternation of power under the rules of a parliamentary system. To the surprise of some observers, Somare immediately stepped aside and transferred the prime ministership in a peaceful and orderly manner to the sponsor of the no-confidence motion (and Somare’s former deputy of two years earlier), Julius Chan. This first transfer of executive authority can be seen as a testament to the

success of the Australian government’s attempts to emulate the British model of decolonisation by instilling a commitment to constitutionalism and democratic process.

A series of other potential crises has also been handled within the boundaries of the constitutional ‘rules of the game’. In 1991 the governor-general, Sir Serei Eri, refused to follow constitutionally-binding advice to dismiss the deputy prime minister, Ted Diro, after the release of adverse findings against Diro by a leadership tribunal inquiry into corruption. Instead, Eri defused this potential crisis by resigning from office, to be replaced by a new governor-general who had no hesitation in sacking Diro. More recently, in September 1993, the then prime minister, Paias Wingti, organised a constitutionally questionable parliamentary manoeuvre: a surprise resignation and immediate parliamentary re-election after 14 months in office in order to circumvent a constitutional provision permitting votes of no-confidence after 18 months in office, thus gaining another 18 months of valuable governing time. In August 1994, Wingti’s own appointment as Chief Justice of the Supreme Court led a full bench of seven judges who all ruled that the resignation was valid but that the constitutional requirements of the re-election had not been met, thus forcing a new election by the parliament which enabled Wingti’s former deputy, Sir Julius Chan, to return to the prime ministership. Again, these transfers of power were conducted in an orderly fashion, and Wingti accepted the Court’s decision. The acceptance of these and other rulings by political actors are themselves a persuasive example of the resilience of PNG’s constitutional structures.

The most serious threat to the existence of constitutional government to date was the military insurrection in March 1997 against the Chan government’s decision to hire foreign mercenaries to assist the government in its ongoing secessionist war on Bougainville (Dinnen, May and Regan 1997). This revolt was led by the commander of the Defence Force, Brigadier-General Jerry Singirok, who stopped well short of a full- scale attempted coup but who nonetheless was able to mobilise sufficient numbers of the armed forces to make his call for the resignation of Prime Minister Chan a serious threat to parliamentary democracy. While the army’s push for Chan’s resignation and pressure on the government to drop their contract with the mercenaries was clearly contrary to the spirit of the Constitution, they were equally concerned that their actions were seen as having a constitutional basis. Army leaders consistently advanced the

rhetoric of constitutionalism to justify their extraordinary actions. Singirok, for example, claimed that “We’ve allowed the democratic process to take place ... I was exercising my constitutional rights as a Papua New Guinea citizen”. His ally, Major Walter Enuma, asserted that the army was actually exercising its constitutional obligation to intervene in the crisis: “we fought against corruption, but the Constitution of the country must be upheld” (‘Army says it will now take a back seat’, Sydney

Morning Herald, 27 March 1997). Moreover, and more importantly, the crisis was

settled by constitutional means, with Chan stepping aside from the prime ministership pending an inquiry into the hiring of the mercenaries, and an acting prime minister, John Giheno, taking his place until Chan’s return just prior to the 1997 national elections. At the 1997 election, Chan and another former prime minister, Paias Wingti, both lost their seats, along with most of Chan’s cabinet.

Political parties in PNG

While PNG stands as a relatively rare example of democratic longevity in the developing world, its party system appears to be heading in the other direction, towards fragmentation and ultimately, perhaps, dissolution. The PNG party system is unstable (no government since independence has survived as elected for a full parliamentary term), fragmented (there were 20 registered parties prior to the 1997 election), highly personalised (parties tend to operate as parliamentary factions, based on one or two dominant personalities, rather than as coherent, broad-based vehicles for translating public preferences into government policy) and increasingly irrelevant (the largest ‘party’ at each of the last two elections has been independents, who have won over 50 percent of the vote on each occasion).

Rabushka and Shepsle have argued that this type of weak party system is typical of ethnically-fragmented societies, which are characterised both by the presence of many groups and the inability of any of them to dominate the political process. In post­ colonial situations, where the rewards of political success become a valuable prize, parties proliferate but multi-party coalitions become difficult to form and hold together: “effective party politics ... does not usually emerge in the fragmented setting; no party is large enough to rule and the multiplicity of culture groups frustrates any attempts to form long-run multiethnic coalitions” (1972, 178). But the PNG case does not support

their conclusion that the absence of effective brokerage institutions such as political parties leads, in such cases, to the breakdown of democracy. In fact, the comparative literature appears to be largely silent on how to interpret instances where continuous democracy seems to coincide with the lack of a meaningful party system, such as appears to be the case in PNG.

Many political scientists appear to view a meaningful party system as a sine qua non of representative government, and often identify a functioning party system with democracy itself (Strom 1995b, 924). Strong party systems, some contend, are both reflections of and indispensable prerequisites for “good democratic performance” (Powell 1982, 74). Diamond sums up the prevailing view of many scholars, arguing that

one o f the most important institutional arenas for democracy is the party system ... political parties remain important if not essential instruments for representing political constituencies and interests, aggregating demands and preferences, recruiting and socializing new candidates for office, organizing the electoral competition for power, crafting policy alternatives, setting the policy­ making agenda, forming effective governments, and integrating groups and individuals into the democratic process (1997, xxiii).

Despite its democratic longevity, parties in PNG provide few, if any, of these aggregative and policy-related functions. Increasingly, PNG’s political parties appear to be little more than parliamentary factions. The scholarly literature suggests that this should present significant problems for the consolidation of democracy in PNG. As Powell notes, “virtually no party theorists favor fractionalized parties not linked to social groups. Multiple parties that represent only elite factions and personal followings receive few favourable reviews” (Powell 1982, 76-77). Huntington similarly argues that a key function of political parties is to present clear choices to voters and to link them closely to the political process; fractionalised and personalised systems which fail to do this are extremely damaging for democratic prospects and are, consequently, found widely in the failed democracies of the Third World (1968, chap. 7).

The empirical evidence appears to confirm these theoretical expectations. Reviewing the findings of their 26-nation study of democracy in developing countries, Diamond, Linz and Lipset found that their cases generally supported the proposition that “a system of two or a few parties, with broad social and ideological bases, may be conducive to stable democracy” (Diamond, Linz and Lipset 1995, 35). Papua New Guinea, however, proved to be an exception to this generalisation. Of the five developing-world cases

classified as ‘stable democracies’, two (Venezuela and Costa Rica) featured two-party systems, two (India and Botswana) exhibited one-party dominant systems, and only one (PNG) had a multiparty system.38 Similarly, Power and Gasiorowski’s comparative examination of democratic consolidation found that Papua New Guinea was the developing world’s only consolidated parliamentary democracy which featured a multiparty system, defined as three or more parliamentary parties (1997, 144).39 Again, the evidence from PNG appears to be a deviant case for comparative politics specialists, and a challenge to the prevailing wisdom. How do we explain this apparent conundrum?

The PNG party system

Like most other modem institutions in PNG, political parties are very recent phenomena. The oldest party, the Pangu Pati, was formed in 1967 and headed by the country’s first prime minister, Michael Somare. It attracted a group of young indigenous candidates and supporters and pushed the Australian government to increase the speed of decolonisation. A rival grouping formed as a counterbalance to Pangu and to their pressure for immediate self-government, the United Party, was supported mainly by highlanders and expatriate plantation owners. Since independence, parties have come and gone with increasing regularity, with three major parties participating in most coalition governments: the People’s Progress Party, led by Julius Chan, which draws most of its support from the Islands region; the People’s Democratic Movement, led by Paias Wingti, which is often seen as a highlanders’ party; and Pangu, which has national aspirations but continues to be identified with the Sepik region. Other parties of influence in the post-independence period include the People’s Action Party, led by Ted Diro, the dominant ‘Papuan’ party; the Melanesian Alliance led by John Momis from Bougainville; the League for National Advancement headed at various times by Tony Siaguru and Barry Holloway; and, in the lead-up to the 1997 elections, the National Alliance headed by former prime minister Somare and the People’s National Congress headed by the prime minister elected following the 1997 elections, Bill Skate.

38 Increasing party system fragmentation in recent years in two o f these cases, India and Venezuela, has