The first of these questions has been substantially answered by Hegarty (1979a): Papua New Guinea’s political independence was not the product of struggle, and in the absence of a struggle the sort of nationalist movements which emerged in other new states, and which provided the milieu in which party systems developed, were simply not evident in Papua New Guinea. ‘The development of nationalism’, Hegarty argues (1979a:188-190), was ‘inhibited by a number of factors’:

More specifically,

In a paper written in 1970, a frustrated political organiser, Michael Somare, presented a similar list of problems of political organisation. ‘The administration’, Somare said, ‘is the giver of all things and people do not care so long as they are at the receiving end. Our people are so accustomed to getting things for nothing that … they do not see why they should organise as political groups’ (Somare 1970:490).

We will return to the question of class and elites below. Meanwhile, the attitude of the Australian administration to political parties deserves some elaboration. There is no doubt that until quite late in the colonial period the administration was antipathetic to any mass movement of a ‘political’ nature – and that could include anything from a ‘cargo cult’ to a local government council breakaway group. In part this was a defensive response to forces which were seen as potentially disruptive of the colonial order. In later years it was increasingly a reflection of anxiety over possible threats to an orderly progression towards national unity and a stable, independent state. Wolfers records that Special Branch police ‘were regularly to be observed taking notes’ at meetings of political parties in the 1960s (Epstein, Parker and Reay 1971:30; see also Wolfers 1970:445) and as late as 1967 (as steps were being taken to establish indigenous political parties) the then minister for Territories, Charles Barnes, was reported as saying, ‘At this stage of its development the Territory would be better off without parties’ (Canberra Times 23 June 1967). During the 1968 elections the derogatory statements of the minister and of some administration officers gave strength to the suspicion and hostility with which political parties were generally regarded (Epstein, Parker and Reay 1971:337) and even in the months prior to the 1972 elections, according to Stone (1976:51), ‘Administration officers outside the House continued to foster the attitude that parties were detrimental to the country’. (See also Somare 1970, 1977.) Nevertheless, in a political education programme mounted before the 1968 election, and continued after it, the administration commended political parties (specifically commending two or three parties over many or one) (see Epstein, Parker and Reay 1971:373-375; May 1976 [chapter 5 above]) and in 1972, after a visiting UN mission had recommended that parties be promoted on a nationwide basis (it expressed fears that differences between the existing parties might solidify on a purely regional [coastal/highlands] division, UN 1971:66), the administration distributed, as part of its political education programme, a booklet on political parties which contained the platforms of the three major parties. Indeed, by the early 1970s it might be said that the administration was propagandising for the institution of political parties at a time when some well informed Papua New Guineans were arguing against parties as being potentially disruptive. Ironically, those who in the late 1960s and early 1970s might have appeared pro-gressive in their support of political parties are now liable to find themselves condemned as having been the apologists for a conservative and essentially ethnocentric doctrine of political ‘modernisation’.

Inhibiting factors aside, however, mass-based political movements did emerge in the pre-independence period. Amongst the various political organisations to appear on the scene prior to the elections for the second House of Assembly in 1968, two – the United Christian Democratic Party (later United Democratic Party – UDP) and the Pangu Pati – might be described as the first indigenous, mass-based parties. The UDP, however, proved to be shortlived. Formed in the East Sepik, where it was identified with the Catholic mission, and ideologically conservative and parochial, it was unable to attract significant support outside the province and faded away after a disappointing showing in 1968. Pangu, of course, was, and has remained, more successful. But it is perhaps worth noting that although the founders of Pangu (themselves an educated bureaucratic elite) represented a broad geographic cross-section of the population, the mass base of the party was (and has remained) geographically concentrated – in the East Sepik and Morobe provinces, Port Moresby and, at least in the early years, North Solomons. Moreover, the most effective organisational network which the party has had – the branch system in Morobe which Toni Voutas developed in 1966 – was based on individual village organisations with little or no coordination below the provincial level (Voutas 1970).

In 1968 parties were evident, but their influence on the election was minimal:

Nevertheless when the second House of Assembly sat in 1968, 10 of its 84 members were supporters of the Pangu Pati, which duly declared itself to be the ‘loyal opposition’.

During the course of the second House two other mass-based movements appeared which subsequently contested elections and have been described by both Wolfers (1970) and Stephen (1972) as political parties. These were the Mataungan Association of East New Britain and Napidakoe Navitu of the North Solomons. But though both movements fielded candidates in 1972 (and the Mataungans again in 1977), these movements were not political parties in the usual sense of that term, in that they were not formed to contest elections. Nor were several other, ‘micro-nationalist’ movements which appeared about this time, even though some of their leaders did contest elections (see May 1982:429-30).

Before the end of the 1968-72 House of Assembly, however, three organisations emerged which clearly were political parties. The first of these, the United Party had its origins in an Independent Members’ Group (IMG) established in the House in 1968 amongst a group of members brought together essentially by their opposition to Pangu’s demand for early independence. The group consisted largely of highlands members together with some of the more conservative expatriate members. During 1968-69 some attempts were made by members of the IMG to create local groups to support a political party centred on the IMG and in early 1970 the formation of a coordinating body, Combined Political Associations (Compass) was announced. Later that year, at a meeting in Minj attended by some members of the IMG and by local government council leaders and representatives of other highlands groups, a Compass organisation was established and a chairman and secretary appointed (both were highlanders); in early 1971 Compass changed its name to United Party (UP). The circumstances under which Compass was established, however, and specifically its dominance by highlanders resulted in the emergence of a second party from within the IMG. In mid 1970 a Business Services Group, under the leadership of Julius Chan and comprising mostly members from the New Guinea Islands, was formed within the IMG. Following the Minj meeting this group founded the People’s Progress Party (PPP), a move which, in the words of Loveday and Wolfers (1976:21), ‘seemed to represent a regional distrust of the highlands leadership implicit in Compass’. The association of highlanders with a ‘go-slow’ attitude to independence, which Compass represented, also prompted the formation of the New Guinea National Party (NP) amongst a group of generally younger and more progressive highlanders. To quote Loveday and Wolfers again, the NP ‘was soon regarded as the highlands equivalent of Pangu’ (ibid.).

The 1972 election was thus, for the first time, contested by parties. About 150 of the 611 candidates who nominated were endorsed or selected and helped by parties (Loveday and Wolfers 1976:74), though some candidates were still hesitant about publicly admitting party membership. Some electorates fielded more than one ‘endorsed’ member from the same party, and no party had a nationwide organisation.

In the event no party emerged from the 1972 elections with a clear majority and, notwithstanding the expectations of the UP (as the largest group in the 1968-72 House), Pangu leader Somare was able to cobble together a coalition government which embraced Pangu, NP, PPP and the Mataungan Association. The UP accepted the role of opposition (though in 1975 some UP members supported the government on critical divisions).

This party alignment was broadly maintained during the life of the 1972-77 House. There were, however, at least four developments of some significance with respect to political parties. One was the formation in 1974 of a Country Party, whose supporters were recruited primarily from amongst UP, highlands members. The Country Party, however, had little effect on members’ voting patterns (Loveday and Wolfers 1976:91) and appears to have faded away by the time of the 1977 elections. A second development was the emergence of the Nationalist Pressure Group (NPG) in 1974. The NPG represented a coalescence of members who supported the proposals of the Constitutional Planning Committee against the modifications put forward by the government (see Hegarty 1975). But although it voted as a cohesive group on ‘national’ issues in 1974-75, the NPG specifically avoided the label ‘party’ and its 18 core members – drawn from the four major parties plus the Mataungan Association and the newly-formed Country Party – retained their party affiliations. A third development was the split and virtual collapse of the NP in 1976, after Somare had dismissed from cabinet its leader and deputy leader, and a move by them to withdraw all NP members from the coalition had failed. The fourth, which foreshadowed the emergence of Papua Besena as an electoral force in 1977, was the election in a by-election of 1976 of a second Papua Besena member and the subsequent announcement of a Papua Party (McKillop 1982).

In 1977 the party mass organisations, which had generally atrophied since 1972, were revived for the country’s fourth, and first post-independence, election. This time, of the 879 candidates who contested the 109 seats, 295 (30 per cent) were endorsed by one, or more, of the three major parties (Hegarty 1977:457). In addition, a number of Papuan candidates stood for Papua Besena, which in 1977 appeared to have evolved from an ill-defined separatist movement to a fully-fledged political party. Observers of the 1977 poll seem to have been generally agreed that political parties had a substantial impact on the election (Hegarty 1977, 1983), though in an interim report on the election Standish (1977:4) concluded that while in the towns competition ‘was more in terms of modern associations’, in rural areas ‘clan voting prevailed’. Nevertheless, in 1977 as in 1972 uncertainties about the political allegiances of some candidates resulted in intense post-election lobbying amongst those who hoped to be able to put together a government. One proposal was for a ‘National Alliance’ comprising former governor general Sir John Guise and members of UP, Papua Besena, Country Party and NP. Another was for an Islands-based Alliance for Progress and Regional Development, led by the two former NPG spokesmen, Momis and Kaputin.

In the event, the successful combination was a coalition of the enlarged Pangu and PPP membership with most of the Mataungan and North Solomons members. After several months of dispute within the opposition over the leadership performance of the UP’s Sir Tei Abal, in 1978 former NP minister Iambakey Okuk (who had been reelected in Simbu with UP support) emerged as opposition leader. After attempting unsuccessfully to bring together his highlands supporters, Papua Besena members and some others in a People’s United Front, Okuk revived the NP and as its leader waged an aggressive campaign against the coalition.

In November 1978, following a growing unease in the relationship between PPP and Pangu (which had probably more to do with personalities and leadership styles than with policies) PPP withdrew from the coalition. Pangu was maintained in office, however, by a split within the UP which brought about half of that party’s members across the floor to the government. In 1978-79 the Somare government survived three no-confidence motions initiated by Okuk, but in January 1980 both Momis and Kaputin withdrew from the coalition, forming a new party, the Melanesian Alliance (MA), and two months later, with their support, a no-confidence vote against the Somare government succeeded. Chan became prime minister as the head of a National Alliance government comprising PPP, NP, MA, Papua Besena, and part of UP. The Alliance was able to hold on to office until the scheduled elections of 1982, but it was, to say the least, an improbable coalition. PPP and NP, broadly aligned in support of capitalist development and foreign investment (though with little personal empathy between Chan and Okuk – see, for example, Post-Courier 21 June 1983), were at one end of a political spectrum from the MA, which regarded itself as being to the left of Pangu and whose leaders were strongly identified with eco-nomic nationalism and the aim of self-sufficiency; and Papua Besena, which owed its origins in large part to fear and distrust of highlanders (Daro 1976; McKillop 1982), was a strange bedfellow for a coalition in which highlands members were a large component and whose deputy leader (Okuk) was a staunch highlands nationalist.

When in 1981 I visited East Sepik and Morobe provinces even Pangu party organisation appeared to be in a shambles. Nevertheless, as in previous elections, party organisations were resuscitated in late 1981-early 1982 and several new groupings appeared on the scene; indeed in the 1982 elections parties seemed to be more salient than ever. Pangu, PPP, UP, NP, MA and Papua Besena/Papua Party all fielded candidates, while two new groups – a Papua Action Party (which had links with the NP), and a predominantly-Papuan ‘Independent Group’ headed by former Defence Force commander, Ted Diro – emerged as significant contenders. About 60 per cent of the 1125 candidates who stood in 1982 were endorsed by one or more of these eight parties (Hegarty 1982). My own observation of the 1982 campaign in the East Sepik suggested not only that nearly all candidates sought a party label (some, indeed, more than one) but that a high proportion of voters could accurately attach party labels to most candidates. As against this apparent evidence of the arrival of parties as an important political force, however, there remained some indications that perhaps things had not changed as much as might seem. For one, party attachment for most candidates seemed still to be loose and it was not rare for a candidate who failed to get endorsement or assistance from one party to turn to another. Second, for some parties and in some electorates party attachment meant little more than the use of a label. Third, in a number of instances party members stood against endorsed candidates of their own party against their party’s interests (though in some cases parties – especially Pangu – supported more than one candidate in order to split the local vote of opponents of their endorsed candidate). Finally, although the available evidence does not permit a strong conclusion, it would appear that while there was in 1982 some increase in party voting, personal and local loyalties were still considerably more important for the great majority of voters (Hegarty and King 1982).

The outcome of the 1982 election was a victory for Pangu, which – apart from the recently established MA – was the only party to increase its representation in the parliament. In August 1982 Somare was duly reelected to the prime ministership, heading a government comprising successful Pangu candidates, UP members and a number of members who were either elected as independents or switched from other parties after the election. Although Momis was the opposition’s nomination to stand against Somare in the initial vote, he was subsequently passed over as opposition leader; this position was given to Diro whose group had thrown its weight in with the NP. But when in 1983 Okuk – who had lost in Simbu in 1982 – was returned in a by-election, Diro stepped down in his favour.

From even so brief an account as this, four characteristics of political party development in Papua New Guinea are apparent.

First is the fluidity of party alignments. Amongst the major parties which have contested elections since 1968 (that is, leaving aside a number of small parties not mentioned here which have flashed briefly across the political scene) only Pangu, and PPP have had a continuous, cohesive existence lasting over several years. The UDP disappeared, within about two years, soon after its failure in the 1968 elections; the UP, the largest group in the 1968-72 parliament, split in 1978 and appears to be on the wane; the NP has come and gone and come again, its future probably closely tied up with the political fortunes of Okuk; the Country Party proved to be short-lived; Papua Besena/Papua Party enjoyed at least one reincarnation but has been faction-ridden for most of its existence as a party, with its founder, Josephine Abaijah, denying that Papua Besena is a party; the Papua Action Party appears to have dissolved after its failure at the polls in 1982; and the Melanesian Alliance has yet to prove its capacity for survival, though in 1984 it seems likely to remain a substantial political force. Moreover, apart from the question of candidates holding multiple party endorsements, in 1982, as in 1972 and 1977, the real test of party loyalty came after the elections in the politicking to get the numbers to form a government. Thus in 1982 the MA lost two members to Pangu after the elections, UP lost three, and in the vote for the prime ministership only four out of seven of Diro’s Independent Group voted with him.

A second notable feature of parties has been their general lack of effective mass organisations. Although most of the larger parties have, on paper, organisational structures based on party branches, in fact the majority have their origins in parliamentary alliances and remain dominated by parliamentary members, and between elections party organisations in the electorates tend to atrophy. The main exception to this – apart from the Mataungan Association and Papua Besena, which began as popular movements without the specific objective of contesting elections – is probably Pangu, but even it is essentially a parliamentary party except when elections are approaching.

Third, as various commentators have observed, parties in Papua New Guinea have not been particularly concerned with ideology. In the period before independence, Pangu (together perhaps with the NP) was differentiated from the other parties primarily by its critical attitude to the Australian administration and its demand for early independence. By 1974-75, however, in the interests of political stability Pangu was in the position of pressing the Australian government to delay the granting of independence, and with the achievement of independence in 1975 this ceased to be an issue. The UP and PPP, on the other hand, were generally regarded as more ‘business’ oriented and more favourably disposed towards foreign participation in the economy, though in practice the differences appeared not substantial, as the record of the 1972-1978 coalition government indicates. (Hegarty, in fact, speaks of a general ‘pro-capitalist con-vergence’ (1979a:199-201).) On the one occasion that substantial differences on important policy issues did arise – namely during the constitutional debates of 1974-75 – alignments cut across party lines. The split which occurred within the coalition in 1978, on the other hand, seems to have had more to do with personal relationships and political styles than with any ideological differences. And the nature of the coalition which replaced Somare in 1979 – accommodating as it did such widely different viewpoints as the ‘open economy capitalism’ of Chan and Okuk and the economic nationalism of Kaputin; the highlands nationalism of Okuk and others and the (anti-highlander) Papuan nationalism of Papua Besena – suggested further that questions of ideology were secondary to considerations of parliamentary office, a suggestion reinforced in 1978 by the split of the UP, and in 1977 and 1982 by the lobbying which followed the polls.

Fourth, even leaving aside those movements and parties (such as the Mataungan Association, Papua Besena, the Papua Action Party) which were specifically locally or regionally focused, and a number of small parties in the late 1960s and early 1970s which never grew beyond their local origins, there has been from the inception of parties a clear regional orientation in political party organisation. As has been noted, the UDP never managed to establish itself beyond the Sepik; the UP, NP and Country Party have drawn their support predominantly from the highlands; PPP and MA have been largely Islands based; even Pangu has little party organisation outside of East Sepik, Morobe and Port Moresby (though the support it received in the highlands in 1982, and its association with the UP, may change this). Mention might also be made in this context of the Morobe District People’s Association (MODIPE), a provincial organisation which was established in 1973 with the specific objective of preventing people from outside the province becoming parliamentary members for Morobe electorates (at this time the Pangu member for Lae was a Papuan). The visiting UN mission of 1971 expressed concern at the regionalist tendencies in political party development (UN 1971:176). Stephen (1972:175) also observed, ‘Pangu and the United Party are regionalistic … The smaller parties are regional too’. Writing in 1972 Waddell (1973a:96) forecast that ‘it will not be ideology or class interests which separate the parties – if there are more than one … regional interests are the most likely source from which political parties will derive their mass base’. Commentaries on the 1977 election tended to support this judgement. Hegarty (1977:454, 461) observed that in the pre-election period ‘considerable social differentiation had become apparent’, but went on to conclude, ‘The basic cleavages in PNG politics are not ideological or class based but regional’. And Premdas and Steeves (n.d.:35) ventured the opinion, ‘It would be difficult for anything but an ethnically-based party system to emerge’, and figures of candidates by party and region provided some basis for such a view. By the time of the 1982 elections the regional concentration of party support appeared to have been diluted somewhat. Jackson and Hegarty (1983), in fact, entitled a paper on the 1982 elections, ‘From geography to ideology?’ Nevertheless there was still evidence of a regional element in party support, as shown in the following table.

Moreover, while the PAP and Papua Party campaigned for Papuan solidarity and a Papuan prime minister, in the highlands the NP complained that the government had been dominated by coastals, and called on voters to help elect a highlander prime minister.

The introduction of provincial government, commencing in 1976, created another level of political institutions and raised the possibility that provincial elections might stimulate the development of existing party organisations at provincial level.

In fact, and although stories vary from province to province (and to date there has been little documentation of provincial politics^[80]), two patterns seem to be common. One is a tendency to reject parties at the provincial level: within several provinces (amongst them Eastern Highlands, Western Highlands, East Sepik) the view has been expressed that parties are ‘disruptive’, are ‘not appropriate’ for provincial assemblies and provincial elections (see, for example, the statement by the East Sepik Provincial Minister for Commerce, reported in the provincial government’s Wama Nius 5(4), April 1983). The other has been the discouragement of the projection of ‘national’ parties into provincial politics has been a frequent antipathy between national politicians and provincial politicians (even those identified with the same party). This antipathy is rooted in uncertainties about the respective roles of national and provincial assemblies, and in a common fear on the part of national politicians that provincial members (seen by at least one highlands politician as failed national political aspirants) undermine their power bases in the province. That is not to say that factions have not emerged in provincial assemblies, nor even to deny that such factions are sometimes identified with parties (for example, Pangu, and non-Pangu factions in East Sepik Province). But I would suggest that where systematic divisions, other than limited personal followings, have emerged in provincial assemblies, the basis for such divisions has more commonly been regional. Thus, for example: Standish (1984) has described the 1980 provincial and 1982 national elections in Simbu in terms primarily of a conflict between Bomai (southern) and northern regional blocs in the province; in the East Sepik an emerging division between the people of the river, the people of the coast and islands, and those of the hinterland has, to a degree, been formally accommodated by writing into the constitution specific provision for area ministers (representing Ambunti and Angoram, Wewak, and Maprik); in the Southern Highlands Ballard (personal communication 1982) has described provincial politics in terms of differences between three main regional blocs – Mendi, Ialibu and Tari. And it seems that similar patterns are discernible elsewhere (Central Province? Morobe? Eastern Highlands? New Ireland?).

Moreover, I believe that the creation of provincial governments, and more recently the emergence of regional provincial political organisations, have tended increasingly to ‘provincialise’, or at least to ‘regionalise’, national politics, a factor which is reflected, for example, in the allocation of cabinet portfolios and of senior public service positions.

If the foregoing has established that Papua New Guinea does not have a ‘developed’ party system, based on fundamental social cleavages, the big question remains: is this situation likely to endure or is it simply a reflection of Papua New Guinea’s present, transitional, ‘stage of political development’?

Several commentators have suggested the latter, generally implying that a two- or three-party system will develop as social classes emerge (see, for example, Jackson and Hegarty 1983). Such a view is in accord with the general thrust of the political development/‘modernisation’ literature of the 1960s and early 1970s. It is also consistent with the mathematics of competitive representative government (for example, see Riker 1962; Groennings, Kelley and Leiserson 1970), which suggests the formation of coalitions to aggregate the preferences of like-minded people.

As against this, the view of political development as an inevitable, evolutionary, process, in which competing political parties play an essential role in organising mass participation in the processes of social choice – and more particularly the suggestion that there is a ‘natural movement of societies … towards the two-party system’ (Duverger 1954:21); see also Huntington 1968:423-433) – has not been borne out by the historical experience of the ‘new states’, nor even perhaps of developed states in which social stratification is clearly articulated. Over the past two decades there has been a pronounced tendency towards authoritarian regimes, in which the functions of interest articulation, interest aggregation, recruitment of a political elite, political socialisation, and so on have been performed by one-party or dominant-party structures, by the military, or by the bureaucracy. There has also been a growing recognition that, where they do exist, political parties in developing states are often, to quote one commentator, ‘little more than coalitions of elites and, at their outer reaches, complex sets of highly personal face-to-face relationships, all momentarily integrated by access to government and its patronage’ (Heeger 1974:70). This pattern – the predominance of personal and factional (kin, ethnic or regional group) politics – may be particularly appropriate to the politics of small-scale, as well as poorly integrated, societies (cf. May 1982; May and Tupouniua 1980).

Addressing the question, ‘what sort of a party “system” could emerge’ in Papua New Guinea, as early as 1966-67 Robert Parker warned that the constitutional, democratic party system, represented by Australia, was rare, and that there was no basis for expecting such a system to develop in Papua New Guinea. Recognising the general tendency to one-party states elsewhere, and in the light of his analysis of interest groups, issues, communications and potential leadership in Papua New Guinea, Parker went on to suggest that there appeared to be ‘powerful considerations favouring the emergence of [a “democratic one-party” system] after independence’ (1967b:15; see also 1966-67, 1967a). In a later paper, Waddell (1973a) also appears to have seen a one-party regime as a likely development, though he himself advocated ‘a co-operative government in which all major, and perhaps even minor, parties are automatically involved’ (1973b:30). Premdas, on the other hand, arguing against a ‘no-party state’, believed that ‘a wide range of party systems is available from which a choice can be made’ (1974:130; see also 1975).

In 1984 it would appear that a one-party, or no-party, regime is unlikely. Although the ‘party system’ has not been particularly stable, parties have been in evidence now for some time and a small number has survived for over a decade. The answer to the question posed in this section would therefore seem to be that, while mass-based, issue-related political parties are not logically essential to the working of a parliamentary system like Papua New Guinea’s, some sort of political coalescing probably is inevitable and in Papua New Guinea parties of one form or another now seem to be a well established feature of the political landscape. Such an answer, however, still leaves us with the other big question, namely what are the likely bases for the articulation of party interests?

Amongst those who in the 1960s and early 1970s anticipated the development of political parties, both in new states generally and in Papua New Guinea, there seems to have been a presumption that, as in Europe (and Australia), the development of parties would be associated with the articulation of social classes and the emergence of a nationwide socio-economic stratification which would supersede politically other divisions of society such as region or ethnicity. Indeed political parties were often seen as agents of integration and national unity. (In the Papua New Guinea context, for example, Premdas [1974], in arguing for political parties, listed integration and unity as one of four ‘critical areas’ in which parties had a role to play.)

Three assumptions seem to be implicit in this view: (i) that the emergence of social classes is inevitable; (ii) that class divisions are generally a more likely basis for political parties than other divisions in society, such as ethnicity or region; (iii) that political parties must relate to certain fundamental divisions of society and be differentiated from one another on the basis of such divisions.

The first of these assumptions is discussed elsewhere (see May 1984); I will not buy into the debate here except to offer the opinion that even though there is evidence of emerging social classes in the rise of ‘big peasants’, the growth of a largely- urban ‘salariate’ or ‘bureaucratic bourgeoisie’, and perhaps the emergence of a rural and an urban proletariat, the prospective picture of social class in a predominantly rural, fragmented society like Papua New Guinea’s, in which traditional elements of social structure are still important, is more complex than that which is offered by a simple Marxist model (cf. Hyden 1980).

Second, on the evidence to date from Papua New Guinea, and having regard to experiences elsewhere (in Europe and North America as well as in the ‘new states’), I am far from convinced that class divisions, even where they are clearly articulated, must necessarily outweigh ethnic or regional differences as a basis for political organisation.

Third, to the extent that Papua New Guinea’s political parties conform to what Heeger sees as often little more than coalitions of elites – and on the basis of the description presented earlier in this paper I think they do – the question of a mass base becomes relatively unimportant; it might then be argued that regionally-focused patronage systems are a more effective way of gaining electoral support.

Finally, to have a small bet each way, it might be pointed out that in Papua New Guinea socio-economic status and region are not entirely without correlation: capitalist penetration has probably been most extensive in the Islands and the Highlands regions, which have also been the primary areas of support for the generally pro-capitalist UP, CP, reorganised NP, and PPP, though the more recently contacted – and thus in some respects ‘less developed’ – highlands provinces are historically most closely associated with the relatively conservative (initially pro-administration) UP and CP; the less advantaged Mamose (North Coast) region has predominantly supported Pangu, while Papua Besena/Papua Party and PAP have in part sought to protect the interests of a largely Papuan urban salariate. Moreover one can think of tendencies which could strengthen such a correlation: for example, growing regional disparities in levels of economic development (cf. Berry and Jackson 1981), competition over central government funding of provincial governments (cf. May 1981), administrative breakdown at the centre, increasing wantokism within the bureaucracy.

For some time commentators on Papua New Guinea have spoken of the imminent emergence of a social class system and have frequently suggested that as social classes finally emerge the country’s currently fluid party ‘system’ will solidify along class lines.

In this paper I have attempted to suggest that there are no good grounds for regarding Papua New Guinea’s present party system as ‘transitional’ between an ‘undeveloped’ and a ‘developed’ system, and that even if there is an articulation of social classes which cut across regional divisions there is no necessary reason to suppose that parties will follow the class divisions. Ethnicity and regionalism are significant and convenient bases for political organisation in Papua New Guinea and are likely to remain so. To the extent that political parties remain essentially parlia-mentary organisations, coalitions of political elites, the more successful undoubtedly will widen the geographic base of the coalition but I believe this is more likely to be reflected in the linkage of personal and regional followings than in appeals to what some observers see as emerging social classes.